For several years, and then more recently had the pleasure to take on the role of running the software business. Just to level set everybody, Broadcom started out obviously as Avago. We did a number of acquisitions in the semiconductor space. We then added on to the portfolio by buying Brocade, which is in the Fibre Channel SAN switching business. More recently in 2018 and then in 2019, bought CA and Symantec. What we're gonna talk about today in the business that I'm responsible for is CA and Symantec. We have two reporting segments I think most of you know, Semiconductors and Infrastructure Software. Infrastructure Software includes Brocade. What we're gonna talk about today actually excludes Brocade, so this is really CA and Symantec. CA closed in November 2018, funny enough, about three years ago, and Symantec actually closed one year later in 2019.
With that, just at a glance, this is a business that does $5.2 billion of annual recurring revenue. We actually are very much focused on serving the largest enterprises in the world. The vast majority of our business comes from the Fortune 500. Our investment thesis is all around R&D operations support. 80% of our workforce is actually totally focused in those areas. You can see that we spend about 14% of our revenues on R&D on an annualized basis. We have a lot of technology and a lot of IP that's been developed over a number of years. We're continuing to invest in our products and our roadmaps. With our model, given that we're an incumbent, we have a long operating history. Been around a long time. Our customer satisfaction scores are really important.
That ties to our support, and we're running in the high 90s%, which is great. I'm gonna spend a bunch of time on this slide. I think this really helps articulate our business model and how we take what we think are fundamentally undervalued companies, buy them from the public market and actually turn them into great businesses, great cash flow generation areas where we can actually sustain and grow revenues over time. We are focused on the largest enterprise customers. You know, what does that really mean? What it really means is we're looking at businesses and customers that are fundamentally actually on-prem first. These are companies that are investing significantly in their own private data centers. They have a long history of operating those data centers, and then we're seeing them expand dramatically their footprints actually on an on-prem basis.
It's not that they don't leverage the cloud, they certainly do, and they're moving workloads to the cloud every day, but they're very much hybrid, and we think they're gonna be hybrid in perpetuity. That's really the focus of our business model. Beyond that, there's a tremendous amount of complexity and heterogeneity in terms of their environments. That means the existing infrastructure that they have, what they're doing in the cloud, all actually tailors toward the incumbent. I'm gonna go into all that in a lot more detail, and then the three GMs are gonna talk about, you know, what they've been doing in the past, but also where their roadmaps are going forward to support those organizations. Typically, they're also very regulated risk-averse companies, big large IT budgets.
Those budgets are expanding actually quite rapidly, which is great for our business fundamentally, but they're also quite risk-averse. You know, what does that mean? They also have a lot of legacy environments. They have new workloads. When you look at all of that, it really does favor the incumbent, which is really CA, Symantec and what we're doing going forward. Our go-to-market model is very much focused on these customers. Who are these customers? 600 of them actually today. We have a highly strategic focus, and we'll get into what that means for the financial model, but it means basically our go-to-market and our selling costs are dramatically lower than a typical software business.
If you look at the stats, 70% of the ARR that we drive, that $5.2 billion I mentioned, is actually from these 600 strategic accounts. They all have the characteristics of the businesses I talked about on the prior slide. 96 of those accounts, when we think about scale, actually have $10 million or more of recurring revenue a year. Actually more than 600 of our accounts do more than $1 million a year of ARR. We organize ourselves from a model standpoint, both with the business units as well as our go-to-market, totally focused on the priorities of these 600 strategic accounts. We obviously have different levels of focus between platinum, gold, and silver in terms of our go-to-market. I won't spend a lot of time on that.
Then behind that, we actually have a long tail of enterprise accounts, which are good customers. We're very much focused on making sure the products that are installed are current, and we're also focused on making sure we deliver great customer support. That makes up the bulk of the rest of the business. In the long tail, which we call commercial, this is a small part of our business today. This is a part of the business that's attriting over time. Our objective is really to grow with the strategic accounts, sustain with the enterprise accounts, and then let the commercial accounts attrit, as I discussed. Thinking about the strategic customers. When we think about how do we grow from here, $5.2 billion of ARR, we really are looking at a couple of things that fundamentally drive growth.
One is the fact that we actually have a tailwind from a market perspective. The markets we're in are growing infrastructure, security. Just as a level set, infrastructure is about a $3 billion business for us. Security is about $2 billion. So just to give you an idea. These markets, you know, when we'll get into this in a lot more detail through the GM presentations, they're growing. We're seeing our customers who have very large IT budgets, and these budgets are typically $1 billion plus. Some of them are $10 billion plus. They're really not as concerned about the budget, but they're mostly focused around vendor consolidation capabilities and what they can do with their incumbent solutions. What we've seen is markets growing actually relatively quickly, especially in these areas.
On top of that, there's a bunch of organic things we can do. You can see the stat here. 80% of our customers on the strategic side actually are already licensed to five plus of our solutions. What we see is two fundamental opportunities. One, we can expand. We can expand our footprint within these accounts in terms of under deployment or vendor consolidation, which with software that's already being used today. The second area where we've been focused is actually on cross-selling and upselling. We can actually take a customer that has three or four of our solutions and then add on solutions that they don't already use today. We have a very focused enterprise license agreement approach, ELAs, you might hear that term, where we're actually providing multi-year contracts where they can take advantage of our solutions across the portfolio.
We're really just at the beginning of that. We know Symantec, we've just integrated and rationalized that business over the last couple of years. CA, we've been running the business for three years, but this is really the upside opportunity in terms of driving growth. It's the market's growing. We can improve our installed base with what we already have embedded in these accounts, but then we can also upsell and cross-sell. We're really looking forward to that. With that in mind, what I wanted to do is actually take a look at some stats. We haven't done a lot of disclosure to date, but I think this will give you some insights into how we've been doing the last year.
These are things that we're gonna track and look at going forward and hopefully be able to report out to you on a more regular basis. But if you look at, you know, large strategic deals, this is a sort of key indicator for us. We've been driving, you know, 330 deals above $1 million over the last 12 months this year, which is up sharply from where we were a year ago, running about 285 deals above $1 million. Now, how does that translate into bookings? Well, you can see we've actually booked, you know, on a trailing twelve-month basis, about $1.8 billion of bookings across those deals relative to the $1.1 billion at the end of last year.
What we really focus on is large customers, where we're the incumbent, where we see opportunity to improve, you know, our installed base, where we see opportunity to cross-sell, and that really translates into larger deals, larger number of deals, and larger deal sizes. This is a key indicator for us. Similarly, we've been very focused on driving multi-year bookings. You know, part of our model is being able to provide visibility not only for ourselves, but for our customers in terms of their budgets and their ability to expand their footprint without having to actually come back to us necessarily and buy more within the timeframe of the contract. You can see how we're driving strategic deals with these accounts.
If we go to last year where we were still integrating and rationalizing Symantec, we're running about a year and a half average deal duration. Today, we're actually all the way up to 2.3 years. Big improvement, not just in deal size, number of deals, but also the duration of the contracts that we're driving. You can see this, you know, trending more towards three years is sort of the optimized amount. I think given the mix of enterprise and some commercial accounts, it'll end up being about 2.5 years. We made a tremendous amount of progress on this front. We're looking forward to continue to do that. Looking at ARR, I know this is an important statistic. It's something that we track regularly and something that, you know, we wanted to report out on today.
You know, when we exited last year, ARR was sub $5 billion, and our total renewable backlog, which ties to the duration of the contracts, was running about 12.9 years. Big improvements here. We still have a lot of work to do, but ARR is actually up to $5.2 billion exiting Q3. If I annualize that's nearly 6% growth on a year-over-year basis. I think as importantly, the renewable backlog is actually up to $14.4 billion, was actually up double digits year to date. Our focus right now is leveraging the market tailwinds, leveraging the fact that we have a lot of upsell and cross-sell opportunities. We continue to improve the amount of usage from an installed base perspective. You know, we think we can grow ARR mid-single digits consistently over time.
We also think we'd continue to improve on that total renewable backlog, which obviously improves the visibility in the business, lets us really, frankly, focus on customer success. Let's focus on making sure that we're doing a good job for our customers from a support standpoint, but make sure we're also driving adoption that the customers value across the organizations. The GMs are gonna talk a lot more about that within their individual business units. Finally, I think just as importantly, we're also focused on quality of revenue. You know, I think when we bought CA, and we certainly bought Symantec, there was still a pretty significant mix of perpetual license models, as well as term license models, as well as some of the SaaS-based cloud businesses they had.
If you look at last year when we exited the year, you know, we were running actually about 43% subscription relative to maintenance. What we really focused on is we've stopped selling perpetual licenses. We're totally focused on selling subscription license on a multi-year basis. You've seen a big improvement on that front as well. Today, 53% of our mix is actually subscription with 47% on maintenance. I expect this to continue to meaningfully improve over time. In fact, I think subscription should be 90%+ of our business over the next several years. To give you an idea, subscription bookings are actually up 25%+ on a year-over-year basis this year. Big focus on quality of revenue, translating older maintenance-based contracts into subscription-based enterprise license agreements.
I think this will again improve the quality of the revenue, the quality of the business as we look forward. Finally, on the software financial model. I've talked a lot about the business model. I've talked about how, you know, we're able to take businesses, rationalize them, integrate them. Fundamentally, though, you know, beyond just the revenue growth and how we're driving revenue growth, you know, what we're really doing is we're reinvesting in the R&D. We're reinvesting, you know, these businesses we bought were running about 17%. We're running 14%. You know, the discrepancy there is just the focused approach. You know, our priorities are all around these 600 strategic accounts.
That allows us to focus our R&D resources in terms of their priorities, so we're not as levered across a broader set of customers, including SMB customers that have different needs and drive more R&D investment. Nonetheless, we're still investing at a pretty meaningful clip. If you look at sales and marketing, though, this is where there's a dramatic difference. When you buy companies that have been incumbents at these largest enterprise accounts for, in many cases, more than a decade, sometimes more than two decades. You're already there. You're already installed. You already are basically doing renewals.
The question is, okay, as I'm driving upgrades, as I'm driving expansion, as I'm driving cross-selling, I don't have to spend a lot of money trying to land new accounts or land new products outside of my core account base. You know, in the past, when we look at software business, and this goes for a lot of companies, they're actually spending far more on sales and marketing than they are on R&D, which is very much different than how Broadcom has operated historically and how we operate going forward. You can see, we actually took a business or a set of businesses that were driving 30% sales and marketing as a percent of revenue, and we've driven that down into the single digits. I think this will actually continue to scale with time.
We see going forward, we can drive 5%+ revenue growth, very much focused on recurring revenues, very much focused on subscription-led revenues. We can drive gross margins in the mid-90s. We can continue to sustain our R&D investment, kind of in this mid-teens level, and we can drive sales and marketing, you know, over time in the 6% to 7% range. What does that mean in terms of cash flows and operating margins? Well, you know, today we're running about 70%. Now, this excludes G&A. You know, the G&A at Broadcom runs about 1% to 1.5%. You know, you can see there's not much discrepancy between operating margin here without the G&A and with the G&A.
Nonetheless, 70% on $5.2 billion+ of ARR that's growing 5% is a business that's on track to generate $4 billion of operating income here over the next couple of years. It's a great business. There's a lot of work to do. There's a lot of organic focus right now in terms of driving that growth. I think we're in a new phase. We've bought a couple of companies. We've gone through a lot of rationalization and integration, but I think we have an integrated team. We've got great leadership. And I'm really actually pretty excited about the future. Hopefully, through this little snapshot in terms of where we are, you get some insights into the quality of the business and some of the metrics.
For most of the substance, I'll turn it over to the general managers. They're gonna walk through each of those businesses for you. First and foremost will be Greg Lotko, who runs our mainframe business. I think just to take a minute on CA, I think when we bought CA, similar to our investment thesis around LSI or when we bought Broadcom or Brocade, fundamentally, we knew we had to get one thing right, and I think that was, in the case of CA, is the mainframe business. Our view was fundamentally different than the public markets. Our view was that mainframe was not only sustaining, but actually it was growing in these strategic accounts.
we saw an opportunity to reinvest and actually increase R&D and focus on the mainframe business to actually drive a lot more value. That's exactly what's happened. You know, fundamentally, through Greg and his leadership team, we have really done a great job, actually, not just stabilizing that business, but growing that business. It's a key anchor tenant in the portfolio, and it's a big part of the reason CA has been such a successful investment. With that, let me turn this over to Greg, and he can go through that business. Thank you.
Thanks a lot. It's not moving. I'm not sure if I did that or you did that, but that's okay. It happens. Think about your day. Think about today. Many of you probably used a debit card or a credit card. Maybe you bought that coffee on the way here. You might have shipped a package. You might have checked a bank balance. Odds are a mainframe has already been involved with your day-to-day. I would tell you would be hard-pressed in any given day for a mainframe not to be part of that day because of the volume of the workload that it runs around the world. I think I'm controlling. Tom talked about our team, and I want to talk about myself and our leadership team.
I was with IBM for 30 years, came to CA and then Broadcom in 2017, and across my career, more than 85% of that time, I've been involved in leading teams that are doing work on mainframe technology, whether that be hardware or software. I'd also tell you that for a little more than half of that time, I've also had distributed technologies or cloud technologies. When I talk to you about mainframe, I talk about it, to you about it in the context of overall IT. Now, for the mainframe experience, in the time I had at IBM, I've been the director, the leader of their two largest database systems, IMS and DB2. I also actually ran the hardware, the mainframe hardware, and either launched or set a strategy and direction for three generations.
I also led the storage hardware business, then brought software into it. That includes DS8K, the large disk and flash storage that's attached to mainframe, as well as the tape business. Now, there's similar experience across my direct leadership team. Across those five folks, there's more than 75 years of experience in the mainframe space and 150 years of experience in IT in general. If you go to the next layer below, the executives in our team, whether they be directors or senior directors, many of them have spent their entire career in the mainframe space. We have a deep bench and a broad wealth of knowledge about the mainframe space. Now, the usage of this platform continues to expand today. 3.5 times growth in MIPS in just the last decade.
Now, for those of you who don't know what MIPS are, it's a measurement of processing capacity on the hardware for a mainframe, and it's millions of instructions per second. 89% of the respondents that we survey saw growth in their usage of the platform in 2020, and 91% of those surveyed say, "Hey, the mainframe is a priority for them going forward." Now, this is going to remain relevant for many years to come. 92% see the mainframe as a platform for ongoing growth. 74%, almost three-fourths, see it as strategic to their overall business.
It's expected that companies that are leveraging mainframe, working in hybrid environments, think about the world as heterogeneous and connecting workloads, that there'll be a two times expansion just in the next three years of those that are leveraging their mainframe environment and connecting it in a hybrid work environment. Now, we tell you if you went back five or 10 years ago, you'd probably be hard-pressed to get IT analysts to agree on anything in the mainframe space. You know, some saw what was going on, some saw the longevity of the platform, and that it would grow depending on what was going on. We're in an unprecedented time when across all the major analysts and across the studies that you see, people are understanding how integral the mainframe is to modern workloads and connecting it across cloud, distributed, and hybrid workloads.
What a lot of folks don't realize is even today, 70% of the world's production IT workloads are still processing on mainframes. 70% of the world's workload. 88% project that the mainframe is going to have a continuing role in their environment for the next decade, and I will tell you even beyond that. Two-thirds believe they'll never fully replace the mainframe, and that's because of the unique value and the capabilities that it delivers to workloads. Mainframes outperform multi-tier processes in a single frame because of the dynamic of the architecture and how it handles workloads. Organizations that transform on the platform and invest see a 6 times return on every dollar that they invest in the platform. That's a great return on your investment.
Now you might ask yourself, "Come on, why mainframe?" You know, clearly not every workload needs to run on the mainframe. You know, a lot of the workloads that started on the mainframe started in the sixties when there wasn't another choice. I would absolutely tell you there are certain workloads that shouldn't be on the mainframe today. They should be off. When people see stats about what's going on in the world and they see the staggering volume, they question mainframe. They say, "Come on, 500 million tweets a day. Almost two billion people use Facebook on a day, and don't we all Google everything, right? It's 5.6 billion searches per day." Now think about it. Those are point interactions. You kick off a transaction, you get one thing back, and we're talking billions here, right?
One single z15 frame, one box sitting on a 19-inch tile in a data center in and of itself can handle a trillion transactions a day. That's the differentiation. That's the power of the platform. That's the workload that it's driving out there. There's no other platform on this planet, don't take it from me, you can take it from IDC, that can handle that same number of transactions with the high availability, the security, and the stability that a mainframe provides. It's that unique architecture that helps it excel. I said it was different before. The type of thing that you have dedicated processors within the frame that handle I/O or certain workloads, and that you're not trying to swap things in and out as often, but you're segmenting out the workload that's there.
If you think about those things that need to handle high volume or scale up rapidly or handle high throughput, very chatty apps, things that would be going back and updating storage in the transaction or reading from it and going back and forth. The variable demand, the way it can scale up or scale down. I know cloud does that, but not without somebody checking on that hardware and rolling it in and out. This capacity is within the frame, and it grows up. The world's fastest commercial processor out there is in the mainframe. It is high horsepower. When you talk about downtime, when you talk about resiliency, I know for many years people talked about five nines of availability, and now the mainframe is up to seven nines.
You can get lost in all the nines and understanding what that means. That's less than 3.5 seconds a year of planned and unplanned downtime. That's staggering. It's unmatched in the marketplace. There's pervasive encryption across the workloads. You can tie that into other security capabilities and deal with data at motion as well as in rest. All of this now, you know, gone are the days where the mainframe had to sit in a dedicated space that needed certain power or cooling. This can all sit on a standard 19-inch tile in a data center and go anywhere you would slide in another rack. Now I started talking about examples of things you might do in your day so that you'd realize that mainframe is touching a lot of these workloads in the world.
I started with debit and credit cards because that's, you know, when people think of the large businesses in the world, when they think of the heavy workloads and that throughput, they think of financial. I wanted to pull you in on that. It's beyond banking. I mean, think about even just in banking, when the experiences of the last year and a half hit and there was economic uncertainty, a lot of us were probably checking stock portfolios and bank accounts a lot more frequently. That drives up transaction volume. Health insurance, this is how they guarantee the security and the privacy of information by using this platform. All the top ten insurers worldwide use the mainframe platform.
In government, when COVID hit, politicians, governments were making decisions to say, "Hey, we're gonna do stimulus checks, and we're gonna start distributing them in two weeks, and they'll all be in your hands within a couple of weeks." Well, nobody talked to IT ahead of time. Thank God they were using mainframes in the background to get those systems configured and tuned, to drive that throughput and have it processed successfully. Retail. The largest retailers in the world use mainframes, and I can tell you a really clear differentiator, and I think we've all had this experience. You go online, you're looking to buy a particular product. You see maybe five or six places that say they have it, and there's varying prices. So what do we do? We go to the cheapest one. We pick the cheapest one, we go to order it, and it says, "Sorry.
Out of stock." We go to the next cheapest one. Eventually, we get to one and they say, "Yeah, it's in stock." We order it. We're thrilled. We figure we're gonna get it in a couple of days. The next day we wake up, we look in our email, we look on our phone, and it says, "Oh, sorry. Yeah, that's on backorder. You're not gonna have that for another few weeks." I can tell you, those retailers are not using mainframe. 'Cause in a mainframe world, you would real time go back to that data center, go back to that inventory system, check whether or not you had the product, decrement your inventory, and then set it up to have the shipment made. Again, think about it. This is the difference between those Facebook, Google transactions.
Multiple things are kicked off when you do something in a mainframe. Now, I happen to know who a lot of those retailers are, so when I want something on time, I go to the ones that are running mainframes. Airlines and reservation systems. Hospitality. Think airlines, think hotels. This is one of the important reasons why they use mainframe, to have that real-time update and be able to manage the ticketing systems. Airlines also actually use it for the parts availability or plane placement. In automotive, five of the top five global automotive companies are using this. Think about why. It's the complexity of all the parts that they need to have in place to run that manufacturing line and have that supply chain working. This is really what the start of the design of the mainframe was all about.
The mainframe started with tracking systems for all the parts to the rockets to do the Apollo mission, and then it expanded beyond there into business. Next slide. Oh, thank you. We participate in about $7 billion of this market space. In every space that we compete, whether you look at IDC or Gartner, we're number one or number two. Gartner has us overall number one. There are really three big players: us, IBM, and BMC/Compuware. The whole rest of the share is by niche players with point products. It's a very stable market. There's not huge market shifts. I mean, think about this, 1% in market share shift would be $70 million. Very stable players that have been investing for a long time and are focused on the capabilities that these customers need.
The business value of transformation on the mainframe. I already talked in an earlier slide about the benefits of modernizing on the platform versus just migrating off or modernization to put a capability down a new technology. When you do that in place on the mainframe for workloads that excel there, you'll get a 6 times return on that investment. Every $1 you spend, you get $6 back in savings. You get about a 20% average reduction in the cost of operations when you transform that platform. Almost two-thirds more frequent updates to your applications as you bring modern DevOps practices to the platform. Now with Broadcom, across our portfolio, we've driven really fabulous results with our customers. 40% reduction in manual effort through automation on the platform.
We recognize that almost 90% of developers out there are using open source tools. They're not using them exclusively, but it absolutely informs our strategy and direction. The idea that you have these developers going out there using things that maybe they learned in college, maybe things that they can use to interact with other platforms, and that really drives into the discussion and the focus we have on opening up the platform. 60% of our customers have engaged with us in what we talk about as beyond code initiatives. Now I'll get back to that in a minute, but I'll stick with the portfolio first. The mainframe is really a system that you have to do everything soup to nuts on. You need to be able to have DevOps tools to develop applications and deploy capabilities.
You need core infrastructure, whether that be databases, transaction management systems, the tools that help you manage and monitor those things, automation, to be able to tie the things you're doing across workload, whether that be in your development processes or your production workloads, and understand with intelligence what's going on across that environment, and you need to keep that platform secure. We actually have two of the enterprise security managers for the platform in our portfolio, as well as capabilities that interact with those delivered from other providers. RACF is from IBM. At the core of this strategy, though, is our open first philosophy.
I want to differentiate to our customers based on the capability we have in our products and the capabilities that we have in our team, not by brittle connections that say, "You have to use this tool with that tool or only with particular vendors." Across our entire portfolio, we're opening up APIs and publishing those. Not just so that you can interact and connect our portfolio, but so that you can connect it to other vendors in the space or bring open source technologies to bear. Now, it's important that we do that in an open fashion because we don't know what the next technology is. We don't know what the next open technology is that people are going to want to use on other platforms and then be able to apply to the mainframe environment.
We also want to make sure that we can attract new developers, new operators to the platform so that they can leverage the capabilities and the core strengths of the platform without having to learn all the specifics or the unique interaction characteristics from years ago. That's why we participate heavily in the open source space for mainframe. That doesn't mean that we're taking our technology and donating it to open source. It means opening up the APIs. It means working on those open technologies and contributing to them so that they have the connections into mainframe, and then ensuring that we're doing things in an open fashion so that customers can interact in a multi-vendor strategy. The reality of businesses today, due to acquisitions and divestitures, is they have a broad portfolio of capabilities that have come from a variety of vendors.
Fundamentally, bottom line, we want our customers to be successful, and we can do that by providing flexibility across our portfolio. We also participate out there where our customers are to learn about the platform and the capabilities, whether that be through users groups or open education initiatives or conferences, and even things that we hold individually, making them available to our customers. Look, anybody can be a vendor. Anybody can develop software and deliver it for the platform. I showed you our portfolio, and I talked about the pillars, and I talked about being open.
The reality is, on this platform, you could just get the hardware and the OS from IBM, and you could do everything else you would need to do on the mainframe platform purely with software from Broadcom. It's a heterogeneous environment out there, not only for the things that are delivered for the platform, but as we open up and the things that can interact with the platform. The real challenge, the biggest challenge that has been out there is how do we help onboard skills to the platform? How do we make this platform more accessible to IT shops to interact and connect those with other workloads? We're a partner. We're not a vendor.
That means we don't just deliver software, drop it off at the door, they install it, we go away, call us if you have a support issue. We partner to help our customers be successful, and that's recognizing that they need help building skills. This is part of the beauty of coming to Broadcom. Broadcom brought a lot of investment to us, and this is how we drove additional value beyond products into our business. Our Vitality Residency Program is unique in the industry. We recognize that customers were having challenges hiring individuals to work on the mainframe. Fewer universities teach the technology, and they were having trouble finding people. Even that wasn't the biggest issue. The biggest issue was how they run their businesses and their own budget.
The reality that they may have a 20-person shop, and when they ask their boss and say, "Hey, I got Joe that's gonna retire, or Mary that's gonna retire in another year. Can I hire somebody so that we can cross-train and skill them up? And when the other person retires, I'll go back down to 20." And they say, "No." They say, you know, "How many are you budgeted for? 20? You're approved for 20. Worry about it when the person retires." That doesn't work. We simply said, "What if? What if we hire the individual? What if we train them on the fundamentals of mainframe?
What if we work with our customers as a partner to determine where that need is, ACF2, Top Secret, Endevor, IDMS, Datacom, give them a deep dive immersion on that training for probably four to six months, and then we deploy them out to the customer for an additional six months to work side by side with their employees to invest in that customer in a partnership so they have that crossover, so they have that learning? We've done this. We launched this a little over two years ago. Our first cohort was about five people who were placed rapidly. We've now run more than 60 individuals around the world, and we're continuing to expand this. This is on our dime.
At the end of the experience, at the end of the year, if the customer wants to hire them, no finder's fee, no services in arrears. They hire them, and they go into their business. This is a win-win. We get somebody who understands our products and is gonna be more successful using them. They get somebody who's skilled up. Education. The next challenge. How do you learn? Where do you learn about this mainframe stuff? We had a lot of online education that was non-instructor-led, so honestly, a pretty fixed cost for us. We were looking at how many customer seats were taken each year. We said, "You know, if we open this up, if we said, 'Look, if you've got license to that software product, have at it.
You can take the class for free, take as many seats as you want, have your chief bottle washer do it, CEO, whatever." In the first year alone, the number of customer seats taken doubled. In the second year, we grew yet again another 50%, and we're gonna grow again this year. More than three times improvement in customers taking advantage of education and having those skilled workers on our products in their shops. It doesn't stop there. Think about those folks that are becoming executives today, that are becoming CIOs, that maybe don't know what a mainframe is, maybe never had that in their sphere of control or what they had to manage. We also launched a program for mainframe insights for senior managers and executives. We record those. They're available on demand.
They run at about 20, 30 minutes, and we dive into topics, whether it be strategic outsourcing, security, terms, capacity planning, so that they can get their feet wet and understand the teams that they're leading. We also launched custom MRI assessments. Now, we don't roll in a scanner and do magnetic resonance imaging on the mainframe, but we did pick that name on purpose. It's mainframe resource intelligence. We have tools and capabilities where we can go into a shop, analyze their environment, looking for security exposures, performance improvements, tuning settings, capacity efficiencies, and we do this not only for our own portfolio, but for competitors as well. Again, with the focus of helping our customers be more effective and efficient. Now, every computer environment has change. You upgrade hardware, you upgrade OS, you upgrade databases, but you might use somebody else's tools.
We have a wide breadth of experience on our team, having worked on a bunch of different products and understanding the platform for many, many years. Too often, customers will think just about the product or the piece part of component that they're upgrading. What we've stepped back and said to our customers is, "Come to us. Tell us when you're introducing any change into your mainframe environment, whether it be hardware, OS, somebody else's database, but you're using our tools, something that could possibly impact security, anything, and we'll do a review in partnership with you ahead of time." This has greatly reduced the incidences or issues that large companies have when they do migrations through their change management. In fact, since we launched this, we have yet to see any customer that's participated with us experience a SEV 1 during one of their migrations.
That's investment and partnership up front. Now, I talked about the environment and that it's heterogeneous. There's products from a bunch of different vendors out there. We have an opportunity to grow our business not only from usage, but when customers use our portfolio versus somebody else's. Now, swapping out to end up with something that might have about the same capability can just get into a conversation about the financials of the product, and anybody can negotiate and make that look appealing. But the rub, the problem here, is usually the time it takes to get to return on investment. Think about what I said about skills. Our customers don't always have the people they need to do that migration efficiently, effectively or as rapidly as they would like to.
We launched a thing called Win No Fee Services, where we said, "Look, if you're gonna migrate off of a competitive product and come to ours, we'll help you. So long as you meet the functional and the performance characteristics, we won't charge you either little or no cost on those services because we're gonna make our money in the ongoing relationship." That's what partnerships are about. What you get is something for nothing. Investing in the customer side by side to help them migrate so they get to a quicker ROI. We've also launched consumption licensing. You hear a lot of people talk about cloud-like financials and business dynamics. Just like a cell phone plan, you can set a baseline and say, "Hey, here's the capacity, the flow of work that I'm gonna do on the mainframe, and I want to set that as my baseline.
I wanna know that as long as I don't go over that's what I'm gonna pay each year." You also want to know if I make extra calls or I use more minutes, exactly what it's going to cost me per minute. That's what mainframe consumption licensing is. We set a baseline. We set a linear price per MSU, which is a measure of flow of MIPS. It's complicated. Don't worry about it. Think about it like cell phone minutes. If you underutilize in any year during the term, you can roll over those MSUs or those minutes to the next year. If you go over, you know exactly what you're gonna pay per capacity use. Finally, we don't know what we don't know, and we don't know what our customers don't know. We know they have challenges. We know they have problems.
We engage them in design thinking workshops in the areas of our portfolio, in the areas of education or skills challenges, or anything else that they would bring to the table to figure out a mutually beneficial solution. Sometimes that involves our product. Sometimes that means we're going to enhance our product or morph or expand the capability. Sometimes it just boils down to changing processes and practices and how they're running the environment, all with the idea of partnership. I talked a bit about open and why it's important to open up this platform, and there's a variety of ways that you can think about this. A lot of people use modern DevOps practices, and they use these tools siloed to different platforms. We don't want that happening anymore. We realize that these workloads are integrated.
We realize that, you know, when you make that bank transaction or that debit or credit on your cell phone, it's hitting an edge server, it's going through the cloud. It might be hitting a distributed server, going back to a mainframe and then coming back. When these businesses are working on their development, they need to be able to think full scope, mobile to mainframe across that entire environment. That's why we're opening up, and that's why we're ensuring our DevOps tools that are focused on the mainframe interact with distributed tools, or you can use the same tools distributed cloud to mainframe across the board. Now, there are also synergies across our portfolio. You're going to hear from Serge, and you're going to hear from Rob. Think about Zero Trust and reducing risk and security.
I said the mainframe is the most secure platform on the planet, and that's why certain businesses go to it. You can do multi-factor authentication prior to interacting with the mainframe. Through our open APIs, you can do that with other vendors, or you can use that with Symantec's capabilities. We also have customers with those workloads that run mobile to mainframe that want to do automation to orchestrate or to have intelligence across that workload, across the same, you know, across that whole portfolio. They want to be able to deal with one vendor or have those technologies tie in together. While we open up to vendors, we also have capabilities across Serge's division and ours that tie together that allow customers to automate workloads across those platforms, as well as have intelligence on how they're performing or if they start to trend with issues.
What I want you to get out of this is that this is a stable space with three big players. Broadcom making significant investment, us differentiating in the marketplace, not only with the full suite of products that we have, but how we're partnering with our customers and how we're working and interacting with capabilities across our divisions, all to drive more value to our customers. Thank you.
Okay. All right. Thanks, Greg. Appreciate you spending time on mainframe. Up next is Serge Lucio, who runs our enterprise software business, and
You know, I talked about with CA, you know, we had to get the mainframe thesis right. I think hopefully through Greg's presentation, you have a better understanding of where that business is and where it's going. It's interesting, on the distributed side, CA was a very fragmented set of products across a number of different domains. I think in essence, when we thought about the investment thesis with CA, it was really, you know, we're gonna get mainframe right, and then, you know, amongst this broad set of distributed products, there's going to be some pockets of real opportunity.
I think over the last three years, you know, the good news is we found a couple of really interesting growth vectors, businesses that have some scale, businesses that are very well suited for our target strategic customer base, and they can actually grow double digits on an annual basis going forward. We've put a lot of investment, a lot of leadership focus around a couple of those areas. We're not gonna talk about all of them today, but in particular around areas like value stream management, where we have a leadership position. I think just as interesting, maybe even more interesting, is around the network monitoring space.
I think what you'll see is there's some interesting synergies between what we're doing on the software side and also what you see through our semiconductor business and actually even some of our security portfolio. With that, let me turn it over to Serge, and he can walk you through those businesses.
I'll start with a brief introduction. Is that all right? Yeah. Okay. My name is Serge Lucio. I actually joined CA Technologies about five years ago. I started leading the strategy and product management for the mainframe division at the time, so I helped shape some of the strategy that Greg just took you through. But my background is really on the distributed side. In fact, I started my career as a software developer, working for a start-up back in France, specialized in software testing for network equipments. We had a successful exit in 2001, and Rational Software acquired the company, and I joined the company in the US, so I moved to the US at that stage.
For about 15 years, I was involved in a number of product leadership roles at IBM across a number of product segments that I manage today. Of course, the testing part of the portfolio of IBM Rational at the time. Subsequently, I got into the agile management project and portfolio management. Spent quite a bit of time leading M&A for the Rational division. I did a few transactions in the areas of application security. Finally, I ended up being kind of a leader for the DevOps portfolio as part of IBM Cloud. Through the 15 years I was at IBM, I ended up competing with a number of the solutions that I lead today. Really, I grew a lot of respect for those solutions.
In many ways, I thought that these solutions were better than the solutions I was leading at IBM. Really what makes me uniquely qualified for leading the enterprise software division is really kind of the breadth of experiences I've had across the many segments that we have in the portfolio. I'm sure most of you probably are wondering, what is enterprise software about anyway? Let me try to introduce you know at a very high level to what we're doing. At its core, we're really about helping organizations go from ideas and build software ultimately to drive outcomes. Fundamentally, you can think of kind of two phases of this life cycle.
On one end, there is a clear phase around customers trying to innovate, and then eventually they deliver software, and they get into this operation phase. Those are kind of the two key areas we're going to be talking about today. If you peel beyond and you get to the next level, you can start to look at the number of tools which are involved in that life cycle. From an idea, you're going to first define what is the investment that you're going to put in that specific idea, right? Once you have an investment that's secured, you're going to plan for the delivery of that. Developers are going to be building it. You're going to end up testing it. That's the innovate phase. Ultimately, you're going to deploy this solution in production, typically in a data center on some servers.
You get in this kind of operation phase. During that phase, you are monitoring the software, you're delivering fixes, you're delivering changes, right? You're upgrading the software. Now, across this life cycle, which really is iterative, right? We are presenting this as kind of this linear process, but in fact, most of our large enterprise customers iterate through this sequence of steps every three months, every month, and for the best, every two weeks. These customers use a number of tools. In fact, according to Gartner, on average, larger enterprise customers we serve use 28 different tool chains. Now, these tool chains comprise oftentimes dozens of products. In reality, what you have in terms of a landscape, we're one of the many, many providers that these enterprises are using today. Why is that? Of course, there's been technology transitions, right?
The tools that historically these customers used on the mainframe are not the same as the tools they are using, for instance, for the cloud. The second key issue that these large enterprises face is that they've gone through mergers and acquisitions. It's very common that they are going to acquire a company with a totally different tool set. Then there is leadership preference, right? New leaders come in, they have different opinions with respect to what tools they want to see their teams use, and invariably, we get into new tools being added to these tool chains. The reality is that the switching costs are extremely high. In fact, if you look at some of our customers, they have been migrating from some of our competitors, from homegrown tools to our own solutions. The time to do this migration is oftentimes measured in years.
Recently, one of our large customers, a large financial institution, tried to replace one of their enterprise monitoring tools, which was a homegrown tool, something that they had developed internally, and took them 18 months to migrate from that tool to our commercial solution. Really, we're kind of in this kind of status quo. It's not to say that customers are not moving across tools, but to a large extent, it is an impossible mission to standardize on a set of tools. Now, with that context, these customers still have a set of challenges, right? They are constantly trying to innovate specifically and trying to digitally transform. Now, when you think about digital transformation, we are kind of in phase two of a digital transformation.
Phase one was very much about automating existing processes, was very much about applying your robotic process automation, trying to have processes which were more efficient, which was about remote workers, right? That was kind of phase one. In phase two, many of our large enterprise customers are really looking at introducing new services to drive revenue growth. Now, through COVID and the pandemic, what we saw was a rapid acceleration of these initiatives. In many cases, it was a matter of survival. If I take one of our large enterprise customers, Chipotle, you know, they were able to quickly pivot their investments right at the onset of a pandemic, and through the pandemic, we're able to continue to grow double digits their revenues.
For many of our large enterprise customers, they realized that they need a new operating model to actually be able to innovate. Innovate in a way that is much closer to what a startup or our startup or our born digital company would be innovating. Now, you know, a couple of weeks ago, I met with one of our customers, a large insurance company in the U.S., and I'm going to use this as kind of an example of really what's happening in concrete terms with many of our enterprise customers. This insurance company, if you think about their IT landscape today, they run two data centers, two mainframes. They are essentially a claim processing company. Right? You can imagine that the value from an IT point of view is processing claims. They're trying to reinvent themselves as an integrated healthcare service provider.
Tomorrow, they are looking at serving millions of constituents completely directly. They are looking at a network of doctors, hospitals, facilities. They are looking at a totally different landscape in terms of their IT. That is really what we're talking about when we're talking about changing the operating model. Now, it really affects kind of the two areas I talked about in very different ways. Let me start with how it affects how people innovate. If you look at this large insurance company today, they're organized very much like most of our large enterprise customers, meaning that they have functional silos. On one end, they have teams focused on software development. They have line of business teams who are basically directing the investments for the software development teams.
IT operations teams, which are managing the operations of the systems, whether we're talking about mainframes, data centers, or the cloud. Security teams. Within each of these teams, you can actually see that there is further decomposition. Back to what Tom was describing earlier, at their core, these organizations run based on their existing processes, right? Their processes is really what defines the collaboration across all these different teams. Now, when you look at born digital companies, which are represented on the right-hand side of the chart, what you see is a very different way to organize teams and organize innovation. Typically, what people are using is what we call value streams. What are value streams? You can think of value streams as collection of teams essentially serving a specific stakeholder.
In my example, it could be a set of teams dedicated to serve the needs of the doctors. It might be, you know, they might think about introducing a new service around telemedicine. Right? If you think about telemedicine, they are going to have video conferencing, they need to develop capabilities to help book those doctors. In this specific example, this customer is actually building some advanced technology using machine learning to be able to summarize a patient's chart for the doctor so that the doctor can be ready within a couple of minutes before meeting with the patient. Each of these value streams essentially is a collection of teams which is dedicated to serving the needs of a specific buyer or stakeholder. They're measuring their performance based on some key results.
In this case, it might be what is the utilization of those doctors. Their orientation is very much about driving business value. On the left-hand side, if you look at these software development teams, they don't really care about the business outcome that they are delivering. They're really focused on delivering on time and on budget. They don't really care if ultimately what is being delivered is meaningfully impacting the business. In many ways, you can see on the right-hand side, these teams operates almost like mini startups within these organizations, right? That starts to create a new set of challenges around how initiatives get funded. The funding model on the left-hand side is very much about funding a specific silo. On the right-hand side, you have a funding model, which is very different.
Very much like a portfolio management, where you're funding these value streams over time. Similarly, because you have way more autonomy at the value stream level, it becomes very challenging to actually have visibility as to what is happening across the organization. For most of our large enterprise customers, they are on this journey to go from the left to the right, and they've done that probably in some pockets. Doing that at scale is really the challenge that most of our large enterprise customers are faced with. That's around the innovation part of the operating model. The second key dramatic challenge is around how they operate those systems. If you go back to my example of this large insurance company, they are going from about two data centers to six data centers tomorrow.
Obviously, they are going to be leveraging the cloud to perform some workloads like analytics. They are going to have mobile applications deployed, you know, across millions of stakeholders. You can immediately see that kind of the complexity of the environment is increasing dramatically. It's not just that you have this more open environment and the scale of that environment, it's also the architecture of the software we're building these days. If you go back just 10 years ago, we were building applications which are essentially three-tier applications. An application would essentially run on three servers. Today, we're increasingly building applications which are based on containers, on microservices. What it means in practice is that in the past, you were managing three systems.
Today, to manage that same workload or that same application, you need to manage oftentimes thousands of containers. We're really talking about one to two orders of magnitude more elements in the infrastructure which have to be monitored. Now, it's not just the scale in terms of the number of elements that have to be monitored, it's also that it's a very volatile environment. In the past, you know, in order to deploy an application, an engineer in IT operations would be configuring a server, installing the software, and that would be it. Today, increasingly with container-based technologies, what's happening in practice is that you have an orchestrator, whether it's OpenShift, Kubernetes, which is dynamically starting these containers every single second. You no longer have this very static environment that you can really understand well. You have this very dynamic and large-scale environment.
The net of it is that there is just no hope for humans to be able to comprehend what's happening. The reality is that these organizations need to go from essentially a human-based process to a process that is fundamentally based on technology, where you are going to be applying analytics, machine learning and automation to get to a self-healing system. Those are really the two big challenges that we see with our large enterprise customers that we're trying to address and where we see the opportunity to drive double-digit growth. Now, the way we've chosen to approach this is recognizing that, again, most of these large enterprises have decades of investment in these tools at the bottom, and there's just no way these customers are going to be standardized on any of our tools.
We needed to find a way to help these organizations transform today and now tomorrow. In many ways, what we've built with these two solutions that you see on the top is really about augmenting what they are doing today at the bottom. We are not going to force these customers to replace the tools that they are using every day. We're not going to force their teams to be retrained, but instead we're going to augment that. We're doing that with, on one end, with the product owners, with our solution, that comprises two of our products, Clarity and Rally. On the other end, for the operations manager, we're increasingly focused on DX NetOps as our solution to help these organizations really transform.
With that context, what I'd like now to do is to actually give you more details on each of these solutions. First, on kind of what are the unique problems we're solving for our customers, how we're solving these problems, and how we fare relative to the competition. Let me start with the innovate phase. Back to my insurance company. 1,000 developers today, they are scaling to about 10,000 developers tomorrow. On average, when you look at our large enterprise customers, the 600 strategic accounts that Tom talked about, these organizations typically have anywhere between 10,000 developers to 30,000 developers. That's about the range. Of course, when you manage this at that level of scale, what happens is that you end up having to manage hundreds, thousands of projects.
The first thing that happens typically to manage that kind of scale is you invest in people, you invest in project managers, and you spend a lot of money in doing that. In fact, one of our large enterprise customers reported to us before adopting our solution that they had about 1,000 project managers producing 100,000 PowerPoint pages on a weekly basis just to provide status reporting. Now, other customers try to solve that through technology, right? They may start to build a homegrown solution which is going to aggregate the status across all these project teams and ultimately roll it up. What oftentimes starts as being a five team member project to 10 team member project, in some cases ends up being 100 people, 100 engineers, fully dedicated to actually just provide a system to report on really what's happening across these thousands of teams.
That's the first source of waste, which is really about. You can think of it, of just trying to understand what is going on, what is the status of the investment? Are we delivering on time, on budget? The second one is that invariably when we're talking about hundreds of projects, you have dependencies. You know, when we talk about these value streams, you can think of, you know, the value stream team that is focused on the doctors as dependencies on, the teams which are serving the patients. Oftentimes the challenge is managing those dependencies at this kind of scale. The second source of waste is really driven by the fact that these organizations end up, invariably with idle time, right? Teams are waiting for other teams to complete some work before they can continue and finish their own work.
Those are two first kinds of sources of waste. The bigger one, and the one that you see on the bottom right of this chart, $1 million being wasted every 20 seconds, is really driven by what we call overproduction. Toyota actually introduced first this notion of overproduction. What is overproduction? There's one source of overproduction, which is just people working on stuff that nobody knows about. Your leadership team doesn't know about, and believe me, I've been there in my own organization. Typically, you have about 10% of your organization where you don't really understand or have clear visibility as to what is happening. That's one source. The bigger one is really the time it takes to actually get feedback.
Keep in mind that when we develop software, we have an idea, right? We have a hypothesis as to what that software is going to deliver in terms of value, but we don't have certainty until we actually deliver this software, and we actually are able to measure or quantify what is the value that is being experienced. The reality is that oftentimes we build things which do not materially impact the business, do not drive value. Right? Back to my analogy as a startup, you build stuff and ultimately maybe you hit the mark, maybe you don't. The big challenge is how quickly can you provide feedback to these development teams so that they can pivot their investments. To just think in your own experience and take a product that probably most of you have used, Microsoft Word.
How many features of Microsoft Word do you use today? I'll bet that you probably are using less than 30% of those features. Yet, Microsoft continues to maintain release after release 70% of these features that none of us is using. That's the big problem. How do we solve that? It really starts at the bottom with Rally. Again, Rally is one of our products, and it's all about providing visibility to the leadership team as to what is happening across, you know, tens of thousands of developers. Now, Rally integrates with whichever solutions our customers are using today. Commonly, we will see Atlassian Jira. Azure DevOps is more and more common these days. GitLab, which IPOed recently, is also an emerging player in that space that provides team collaboration.
Rally is able to ingest the data from all these different tools to first provide a real-time view as to what is happening. Are we delivering on time and on budget? The second is that once we have this data, we're able to start to actually perform some very interesting analytics, specifically to identify any kind of dependencies that may exist across these different projects. Oftentimes, we're able to detect way early that there are dependencies which are not explicit in the plans. Rally is fundamentally going after kind of the first two sources of waste that I talked about. Now, we all have to recognize that there is another kind of value stream. We call that this bottom layer, kind of this delivery value stream, but there is this operational value stream.
Again, if you think about this operating model, we're funding in a different way. We're funding these different value streams. Over time, depending on the performance of these value streams, executives will want to pivot or shift investment across these value streams. All that work typically is happening at this operational layer, where you are deciding on the funding, you're defining a roadmap and expectations with respect to when you expect specific solutions to be released. Ultimately, Clarity, which is our solution at this level, is integrated with financial systems. We can integrate with financial systems, SAP, Oracle. We can integrate with user experience management products to actually get feedback as to what is the value that is effectively delivered.
The close integration with Clarity and Rally is really what's enabling us to constantly realign what the IT teams are doing as it relates to driving or maximizing business outcomes. Let me use an example of one of our customers, Ford Motors. A few years ago, really at the onset of their Mustang Mach-E program, Ford really wanted to change the way they would drive this program. They wanted to go from a classic model, which is funding projects, to a model that would be really embracing value stream management and funding the product. Across their 700 or so teams, really what they wanted to do is to measure these teams on three dimensions. The first one was predictability.
If you think about predictability and what I just described, one of the big challenges around predictability tends to be around two things, dependencies across these different projects and the ability to effectively forecast or estimate what is going to be the cost or the time to actually build something. The second one was responsiveness. Through this kind of program, invariably, you need to pivot. You need to add more investment, shift investment across different parts of a project or program. That was the second one. The third one was productivity in the sense of, not just measuring the lines of code that these teams produce, but measuring what is the business value that ultimately this program delivered. Now, we have this unique position that, for the most part, these two solutions, Clarity and Rally, are SaaS solutions.
Clarity can be used on-premise, but the majority of our customers use these two solutions as a SaaS offering from us. We have these unique vantage points where we have visibility across hundreds of thousands of teams. One of the values we're also providing to Ford is actually to help them understand how their teams, their 722 teams, really compare to really our teams across many industries. Over the last five years, we've been able to get these 722 teams to perform in the top 25%, again, across the entire population of teams that we're managing today. Just to try to summarize our key differentiators. The first one is really around this notion of enterprise-wide insights.
The ability to aggregate data irrespective of which team collaboration tool you are using, whether it's one of our competitors or our own team collaboration tools. The second one is to tie that to the value that is being delivered by these teams. When you think about Clarity, we integrate with these financial systems. We integrate with systems which are going to be looking at customer churn, which are going to look at Docker utilization. We can provide that feedback almost real-time to these development teams so that they can pivot, so that they can focus their investment in the right direction. That's the third value. It's our ability to realign, constantly realigning those teams to maximize the business output. Now, many of our customers are turning to our solution because we're really unique in the industry.
For the last 5 years, Gartner has been recognizing our solution as a leader in their Magic Quadrant for enterprise agile planning. I think we could go as far as saying that we've invented this space. If you go back 15 years ago, Rally was the first solution that really provided this kind of end-to-end enterprise agile management solution for large organizations. In many ways, Rally invented the de facto standard that's used today across the industry, SAFe, which is the Scaled Agile Framework. Clarity has been a leader for over 20 years in strategic portfolio investment and management. As the two are converging, what we clearly see is the renewed integration that we have across Clarity and Rally is really positioning us to continue to strengthen our position in this Magic Quadrant.
Now let me turn to talk to you a little bit about our DX NetOps solution and how we help our customers operate. I'll start really with the basics, which is that when you interact with a piece of software in the cloud, let's say Office 365, this software is deployed in typically multiple data centers, right? It's running on a server that's connected to a network. Immediately, I want you to understand that it is a complex problem, that when we talk about monitoring the systems, we talk first about monitoring the systems at different layers. You need to monitor the network. You need to monitor the servers where the software is running. Of course, you monitor the application and the health of the application. Increasingly, we're monitoring also the users.
We're trying to understand how these users are interacting with the systems. Are they actually getting the value that we anticipate them to deliver? Are they using specific features? Each of these monitoring technologies is very different. Naturally, you have classes of tools which are used for each of these layers. Now, at each of these layers, there's further fragmentation. In fact, we recently conducted a survey with our large enterprise customers, and more than 50% of our large enterprise customers use more than six network monitoring tools. What's interesting is that as you go from network to infrastructure to application to end user, there's actually further fragmentation. In fact, most of our large enterprise customers use dozens of tools when it comes to monitoring end users.
Very complex kind of landscape in terms of the number of monitoring tools which are in effect deployed at our large enterprise customers. Now, the other thing that I think is important to understand is that when there's an issue, it's rarely an issue that's going to be isolated. Let me use a very trivial example. Let's say that you have a fiber optic link that breaks in a data center for whatever reason. As you can imagine, this fiber optic is connected to a router. This router is likely going to have some alarms identifying that there is something wrong here. It's very likely that the servers which are leveraging this router are going to experience maybe not a loss of connectivity, but more likely they are going to see that maybe bandwidth is degraded.
Ultimately, if we're talking about video streaming, you may end up in a situation where it starts to affect end users. You know, people start streaming video and then stop because the quality is poor, right? You can hopefully see that there's this domino effect that in fact a single issue oftentimes results in dozens and sometimes hundreds of alarms across the many, many tools I was talking about. You live in this very kind of noisy environment. These IT operations teams really are dealing with tens of thousands, in some cases, one of our large customers, one of the largest financial institutions in the U.S., is dealing with over 100,000 alarms every single day.
Now when you look at these alarms, what's really challenging and shocking is that oftentimes if you're an operator looking at a specific issue, it's very difficult to understand what is the impact from an end user point of view. There is a very different, very significant difference between not being able to post on Facebook for a couple of people and Facebook being down for everybody, right? The big challenge is oftentimes that you don't see the signal, you are not able to parse noise versus signal, and everything is very noisy, and you're not able to prioritize the work to minimize the impact on the customer experience. Now the other reality is that today, the mean time to repair, so the time it takes from having an alarm to that alarm being remediated remains around eight hours.
Finally, if you go back to Facebook a few weeks ago, it took them six hours to remediate to their big issue. If you look at Roblox, for those of you who have some kids, it took them like two days to address their issue. It's not uncommon to have these very long mean time to repair. During this time, of course, you're losing share. How do we uniquely address the problem? At its core, it's actually very simple and very complex. The biggest challenge is to be able to correlate this information. Again, you have information that's coming. Most of our large enterprise customers have tons of data. They have tons of monitoring data. They have all these tools monitoring everything. But the challenge is that all this data is completely siloed.
What's happening at the network level, the infrastructure guys don't know about, and you can imagine that you have basically everybody's looking at the forest, nobody's looking at the tree. The biggest challenge is really about being able to correlate this information from the network to the infrastructure, to the application, to the end users. Now, once you are able to do that, and I'll go back to how we do it, you can start to do very, very interesting things, right? Because you can start to look at all these alarms which are coming essentially from the same cluster, and say, "Well, that's these 200 alarms are actually related to the same situation, to the same incident." What we've experienced is that we can reduce the number of incidents that our customers are looking at by over 90%.
Significant reduction in terms of labor costs for them. When you think about one of these alarms oftentimes being anywhere between $10 to $100 to actually be able to analyze and remediate. Once you have that, because we're able to correlate this information all the way to the end user, you can start to quantify what is the business impact, what is the user impact of these issues. What that means is that you can start to prioritize all these situations. What are the ones which have the most meaningful impact, and the ones that you should be looking at first. Now, because we've correlated all this information for each of these situations, we can go all the way back to what was the first alarm, which is likely the root cause.
Based on this alarm, we can start to categorize those alarms and use machine learning to identify what kind of automation script or service process can be applied to remediate to this issue. All of this basically stems from just our ability to correlate this information. Most of our competitors will tell you it's easy. Just ingest all data into, you know, Splunk or Elastic, choose your favorite solution. Believe me, we've tried to do that ourselves. We initially used Elastic to do that, and it just doesn't scale. It does not scale for the complexity and variety of data that we have to ingest. We had to find a new way.
In the last three years, we've spent a lot of time actually building our correlation engine to what it is today, and to a point where recently one of our large enterprise customers, which spent over a year implementing one of our competitors, implementing over 1,000 rules in that tool, was able to just turn on our system, and our system just worked out of the box. That's our secret sauce. It's all about being able to correlate that, and more importantly, to correlate that all the way to what is the end user impact. Let me use a concrete example, O2. O2 is one of the largest cell phone providers in the U.K., 30 million subscribers. Just like many of our customers, they're trying to introduce new services, new source of revenues. In their case, they wanted to introduce an SD-WAN service.
Think of it as 15,000 Wi-Fi antennas, public access points all over the U.K. that will enable small businesses to access the internet. Now, you can immediately see we're talking about millions of devices connected to 15,000 access points, and you need to be able to not just monitor the access network and those devices, but also what's happening in the back end. You know, the ability of these people to eventually get to the internet. Very complex environment, a very diverse set of equipment that they had deployed across this environment. But the biggest issue that O2 was faced with is when something goes wrong, is it impacting a low-value customer, a high-value customer? Are we breaching on a quality of service or an SLA that we've extended to a customer?
Our ability to be able to correlate, again, all the way from the network to what is the end user impact, was absolutely critical in their case. It was not just about being able to provide them with a single pane of glass, end-to-end visibility from the access network of devices all the way back to the backhaul, but also our ability to quantify the business impact that really made a difference. In their case, we've been able to demonstrate our ability to reduce by more than 75% their mean time to repair. Just to summarize, for us, it's first and foremost about unifying the relevant data. The second one, and hopefully you understand that now, it's about quantifying what is the impact. These alarms are not all created equal. Some are really critical, and you need to react very quickly.
Some really do not matter because maybe they are taking care of the application layer. Maybe the application is actually designed in such a way that it's resilient to issues that may be happening at the network or infrastructure layer. It's about pairing that with automation. Because ultimately what you want to get to is a system which is almost self-healing, a system which has actually learned how to fix those own issues. Ultimately, it's all about driving the customer experience. Now let me talk about these customers and I'd like to talk to you about the O2 customers by taking kind of the other side. If you are one of these O2 customers, let's say a small bank in the U.K. today, and for a reason or another, you are trying to access, let's say, Office 365.
Let's just imagine that you're not able to use Office 365. Your first instinct is most likely that you are going to try to Google or go to another website. Let's imagine just for a second that you are successful at going to Google. Your second instinct is going to call Microsoft Support and ask them what's going on. Microsoft tells you, "Well, everything is green on our side." Where is the issue? That's actually a scenario which is more and more common, and one of the scenarios we are really excited about. Let me talk about this subway map. This subway map in the center is the Internet. We all take the Internet for granted, but it's a very, very complex network.
In fact, when you are accessing the network, when you are using Office 365, you're likely traversing about five to six internet service providers. Of course, you only care about AT&T or Verizon that provides you with service, but the reality is that you have this chain of connectivity that gets you all the way back to the data center where Office 365 is running. The reality is that you don't even decide where your traffic is going through the internet, right? As you can imagine, just like a subway, you have links here where you have a lot of bandwidth, that you can carry a lot of data. You have links which are really good at latency, so the traffic goes extremely fast. Depending on the application you're running, well, your application may be actually very dependent on the characteristics of the network.
The reality is that we increasingly depend on the Internet, and this is due to many drivers. The first one is obviously the SaaS adoption. More and more our large enterprise customers are relying on Workday, Office 365, right? A number of the solutions that we use to consume that were in the data center are now running in the cloud. The second is that many of our large enterprise customers are moving workloads to the cloud, but as Greg described this to you, many of these workloads are actually tethered to the mainframe. There's a lot of connectivity between the cloud and what's happening in those data centers, whether it's a mainframe or distributed systems. Connectivity between the cloud and the data centers is actually critical for these customers. The work from anywhere is another key driver.
Increasingly, what we see is our large enterprise customers have more than 50% of their employees working remotely from home, where they don't have any control or visibility as to what's happening in the home. Even in parts of a network that historically we used to control, say the connectivity between a branch office and a data center, we increasingly rely on SD-WAN technologies. Why? Because it's cheaper, right? Historically, you would be renting essentially a dedicated line just for you. Going to the broad Internet and using SD-WAN is cheaper, faster, and more flexible in many ways. That's kind of a reality that in many ways it's very difficult to troubleshoot what's happening.
Just to help you understand, let's say you are working from home, you are trying to access Office 365, and you go through the path. The issue might be related to maybe something that's running on your computer and slowing you down. It might be related to maybe bad Wi-Fi reception. Maybe you have issue with your provider, whether it's, you know, Verizon Fios or Comcast, right? You might have an issue which is really on the access side, and maybe it's just you because a branch fell on your fiber-optic line, or maybe the entire neighborhood is affected by that, right? At each of these stages, basically today, for most of our larger enterprise customers, it's essentially a black hole. They do not know what is happening.
Now, what's interesting is that Broadcom is actually everywhere in that network. Our chips power all the routers, all these links across the internet. Our chips are there. Similarly, when you look at the edge, we power a lot of the Wi-Fi devices out there. We power a set of boxes which are in your house. Increasingly, and you will hear that from Rob, this network is being secured using technologies such as proxies, cloud access security brokers, which are powered by Symantec. We have a lot of vantage point around this network. We believe that we have an incredible opportunity to actually spotlight across this entire network to help our customers really troubleshoot these kind of very complex problems.
EMA, one of the leading analysts in the network monitoring space, recently ranked us as the highest with respect to our technological solution. They cited really three things. First two, hopefully right now you heard me, scalability. We're able to scale to millions of devices. Two, our ability to have advanced analytics to actually parse signal versus noise. Three, our unique ability to integrate directly into the telemetry that our chips are generating so that we can understand what are the network conditions and how the traffic, the congestion, the latency, the packet losses, all these different things which are happening on a real-time basis are affecting an application. We believe that this is kind of a next area of growth for us around the network.
Now more and more of our large enterprise customers are turning to Broadcom to solve these two challenges, whether it's around are they transforming their operating model to innovate or to operate. Fundamentally, I believe that it really starts with a strategic relationship that we have with many of these large enterprise customers. Back to what Greg was describing, we're serving 600 accounts, so we can form a level of relationship with these accounts which is unmatched candidly in the industry. Second is that we've been dealing with these problems for decades. It's not just me, it's many people in my organizations have been dealing with these challenges for decades. The silos I was talking about, believe me, 20 years ago, we were talking about those things.
What's different today is that it's a bit of a live or die for many of our large enterprise customers. They recognize that they have to transform, and the pandemic has just accelerated that motion. At the end of the day, there are really three things that make us absolutely unique and uniquely positioned to help our large enterprise customers. First, we embrace the ecosystem of tools that they are using today. We're not forcing them to move off of their existing tools. We recognize that they have decades of investment in those tools, that these tools are tightly integrated with one another. Two, we really are focused on scale, right? In fact, many of the challenges I've been talking about during this presentation do not exist if you don't have five, 10,000 developers, right? They only exist at this kind of scale.
For us, scale is absolutely key. But with scale comes the need or the imperative to actually perform analytics. You know, you cannot deal with that kind of scale by just providing dashboards or reports. We need to use advanced analytics and machine learning to be able to actually provide the right insights. On one end, when we look at Clarity and Rally, it's all about improving the resource utilization. It's all about removing the waste that I've been talking about. When we talk about DX NetOps, it's all about reducing the amount of labor, first by being able to collapse all these alarms into these like situations or incidents, and we can. We have proven repeatedly that we can reduce that by 70% to 80%.
with that, being able to provide a probable root cause for our customers so that they can accelerate their ability to repair and return to normal. Look, we're very excited about these two opportunities. For most of our large enterprise customers, again, they are just at the beginning of this journey, and we believe that we are absolutely uniquely positioned to help them. Thank you.
Thank you, Tom, Greg, and Serge for your presentations. We are running a little bit ahead of schedule, so we will take approximately 20 to 25-minute break for lunch. Box lunches will be available outside of this room, so please help yourselves. Thank you.
Your IT environment is incredibly complex, and this complexity can quickly create chaos. That's why at Broadcom Software, we dedicate all of our resources and expertise to help you navigate the complex with business-critical software solutions that modernize, optimize, and protect the most complex hybrid IT environments. Because when you tackle complex problems at scale with a partner committed to your success, you see things from a new perspective and move your business forward. Navigate the complex with Broadcom Software.
Good afternoon. Our event is about to begin. Please take your seat.
Okay, let's get started again. Hope everybody enjoyed lunch. Up next is Rob Greer, runs our Symantec business. I think what we found really interesting about Symantec when we made the purchase and did the diligence was when we looked at the CA customer base, and we started formulating the strategic focus on these 600 accounts, and then we compared that to actually the growth rate and all the exciting things happening in security, we saw an opportunity to marry those two things. It's been an interesting journey actually over the last couple of years with Symantec. We've gone through a major restructuring, integration, rationalization. There was a long tail of customers outside of our strategic focus.
What we found compelling and what we're really excited about going forward and what Rob's gonna cover is that when we look at customers that have this significant on-prem footprint and are investing in their own data centers, and there's security needs that are unique to that, and you marry it with what's happening in the cloud. What Symantec did over the course of a number of years through acquisition is they really set up a portfolio that was extremely well suited for that account base. I'm excited to introduce Rob. He'll walk you through all those opportunities ahead of us. Thanks.
Thank you, Tom. Nothing like presenting right after lunch, get everybody's attention. Awesome. Let's see if this clicker is gonna work for me this time. Okay, gotta press it really. Okay, here we go. My name is Rob Greer. I am the General Manager of the Symantec Enterprise Division, and I'll spend a lot of time today telling you what that's about and why that matters. My background has been one of a boomerang, twice, actually. I have been with Symantec three times and have returned twice, all voluntarily over the last 16 years. I have to say, I'm most happy about this version of Symantec, and hopefully, you will see the same way at the end of this discussion.
Some of you might have interacted with me when I was the Chief Product Officer at Forescout. I helped take the company public in 2017. It was pretty awesome being back here four years later. The team with me these days at Symantec, many of them have been with Symantec for multiple decades, and these are some of the most uniquely qualified individuals that help protect the world, frankly, against many of the threats that we see. I'm really excited for the opportunity that's been given to me, and look forward to engaging with you guys a lot in the future.
Let's give you a little bit of context around what we were like as Symantec Corporation, basically everything in the market, from a small consumer to a small business, a medium-sized enterprise, to the largest enterprises. Back then, when we looked at our investments, it was like peanut butter, if you wanna use that term. It was really hard to make big bets in one of these areas without impacting one of the other segments. You know, as Tom mentioned, you know, we have had the opportunity as being part of Broadcom, and now Broadcom Software, to really laser, put laser focus and investment around the key use cases. Our largest organizations, enterprises, both public and private, need us to focus on.
They have very challenging environments to make secure and compliant. As you think about it, many of you work for large institutions, and there's a legacy that institution must keep. You can't just, you know, leave the old house behind or the old data center behind and go into a new environment because you have all these applications, you have all this data. It's a very complicated environment. Fortunately for Broadcom and for Symantec Enterprise, we have capabilities that can protect what they've relied on us for several decades, a couple decades, in some cases, as well as where they're headed. I'll spend some time on those key use cases, so you really have an appreciation for where our relevance is.
During Q&A, I'm sure you'll have a lot of questions. This strategy that has been revealed to you today by my colleagues has really worked well in less than a two-year period for Symantec. When you think about our coverage in the Fortune 500, a little less than half our coverage, we have about 773 customers that are residing in Forbes Global 2000 organizations, all share those characteristics that I showed you on the last slide. The part that really excites me the most, being the general manager, is 68% of those current, you know, Forbes Global 2000 customers that we have today, they have two or more of our franchises.
Our franchises are endpoint security, which Symantec was known for, has been known for since its inception, our information protection, network security, which we recently got into about five years ago, and of course, email security. When you have customers that have already invested there, if you're having the right dialogue with them and they're adopting your technology, what an opportunity to expand that. Because in today's world, large organizations don't want hundreds of tools. In security, the more tools you have, the more options there are for bad actors to take advantage of the lack of coordination that happens. You just have to read the news today and last week, every week, there's something out there. If you add too much complexity with too many tools in that environment, you're just making it easier for the bad guys.
The part that I also wanna kind of underscore is we have a huge footprint. We have 150 million enterprise users deployed today. Okay? We're getting lots of telemetry to help protect and inform our customers. Over the last year, since 2019, over the last two years, our cloud adoption use cases, specifically to Symantec Enterprise, has increased 40%. 20% of every dollar goes to R&D. This is an investment area for Broadcom. I mean, and you have to in cybersecurity. Cybersecurity isn't something that you can just set it once, and it's just gonna protect you because every second there's something out there that is mutated that you have to be thinking about.
We're not the only ones that, you know, have seen some improvements in the last few years and have recognized it. If you look at Forrester, they recognize us as, and have for many years, a leading unstructured data security platform. Many of you guys know of it as data loss prevention. We're also well known and recognized around cloud security gateways or cloud access security brokers and the combination of that. Many of you follow Gartner and others like IDC. They refer to this category as SASE, Secure Access Service Edge. You also get confused because we like words in our industry, Zero Trust or zero trust network access, which is basically a fancy word of saying secure remote access from anything, from anywhere, from any device. These are more.
This use case, frankly, is very relevant today, given what we've gone through the last two years. Symantec, as Tom stated, our cards, we have a good set of cards. We were dealt a good hand of cards. The people before us, I was actually there when Vontu was acquired and actually ran that business. Had a foresight around what is required as more and more companies look to the Internet and to certain services that were not within their perimeter and thought, "What is the most important thing to protect?" It comes down to data. Nowadays, you're also seeing operational technology and infrastructure as another element that is elevated thanks to the connected nature of the world.
When you look at the technologies that we've acquired, DLP, CASB, Secure Web Gateway, Mobile Threat Defense, Zero Trust Network Access, a lot of these technologies, they have actually now been integrated. In fact, all of these have been integrated in the last two years. I'm gonna talk to you about why that matters and why we have a competitive advantage with the customers who've already invested in our much of our on-prem footprint. The last two years, those innovations have come out rapidly, but with a lot of relevance because our 600 plus customers that we spend most of our time with, the strategic customers, they want a single endpoint security platform.
They don't wanna buy an EDR product from CrowdStrike and, you know, another antivirus vendor from Trend Micro or name your vendor and then try to bring it together. Many of our customers have been with us for many years, and they need evolution of the types of attacks that have come to fruition over the last several years. The way to deal with that is to simplify things. You know if you have cloud use cases, you need to address cloud use cases. If you want to manage things on-prem and in the cloud, which is hybrid, which is where we believe the end state is for our customers, you need to address both what they have today and where they're going. We have released an integrated endpoint security complete solution that is device. It's per device, per user. It is.
You buy it's subscription, you can run your management on-prem, you can run in the cloud, you can go at the speed that you want. We did the same thing with our web protection suite, which with the foundation from, you know, the industry leader in proxy when we bought Blue Coat. Lot of great intellectual property, a lot of very loyal customers. They feed many of their security policies through our appliances. Many of them on-prem, many of them virtual appliances, many of them in the cloud, most of them all three. They expect to be able to do what's required to protect their business and be compliant. We've innovated on something that's really newsworthy. It's ransomware, adapting to protect our customers against ransomware.
Our customers all have different usage patterns for their devices, for their laptops, whether it's Mac or PC, and we are able to know what normally happens and then adaptively create policies that are specific to that environment. It's very, very awesome innovations that have come out of that team. Of course, during COVID, all of us experienced what it's like to all of a sudden be thrown out of the office and remote. Well, majority of our large enterprise customers didn't already hand out laptops and phones and to everyone in the organization.
We were able to bring to market what we call Mirror Gateway, which is the ability for you as an end user who's never been given a corporate device, go home to your MacBook Pro, open up your Chrome or your Safari, and get access to applications that are either hosted inside your organization or as part of your SaaS service, maybe it's a Salesforce or Workday, and not put your company at risk around DLP, around compliance, and around ransomware. The way we were able to do that is integrate the Elastica CASB with the Fireglass web isolation, with the Web Security Service that we built native, you know, organically with our Blue Coat business. The area that I'm most proud of though, the last two years, is around our digital transformation.
We spend a lot of time getting in front of customers talking about digital transformation. Well, we went through it, and we went through it with a great partner. Google helped us transform. Many of it we had to rewrite native software code. Some of it we had to lift and shift because it was already there. We used to run our own data centers. A lot of our R&D time and money, we had to build out capacity before it was actually used. It wasn't elastic. The last two years we have partnered with Google. We've provided scalability. We are now more prevalent around the world than any other network in our space because we're leveraging the YouTube network.
We have virtual points of presence in areas where there aren't Google data centers, and we've innovated and able to in order to give customers a localized experience. You're in your local language for your environment, getting access in a secure way, thanks to the Google platform. Our team gets to spend time working on security, making the products more usable, integrate key use cases with my colleagues, you know, around how do we secure access to mainframe, or how do we share threat intelligence with our network, AIOps platform. Those are all the things that our customers ask us for.
When you bring it all together, we are able to do something for large enterprise customers that I know of no other vendor who can do, which is take care of their devices, take care of their data, take care of their applications, regardless of where those workloads are today, and help them future-proof that transition to the cloud at their own time. We've simplified the user licensing model. In the past, we'd make it more complicated, where customers would have to know how many cores they have or how much bandwidth.
Now you just say, "How many users do you have? Great. $X per month, and we're gonna solve those use cases." The foundation of our platform, best-in-class endpoint security, best-in-class data loss prevention, best-in-class proxy and network security, best-in-class email security, all leveraging the Symantec Enterprise Cloud, which you'll hear more of as we move forward, is what our customers are looking for. They're looking for an on-ramp, a bridge to cover their entire landscape. It's not just about data centers that are moving, it's the devices that they're using. You have to support Windows and Mac and Linux equally. You can't just be all in on Windows. You know, customers want currency on browser support, and Microsoft Edge is now the standard versus Safari versus Chrome. This is very complicated, and many customers built these processes, and they look to Symantec Enterprise, part of Broadcom Software, to be their partner.
We're partnering very closely, similar to what Greg's doing in his side of the house. What are the key budget areas that customers fund this? Now, I was talking to one of your analysts here in the room during the break. You know, the board is demanding coverage on a lot of this stuff nowadays, so that's helping. But the reality is, there's implications if you don't have the right cyber hygiene, and that is there's regulations around the world. Some of them are geographic, some of them are specific to industries like, you know, if you think about PCI compliance or HIPAA. Every country, every industry, potentially any states in the United States, there's different compliance issues.
This is the number one budget driver that you're gonna see for any cybersecurity company that's focusing on a large enterprise. The second most important challenge that drives budget is something we're all accustomed to now. You know, remote work. We're probably all a lot more efficient now than we were before, just to be honest, because we're always on. To be on and to get access to key applications and data and collaborate, you need to be able to do it in a safe way, a performant way, and you can't impact the user experience. This is difficult. It's very challenging 'cause you're dealing with applications that don't lend themselves to these new containerized serverless deployment. Not all of us get that opportunity. Most of us in the room have to deal with older apps and newer apps at the same time.
How do you deal with that in a Zero Trust way? Many of us in this room probably experience there aren't enough VPNs for all the customers or the employees. There are not enough devices. This drove a lot of budget that came out of nowhere. Then lastly, the one that we all hear about, the one we hear in the news is around protecting against being the next victim of ransomware, being the next victim of taking down a critical piece of infrastructure like a pipeline or hospitals that are being ransomed to get access to patient data. This is real-world. OT, IT convergence, it's here, and all of us should really think about how are we gonna stay ahead of the bad guys. That's where we want our resources. We're no longer building data centers.
I get to put our engineers focusing on that, on those issues, not how to rack and stack on a 19-inch rack. Let's go through some simple use cases. We'll add a more technical level, so you guys can kind of associate a little bit more about what we're doing, okay? Large enterprises, as I mentioned, deal with lots of different compliance challenges. People need to behave a certain way. You have to handle data. You can't get on conference calls and talk about personal data, personal information, right? You have to train your people. You have to.
Your devices need to have the right protections or patches, or you need to prove that you're encrypting, you know, the device if it gets lost, because if it gets lost, your company is liable for that for potentially data that they're required to protect. Then, of course, the applications which are sitting in your control and out of your control. Most of the time, it's a hybrid of both. The vulnerabilities that are coming out every day that many of us don't control, open source. You know, a lot of the supply chain attacks you've heard about, you know, you can probably go back in time and see some open source code that needed to be updated, and everyone has been impacted by it, right? What are companies looking to do?
Well, they need a consistent way to go and create policies such that they can get the right visibility as to are they compliant or not, okay? That's number one. That's just to keep the lights on. Most importantly, as data is moving in all these different directions, from device to device to databases you control to you know, box on the internet, all these different particular applications, how do you know where your data is? How do you know what data you care about? Which data must you care about improve? This is what our large enterprise customers are dealing with.
Symantec Enterprise Cloud, because we care, we understand the on-premise nature, we understand all the operating systems that matter, and we understand how to deal with cloud workloads and traditional on-prem workloads. No other company has this ability, and this is why many of our customers have been working with us to express the needs they have so that they can expand and invest more and more in our franchises. The second key pain point I talked about earlier is around remote access. Well, what kind of remote access, Rob? Well, first is the one that you guys are used to, and that is but you're accessing cloud, let's say CRM technology, like a Salesforce, for example.
Many times, traditionally, the customers can't deal with this use case at scale because they have to route all that traffic back to the corporation through a data center, okay? That's challenging. They're asking companies like Symantec, "Hey, we'd like to enable secure access to these workloads, but we have all these compliance and security regulations that, you know, we can't go through our data center. We just don't. The cost to users are not gonna they're gonna go on strike," you know? That's the number one. That's one of the number one use cases that we've all experienced. The second is around customers, we're all patients in our doctor's office. You need to go in there and upload check in, put some information that's personal to your health situation.
How are you gonna do that without putting your users and your organization at risk? The third most common one, because we lack resources, all of us lack security resources. It could be a mainframe contractor that's in there doing an MRI, for example, and they need to do it from a location and a device that's not in the control of the organization that is contracting them. Leveraging the Symantec Enterprise Cloud, we solve all those use cases. When we're selling to customers now, unlike the old Symantec Corporation, we would sell the bill of materials. Oh, you need WSS and you need Secure Access Cloud, and you need Mirror Gateway. Now we say, Symantec Enterprise Cloud $X per user. The simplification that we're now able to bring...
Oh, by the way, take your time on that legacy infrastructure that you have, and if you need help, we've got partners and our own people to help you get you there. That, you know, for all the FUD out there, just recognize there's a lot of noise in this area around SASE, Zero Trust. For large enterprises, they can't just adopt that solution and turn off all the responsibilities they have. It's not that easy. That's why we feel like we're in a unique position to help our customers and grow value in this part of the business. The last pain point, I wanna walk through this. This might be simple for many of you, and for others it might not, but let's talk about threats.
How does it happen, and what can Symantec do uniquely to help our customers? Well, first, we've all experienced the phishing email. Those tests we get all the time here, and I'll tell you, they're very challenging if you're not careful. On your mobile device, you can't hover over a link there. If you don't recognize it, do not tap on it, trust me, because you are putting your company at risk. In this use case, what happens when a user taps or clicks on something they shouldn't, it can easily get exploited by vulnerabilities that are just pervasive across all these different operating systems. It could be, you know, one patch back on a Mac, a MacBook. It could be, you know, somebody's running a standardized Windows 10 that Microsoft's no longer patching a particular library.
This stuff happens all the time, and now users get tricked on tapping on that, and then at that moment, the user may not know anything happened. It's happening transparently, where all of a sudden that endpoint is communicating to a control, a command and control, center where the bad actor is orchestrating how to get data that's interesting. Most of the time it's targeted, by the way. It's not just a spray and pray. They're looking for something. They usually have information that would allow them to then find the stuff that matters and then ransom it ultimately. This happens, the email's one path. It could happen in SaaS applications. It can happen many different paths. It's the incursion that's most common is email security.
That's where Symantec, you know, we have a really strong email security platform, and when you combine that with our Symantec Enterprise Cloud that I just discussed, we're in a very unique position. We're able to do that and share telemetry because of our Global Intelligence Network that's sitting on the Symantec Enterprise Cloud. Okay? What this does, it collects telemetry, and with the new Symantec Enterprise Division that I've literally merged the Blue Coat, WebPulse, and content analysis network-focused team with the Symantec well-known STAR team working as one. This data lake is able to now enrich customers, not just with one product, but the more products you use, more capabilities or use cases you have on Symantec's Enterprise Cloud, you're gonna get the benefits of all this telemetry. We're talking about 11 trillion elements today of telemetry.
You have to use AI, you have to use machine learning. Just as important, you need a trusted advisor on the phone to help our customers deal with some of these. Most of these attacks have never been seen before. We have a group of threat hunters that spends time with our strategic customers and help them think through this stuff, 'cause this, it's new. We work closely with the government, DHS, sharing information. Recently, WastedLocker, which we blogged on and got a lot of coverage, and Symantec's been used to that over the years. You should expect more and more. We're heavily invested in being relevant around our security research. We were able to protect all of our customers that we know of from getting hit by this ransomware, WastedLocker ransomware.
One customer had to pay about a $10 million ransom. This is real money, and you combine that with disclosure and all the wasted time that executives have to deal with, and the board, and paying third parties like Mandiant and others to come in and help them, it adds up. It's not fun for our customers to go through. We are fortunate because we understand their pain, we understand the dynamics of their environment, and that's a lot easier than just outsourcing it to a tool vendor. If you bring it together, you know, the foundation for the Symantec Enterprise Cloud, it isn't something we just built yesterday. We have evolved best-in-class technology that was acquired best-in-class.
In the last two years, we've integrated it beyond the speed at which never was done at Symantec, and I can vouch for that, being there, you know, seven years prior in different stints. I'm pretty excited about that investment. You're probably saying, "Yeah, Rob, you know, I cover lots of cybersecurity companies. You sound like they do. Tell me what's unique." If I click, there we are. First off, I want to underscore hybrid is our view, and, you know, go talk to your CIOs, not the chief information officer, not the chief investment officer, at your organizations, and ask them where you think the end state is. It's hybrid.
Hybrid work, hybrid cloud, and our competition, and I would call it co-opetition 'cause we coexist with every one of these vendors that you guys cover. I will tell you, they all have great technologies. I'm not gonna sit here and say they don't have good technologies, but they all are focused in a different. Their mission's different. You know, our mission is to take best in class, embedded security technologies that are proven and evolve those use cases to the ones in the cloud in a way that is convenient, in a way that is familiar, in a way that they know they can pick the phone up and talk to someone who understands what threats are, okay?
When you look at the list that's to your left, Microsoft has done a very good job on improving the security on their platform and integrating, in some cases, with third parties. Let's be honest, they're in business to outsource data centers and deliver the best, you know, productivity platform there is in the market. That's what they're in business for. We're in business for security, and we secure their products alongside their E5 licenses all the time. Most customers will tell you, they can't just go all the way to Microsoft and address their use cases. I wanna make sure you guys understand that. I sure I'll get some questions later on that.
Palo Alto Networks, Zscaler, well-known in their areas, and I would say that we differentiate based on our information protection and our threat protection technologies at scale worldwide, and it's integrated in our ecosystem. Yes, we do integrate with other technologies through our Integrated Cyber Defense API framework. Let's be clear, most customers aren't looking to retain all their vendors. They're looking to consolidate, and that's what we're driving our customers towards with bundling that makes sense, that's a win-win. Then, you know, CrowdStrike, I would just say that they just, they're not as focused on the breadth of what I just described, and they're cloud-only at this point.
As you look at these companies, we coexist, and we have the on-prem footprint that many customers just need to know that they don't need to look elsewhere for in order to accommodate their entire needs. Over the upcoming year, we're investing more and more in our go-to-market alongside the Broadcom software sales team to reestablish many of the relationships with the CISOs that might have gone absent during the pandemic and the integration. I'm really excited about where that's gonna go. From an R&D perspective, beyond the go-to-market that I just said, we're never done in this industry. This is one of the coolest places to be 'cause you have to innovate to stay ahead of the bad actors and nation-states that are always trying to get a leg up.
When you look at next-generation cloud-based attacks, leveraging much of our innovations around deception-based detection, integrating this adaptive technology that I talked about earlier, but more broadly into the network use cases and the cloud use cases is an area that you're gonna see us communicate about as we move forward. Around data loss prevention, if I were to tell you one of the most unique things we've got is our data centricity. The fact we know what data matters, and the fact that we can have that context when we're trying to make decisions on what we're gonna protect in a very short window, our DLP is untouchable. It's because of the investment that we've continued around that Vontu technology over the years. We're not done. You have to auto-discover. Data's everywhere. It can't be done manually.
We're leveraging new techniques to get and innovate around that area in this world of hybrid that I've just been communicating about. Then lastly is around platforms. We've acquired a lot of companies. Yes, there's consolidations that, in some cases, that we should consolidate, but many of those we've already done. It's about how you represent the data, how you show that context. When customers wanna manage everything in the cloud, let them. When they wanna still manage elements on-prem, allow that to happen in concert with what's going on in the cloud. It's not an and. It's not an either/or, it's an and. That's where we uniquely differentiate from every one of the vendors who I discussed earlier. Then around the endpoint, which we have an amazing footprint, 160 million devices.
more we can do to bring that data intelligence, data centricity in a more simplified user experience at the agent, that's what you're gonna see us continue to innovate there. As you know, you can't put agents on everything, so you need network to be able to accommodate that when you can't put an agent on the device. My last slide kind of brings us back to the beginning because, you know, being that you're in the investing space and you care about, well, where's the upside? There's a lot of upside in our current customers. Just within the current Symantec Enterprise Global 2000 customers, which there's a ton of overlap in the sense that 68% of them today have at least two or more. If you look at the middle bucket, there's two or three Symantec franchises with just 58%.
Such an opportunity to move them up to have the entire stack. They're going to, if they see the value that I articulated in those use cases. We have to communicate that value. We have to help them get there in a partnership. Of course, there's many customers who maybe are using us more tactically in a particular area of their business, leveraging the power and the relationships that we get with Broadcom Software. With what Greg and Serge are doing, we can expand even broader beyond a single franchise. With that, I think I'm nine minutes early, and I went fast. I'll bring it back to Tom. There we go.
We're gonna just pause for a minute and get set up, and then we'll run through Q&A. Just give us a minute.
Hopefully, these chairs.
They're comfortable.
feel like this thing's gonna fall down.
How is same feeling? Not too bad.
Need a footstool. I think, G, you're gonna pass the mic.
Tom, do you have a preference where we start?
Go ahead. Sure. Right here. Oh, hey. Good, John. Hey, John.
I was closest, Tom. Appreciate all the insights. I almost feel qualified to spell the word software now, but.
Mission accomplished.
Tom, I'm wondering if you could just take a step back and help us understand, if you were to sort of build a bridge PowerPoint chart, where the starting point was the growth rates of these businesses before you acquired them, which was effectively 0%, to the growth rate now of greater than 5%. What are the two to four bars and the magnitude of those bars that get you from beginning to end? As you discuss that, I'd kinda be interested to know what you did with incentive mechanisms at both companies, because we're all a big believer in this room. That's what's rewarded is repeated. Was it just a matter of reincentivizing the teams? Or help us understand kind of that bridge.
Sure. You know, it's not dissimilar to what we did on the semiconductor side for years, guys. You know, when we go in and we buy these companies, we're very much focused on identifying where we can get the best return on investment with the portfolio that those companies have. We take a very different approach. Whether it be CA or Symantec, not dissimilar from Brocade, which you're familiar with, or some of the other companies. You know, first and foremost, it's portfolio rationalization.
When we go in, you know, we did a number of divestitures around CA and Symantec, and that allows us to get focused, and those are businesses that don't fit our model and probably have more value, you know, with somebody else than us. Two , as you saw today, we streamline our business model to focus on, you know, where we think we can win, and then we go play to win. When you think about, you know, what we did in terms of taking companies that were serving thousands and thousands of customers, and really focusing now only on 600, you know, that allows us to obviously rethink, you know, what the revenue growth profile is gonna look like. When you think about a bridge, you know, we also focus a lot more on recurring revenue. I talked to you about subscription. We're still in the early days of doing that. You know, we've gotten out of a lot of the, you know, one-time perpetual license sales.
We've gotten out of selling services just for the fact of maintaining revenues in the services business. It's not generating any profits. You know, I think it always takes a couple of years to work through, you know, that redirection. Now that we're sort of one organic business and we have a number of these growth vectors, we've got the market behind us. We've focused on an ELA and PLA model where we're selling across solutions. You know, that's how we're getting to this, you know, mid-single-digit growth rate expectation. Could we grow faster? I think we could. I think it would require more investment outside or certainly where there'd be more question around are we really well suited to support those customers.
What we feel great about, as I say here today, is, okay, you know, these 600 accounts, they are hybrid. They're typically on-prem first. They leverage a lot of what we already have. We can expand a lot faster within those accounts by taking less risk. We don't have to sell as much. We're really just focused on adoption. And then, of course, we can cross-sell and term and upsell in terms of the solutions they don't have. And especially post-pandemic, you know, I think we got a lot of benefits from capacity expansion during the pandemic, and that was great.
We also, you know, that was a positive, but I think the negative is we didn't have as much opportunity to engage with our customers, especially when Rob talked about going back out and engaging with CISOs and trying to sell Symantec Enterprise Cloud. Now we actually can go and upsell these capabilities. We've been able to continue to invest on an R&D side, but I think there's a lot more opportunity to come. You know, it's to me, it's mostly focus, John. It's when you're chasing a lot of different types of customers, a lot of different verticals, you know, scale, you know, across a broad portfolio.
Yeah, you can, you know, you can win and you can sign a contract, but often you know, sell software, you don't get adopted, and you just have a tremendous amount of churn in the business. The churn has continued to go way down, you know, especially if I look at the strategic accounts, the churns come off significantly and, you know, we're really focused there. You know, we're getting to Bedrock. You know, the Bedrock, you know, on the, in the outside of strategic and enterprise, you know, commercial, you saw makes up basically 5% of the portfolio. All of those things are contributing to the mid-single-digit growth rate that we talked about. Go to the back.
Thank you, Tom. Thank you all for a very informative day. Tom, I had two questions, one kind of on the business itself and then one on the bigger strategy, you know, from a broader Broadcom perspective. From a broader Broadcom perspective, how would you explain to a layman the synergies between semiconductors and what you presented today? Right? That's question number one. Then on the specific, you know, since we, you know, tend to simplify things and we understand them in terms of units and content and pricing, could you give us some color around those dynamics? You know, when you say 5%, is that mostly from content? Is that more from pricing? And then, you know, is there a lot more headroom in any one of those vectors? Thank you.
Sure. Synergies, you know, I think are at two levels. Some of this has been realized, and I think some of it is still to come and there's areas to explore. I think at a fundamental level, you know, when you look at the semiconductor business, there has been a lot of disintermediation between, you know, the OEM, traditional OEM model and the end user, especially in the cloud and increasingly on the telco side. Where you haven't seen that traditionally, but where we think that will continue over time is on the enterprise side. When we looked at, you know, platform technologies on the semi side, whether it be switching and routing, and you guys are familiar with all that, we saw an opportunity to get closer to those end customers.
If you remember, you know, those end customers are also driving not just their direct private cloud spend, but they're also the ones spending money on AWS and Azure and GCP. So having a seat at the table and really being able to be intimately involved in what those private data center build outs look like, where their priorities are, being able to inform them directly as opposed through an intermediary like an OEM around what our roadmaps look like and being able to architect, in partnership with them, I think has quite a bit of value. I think it's more qualitative, frankly. It's harder to quantify, which I know is sometimes frustrating, but I think it enhances the overall value of Broadcom.
I think we've seen the benefits of that, especially now that we have a very defined view in terms of where the enterprise spend is. I think as all of you know, on the semiconductor side, enterprise spend is still a very meaningful, and I think will continue to be a very meaningful part of what drives the semiconductor business. That spend is directly correlated to the software enterprise business that we're running here. I think the intimacy there between the two is valuable, and a lot of that benefit is being realized or has been realized, and I think, you know, we clearly are a major enterprise software company with over $5 billion recurring revenue. I think that's here to stay.
At a second level, now you're getting into, okay, product synergies and where can we, you know, leverage software and embed that in semiconductors or vice versa. You know, Serge talked some about that, and if you wanna talk more about it, we can. We're still exploring those opportunities. You know, we see, you know, opportunities. We have a, you know, a proxy business through Rob's business on the security side, network monitoring. We're doing work, you know, relative to switch and route platforms. You know, have we realized anything there that's material? No. The opportunities do present themselves. Our customers are, by definition, risk-averse. They take time. They move extremely slowly. It's the exact opposite of dealing with, you know, some of our large mobile phone customers that move very quickly.
You just have to be patient, spend time articulating the value. Then also there's a lot of feedback back to our teams in the respective business units in terms of how we would change our roadmaps or adjust our roadmaps to respond to those opportunities. You know, fundamentally, I think there is some really nice synergies between the two businesses, especially when I think about being able to feel comfortable that we have relationships with enterprises that make up such a big part of the semiconductor end market spend. On the 5% growth in parsing, you know, content or usage versus pricing, I think, you know, when.
Whenever we go in and we purchase companies and we rationalize and we're structuring them, you know, often there is a level of discounting that goes on with the very, you know, businesses that we see a lot of value in in order for prior management teams and boards to try to show growth or drive growth in other areas. I'd say earlier on in acquisitions, often there is some fair market value adjustment we have to do relative to where businesses have been priced traditionally. Does that create some level of consternation and churn and that kind of thing? Yes. Fundamentally, these are mission-critical businesses, and we're adding a lot of value to the customers. We spend a lot of time articulating the amount of investment that we're actually putting back in the business.
That's what translates into the second piece, which is really usage and expansion. You know, I'd say you know at this point, given that we've gotten through most of that rationalization and integration, CA's been under the tent for three years. Symantec's now been with us for two years. We're really kind of past that first phase, and really that's what I'm excited about. I think one of the fun things about putting this day together and having you know Greg, Serge, and Rob you know articulate you know what they're doing and really show you the substance of the opportunity is it's frankly all about organic growth and organic performance. It's not really pricing, it's really you know how do we expand our users? How do we drive adoption?
How do we continue to sell, you know, new capability that the customer may not have adopted today into those accounts that already are very familiar with Broadcom's offer? I think that's really the path forward in terms of how we drive growth.
Great. Thank you. Brad Zelnick with Deutsche Bank.
Hey, Brad.
Thanks so much for hosting today.
Thanks for coming.
Really, really helpful. I have two questions. Tom, first for you. Clearly, the recipe in software has been successful. Hoping you can comment on your broader aspirations, the type of targets. I know you're not gonna give us a shopping list today, but even if you could maybe help in terms of where you might not go. Like could we see Broadcom go into application software? Is there a reason why this only, you know, applies the strategy to infrastructure and security? And then my questions for Rob. Rob, it's great to see you.
Yeah.
It's great to see Symantec Enterprise doing so well. I just reflect back on many years past where there were a lot of smart people telling a similar story, and the promise was always consolidation back then.
Mm-hmm.
From John Thompson to Enrique to Greg Clark, Mike Brown. You've known them all.
Yeah.
Why is it today, you know, reflecting on all your years and all these experiences, that we're finally seeing this one difference come true? Consolidation is now happening. Why is it that you think it's sustainable? Maybe just lastly, as part of that, Symantec had always done a lot of tuck-in and some larger deals with Blue Coat, obviously. Is there a need to continue to do more deals just to keep up with the arms race that we know as security? Thanks.
Sure.
Go ahead, Rob. You go ahead, and I'll take the second.
Yeah. I'll answer that. The good news is that the cards we were dealt with the great, I think, foresight that Greg Clark had and before him, Enrique and John, have set us up for really good, you know, success to come and already have gotten some of that success. We don't have a gap right now in our capabilities that. It, it's about execution, it's about integration. It's about, you know, in some cases, making a decision on which engine to use versus another 'cause we got it from two different places. And we've done a lot of that, and we did it because of the culture of Broadcom. It's disciplined. You know, you have limited resources and we invest a ton in R&D.
The type of inspection that's required to justify your investment, you just have to bring it together. It's just the way the model works. I think that is different than when I was at Symantec in past years, where you know, a lot of cases, it's all the new stuff you're building at the expense of what your customers paid for. That's the challenge when you're dealing with Magic Quadrants or whatever the latest you know, industry analyst is putting out there, because our customers actually look to that to make buying decisions. If you're lacking that next shiny object, which Symantec in a public market had different challenges.
You know, a lot of perpetual licensing, very lumpy, if you think about deals that you had to get done in a particular quarter to meet, you know, objectives and that sort of thing. We don't have that problem here because a lot of our focus is about adoption and getting our customers to stay with us as they transform to this hybrid world. All of them are great leaders, so don't get me wrong. It's the focus and the ability to just target large enterprise customers that have common problems has really been the biggest impact.
Thanks, Rob. I think on the strategy and I think what you're really asking is from an M&A perspective. Here's how I think about it. You know, 10 years ago, when I joined Avago and started working with Hawk on an M&A, and we rolled out sort of a game plan. Okay, how is this going to? What can we do? It was really more around, okay, there's compute, there's memory, there's analog, but this is a highly fragmented communications connectivity you know segment of the semiconductor market.
You know, we think we can plot out, you know, a consolidation strategy, which is really not consolidation, but really focused on, okay, we can find, you know, these great franchises that can stand on their own two feet, and then we can get a lot of synergies by integrating them into one platform. Ultimately, they were selling, you know, all these businesses were selling to the same customers or selling to the same OEMs. You know, that's something we saw, you know, early and were able to execute on, thankfully. A lot of the bigger companies that are in the space didn't get in our way. You know, we're all thankful for that today.
I think on the software side, it's not dissimilar in a sense where, you know, we see an opportunity to, you know, now look at the software world through our lens, which is, okay, the largest enterprise in the world, what are the unique characteristics of those customers? You know, what are their needs? How can we construct a portfolio that serves them, and serves them over time with, you know, with what we have today and where we're investing organically? Then, you know, if it makes sense, adding on from there. You know, is it about, you know, double-digit hypergrowth? No. Is it about cloud only? No. Is it about serving, you know, tens of thousands of customers? No.
It's really about, you know, mission-critical software that serves these common set of customers that have unique characteristics. You know, between balancing, you know, what we invest from a go-to-market standpoint with continued investment, you know, in the R&D part of the business, you know, we think we can sustain this, you know, mid-single digit type growth rate, especially as we start, you know, integrating, you know, these different solutions across enterprise license agreements more and more. You know, I'm not gonna get into targets or industries and that kind of thing, but it has to have a common thread around what are the unique needs of the customers. As long as that is there, then, you know, I think from my perspective, the model makes sense. Go ahead, Harlan.
Yeah. Good afternoon. Thanks for hosting this very insightful event. Harlan Sur, J.P. Morgan. Going back to the synergy, I've recently had meetings with a large enterprise IT and data center team. They're looking at moving their networking infrastructure to more of a white box model and away from expensive OEM networking boxes. They're gonna be using our Tomahawk and Trident family of chipsets. Does your seat at the table and engagements with these large corporations on the software side and directly with some of the hardware teams via the AIOps platform, has this helped to facilitate or accelerate this move to white box switching and routing? Do you see more and more of this disaggregation of hardware and software dynamics, i.e., the cloud model, starting to become more of a discussion point with your largest enterprise customers, just given your close engagements?
Sure. Let me take the first part of that, and then, you know, given your question in one of the prior ones, why don't maybe Serge can sort of go back over some of the things he's seeing on the network monitoring side. You know, look, when we bought Broadcom called Broadcom Classic, you know, the number one franchise in that portfolio was the switch and router business. I think what was underappreciated by the market at the time, and certainly underappreciated, but as we saw sort of the engagement model with the cloud playing out, a huge opportunity was really this disintermediation you're talking about.
We had a unique vantage point because we were in the ASIC business, still are, you know, selling ASICs into those very same OEMs. We saw the value, you know, that the Broadcom team was driving around, okay, how do I directly engage with cloud? What are their use cases? Why is it different? That's where this all started, in my opinion. From there, you know, it's evolved. It's certainly evolved, and I think the next sort of vertical or segment of the market that is catching on to this, frankly, before the enterprise, is really on the telco side. You know, they're seeing the benefits obviously of that disintermediation and, you know, taking advantage of being able to be early adopters of next generation technology.
That's of real value to them. I think that's frankly still pretty early. On the enterprise side, you know, we actually see, you know, a similar dynamic to the telco space. I think it's gonna play out over time. I think especially for the largest enterprises that are spending, you know, billions of dollars a year on IT, it's particularly applicable. The whole concept of white boxes and, you know, these standard platforms and, you know, the benefit of leveraging, okay, I'm gonna go buy a box that's the same box that's used at AWS for my own private data center, it's very logical. I think we're really early there.
If I go back to kind of the synergy between software and semiconductors, certainly having a seat at the table and being able to be in front of those developments and being able to think through, okay, well, here's our roadmap, you know, for the cloud that you guys hear about always, and they're obviously talking about that roadmap, you know, well in advance of when we disclose it publicly. Those are the same things we can do with our largest enterprises. That's something we weren't doing before.
I think it's an exciting trend in development, you know, where, you know, this whole, you know, white box market and getting more value for the silicon and the software that sits on top of it, where it started in the cloud, it's moved to telco, and now it's gonna eventually move into the enterprise space. I think it's fantastic for the business, and it's an exciting next step. I don't know, Serge, you wanna spend a little bit more time on network monitoring and kinda some of the things you're seeing there?
Yeah, sure. So I think what's been really interesting is that with the complexity that we see in our customers' environments, right? They've been reinvesting in kind of a network designing their own data centers. What we also see is that there is kind of a convergence between kind of two traditional roles that exist within the enterprise. On one end, historically we've had kind of network operations teams kind of managing the network, running the network, which were very different from network architects, which were kind of responsible for the technical choices and decisions on really the selection of devices.
These two spaces are increasingly coming together, the driver for that is really that when you think about what I was talking about in terms of the scale and the complexity of these environments, we are moving to what many are calling intent-based network management. What does it mean? It means that increasingly, we're trying to, at design time, define policies around how we intend the network to behave and automatically provision configure the network. Then over time, as we monitor this network, as we understand what's happening on the network, we leverage kind of this data to reconfigure and optimize the network.
With these two companies coming together, naturally we from a software point of view are starting to be dragged into conversations which historically we have not been involved with around network architecture configuration, network automation, and that's one of the areas we're further investigating. Naturally, as many of these devices are powered by Broadcom chips, we're looking at synergies where we can directly get telemetry and flow data from the chipset to be able to automatically orchestrate what's happening at the switching and routing point of view. It's kind of early days, but we see the trend and the convergence, and it's an area we continue to invest with, especially our networking and switching and routing divisions.
AI ops teams interfacing with the silicon teams and saying, "Here's what our customers are doing, here's what our next generation platforms are working on. We need you to build these capabilities into your next generation switching or routing or gateway chips, and so on?
I think at this stage it's more about as you probably know, we've been adopting SONiC, right? Today, what truly is unique about our solution is we can directly integrate with open APIs which are provided by SONiC. We can work with the switching and routing division to start to build specific analytics around things like microburst, for instance, to understand how congestion, for instance, is affecting packet loss or is affecting latency. It's way more about how we are consuming the data. In phase two, what we are seeing increasingly is that there is an appetite for us to get into really kind of a day one provisioning and configuration and the ongoing kind of orchestration of a network.
We're having discussions with respect to how much is basically in the chip and the operating system, and how much is going to live in kind of the network management plane that we kind of own today.
Thank you.
Wanna go other side.
Great. Thanks for the great presentations today. Really appreciate it. A question for Rob and then for Tom. Rob, so when Serge is talking about the kinds of hooks that potentially are on the chip side, is that something that applies to your business also? Tom, another question on synergies, if you haven't answered enough of them. You know, I think I guess what a lot of people are kind of maybe trying to get you to say is that, you know, I mean, if you think about the companies out there with hardware and software platforms together, you think about Apple or like an NVIDIA, you buy the iPhone, you're buying the software stack. You buy NVIDIA's chips, you're buying their CUDA software stack.
Like, why wouldn't that be the vision for what you guys are doing? Why wouldn't it be, yeah, you're buying the Broadcom networking chips, and you're gonna have the software stack. Like, why wouldn't that be the, you know, the natural vision where, you know, given all the M&A that's going on, that that's where ultimately you're gonna drive the company to?
Let me take that. I think your answer is pretty simple, Rob.
Yeah.
Let's start here. If you remember, Broadcom is not a single product or single platform company, okay? Broadcom is a portfolio of franchises in discrete sectors within technology today, namely communications and connectivity semiconductors, and now you know, mission-critical enterprise software serving a very distinct set of customers. When you think about platforms and chip platforms and all of the software that you can embed in terms of you know, design kits and other SDKs and all the things you can do up the stack on the semiconductor side, that is different in many respects than what we're doing in enterprise software.
When I think about, you know, the synergies that you're describing, that would be more forced synergies as opposed to, you know, creating value from a sort of a natural evolution of where we're going. Now, that being said, in the semiconductor world, and you know, obviously Serge just mentioned it around SONiC and some of the things that we've been doing, you know, with some of our platform plays, which is primarily on the switch and route side, we do some of that. That's a natural evolution, but that's obviously just one business within a pretty broad portfolio, just in semiconductors.
On the software side, you know, in most respects, you know, these are companies that we've purchased and, you know, we've always valued, you know, ourselves on, okay, their ability to actually, you know, stand on their own two feet. You know, we do see the synergies. I've talked a lot about that today. Just as importantly, you know, these are great businesses in their own right. They've been restructured and rationalized around a business model that we firmly believe in. We're reinvesting in those business organically through quite a bit of product development, and we're doing a great job, I think, of serving our target customer base in a way that's unique relative to the competition.
You know, I see the synergies, but at the same time, I also see the organic opportunity that we have, on the software side. Rob.
Yeah, just to comment on, we have had discussions. Clearly, we actually use Broadcom chips in our proxies, for example. Our STAR engines, which actually power other vendors' technologies, we have the ability to deliver those kind of services, but there's nothing really to announce about that at this point. We're always bringing our R&D teams together to explore those types of things. Again, within an alignment with what Tom said, you know, I have to stand alone as the best cybersecurity technology and cloud for our target customers first and foremost. If that doesn't put at risk that mission to go and do some of these opportunistic things, then we'll explore it.
Hey there. It's Matt Ramsay from Cowen. Thank you for everything, and I'll go on the thesis that a simple question is a good one. We'll see how I do. Again, the grains of salt need to come out. It's a semiconductor guy asking about software.
Tom, in one of your slides, you said that 80% of your top 600 customers take at least five of your ten solutions, right? I wanted to explore that a little bit. Within, say, the average of five that they take, are you single sourced, close to single source? How fragmented is that? And then on the other side, the five that they don't take from you, is that one big behemoth that has the other five? Is that fragmented that you can Pac-Man up? I just wanted to, at a really high level, explore that a little bit. Thank you.
Yeah. That's actually a great question because this is one of the primary differences that we see in the reality of semiconductor versus software. I think on the semiconductor side, we're more used to. I won't call it a winner take all model, but certainly, you know, the amount of share that you can have in one, any one particular area of competence can be really high. In software, there are some examples of that, but if I would say for the most part, there's a lot more fragmentation. Why is that? I'm also newer to software, so I can say this, is I've observed that, you know, decisions get made often in the line of business or for a specific, you know, company in a certain region.
What ensues is a bunch of, you know, consolidation. They buy companies and things of that nature. You often do see vendor proliferation. I mean, I think if anything that's one of the major themes and characteristics of the markets we're in, is there's been a lot of vendor creep over the years. I think the idea that, you know, people are looking to go to the cloud and use cloud workloads only makes that, you know, explode even further. The opportunity and what we're trying to articulate on that slide is, one, we've been very careful about, you know, the companies we've purchased in that they have quite a bit of synergies in terms of the end customers that we're serving, especially as target 600.
On top of that, going forward, you know, we're really in the beginning innings of trying to take advantage of exactly what we talked about, which is, well, okay, how do we solve for all that vendor proliferation? That's not something really the customers really like, to be honest. They can really take advantage of one platform provider that has all of these capabilities that might be using a particular set of, products and solutions from us, but only in one line of business, and they can expand into their other five lines. That's the one use case, that's the expansion piece. The second piece is how do we go get them to adopt, you know, technology and capability that they actually haven't used before.
They may be using Greg's mainframe software for the last, you know, two decades, but they never really got introduced to Rob's business other than maybe through DLP or through proxy. Well, okay, how do we then go transition that to a Symantec Enterprise Cloud engagement and marry that with the mainframe ELA that we have? Those are real opportunities that are in front of us that we're working on as we speak. I think as we think about what drives that mid-single digit growth rate over time, it's gonna be not just that the market's growing and that there is, you know, a natural tailwind, but it's gonna be this ability to expand with what we already have and then go and cross-sell and up-sell around a lot of the.
Frankly, a lot of the good work that these teams are doing around product innovation.
Hey, guys. Thank you, first of all, for a great Analyst Day. I had two questions. One's for Greg. Greg, you made a great case for some of the Facebooks and the Twitters of the world to move to what you called the IBM Z type mainframe computer. The reality is they don't.
Yeah.
You made a great case why they should, but maybe you could explain to them what is the assumption that they're not moving, you know, why are they not moving to it and why they're not able to see the value to it. For Tom, I think you had a slide, Tom, that you're at 46% subscription trying to go to 90% subscription over time. Could you maybe elaborate on the timeframe you're thinking about? That's a lot of share gain. Does the 5% growth rate capture that sort of share gain? Should, in the near term, while you're in that process, the growth rate may be a little bit more elevated?
Go ahead.
I actually wasn't trying to make the case that they should go to mainframe. I would tell you something like Twitter, where it's individual bursts like that, I probably never would recommend it. Somebody that's storing a lot of images and has to do retrieval, boy, the mainframe paradigm would work a lot better. I think, you know, I started talking about the percentage of workload on the mainframe, right? 70% in the world, and I know there was a question in the room I've gotten at back channel about, you know, gee, there's a lot of people that are saying cloud is 20%. Why? You know, and how do you reconcile those things? I think it's tied into this conversation.
There are a bunch of workloads today, new workloads rolling out that would perform more optimally on the mainframe. The reality is if you're not already there is some level of bias that's saying the entry point to start is cheaper on the cloud. Even though over time, you would see that the ongoing cost of that business, and as you expand it, is gonna be more expensive to run in the cloud. When we say or when anybody talks about the percentage of workload that's on mainframe in the world, we're talking about volume of data, we're talking about transaction throughput. We're not talking about the number of applications. I would tell you probably 10 years ago, I think it was 75%-80% of the workload that was on the mainframe.
It's not that the amount of workload on the mainframe is declining. I don't want any of you to be confused by that. It is consistently growing and expanding. What's just happening is these other technologies that have come along have entered into the marketplace and then they've grown. You look at cloud and what's going on there. When we talk about cloud growing, it's not just what you think about in the cloud. It can be cloud on-prem, and people start talking about hybrid and connecting them together. It can be as simple as a bunch of these software companies that said, "Hey, I'm no longer going to ship you a disk or a key with the upgrade to my software.
You're going to download it from my website." Now they claim that as cloud software and cloud revenue. Mainframe's not declining, it's growing. There are workloads that I see today in the world that should start on the mainframe, and they don't always. Some of them do. It's a small number each year. But there is significant growth going on, and it will well into the future.
Harsh, on the subscription transition, put it that way. You know, when we bought CA and certainly Symantec, there was you know, residual effect of the fact that a lot of perpetual license had been sold traditionally, which meant, you know, from a recurring revenue basis, there was a fair amount of maintenance. You know, the big driver of subscription growth is really the upgrade cycle that we're driving across, you know, these three businesses and others. And that's a big deal. It's actually probably the leading indicator in terms of, you know, where we see the portfolio going in terms of its health over time.
We've started out, you know, last year, we put the business and the software group together, basically focused on we're just gonna sell subscriptions. When we go sell an ELA or a PLA, we're gonna sell a subscription solution, and that's gonna replace often a number of standalone contracts that are typically on maintenance. You know, we're at 43%, as you said. I think we're more over 50%, you know, exiting Q3. That number is going to continue to grow. I don't have a specific timeframe in terms of when we're gonna get there, but the goal is actually to sell, you know, exclusively subscription licenses over time. I don't think, you know, there'll always be some residual, you know, maintenance revenue, I suspect.
You know, I think our goal is clearly to get that up above 90% over time, and if we continue to execute on our plans, we'll get there. Hey, Ross.
Hey, guys. I'll go with Matt's line of being a semiconductor guy, so forgive me. One of the financial metrics that you put up, Tom, I think the most impressive one was the sales and marketing going from 29% to 7% of sales. The question for these three guys, and forgetting that your boss is sitting to your right. Is seven percent enough to do what you wanna do? Did you have to leave anything on the table by cutting it that hard? And how do I reconcile that low number with each one of you had great presentations on increasing your customer interactions. I think your value beyond the code thing, Greg.
Mm-hmm.
You talked about. That seems like a lot of sales and marketing effort to me. I assume your answer is just gonna be, "Well, we're just focused on a smaller amount of customers, so that allows us to do it." At the end of it, is that number the right number? Are you walking away from anything? How do you give those customers enough service at such a low sales and marketing budget?
I think you spend an inordinate amount of money in sales and marketing going after smaller customer footprints. We can get way more value to our business and drive more value to our customers focusing at the top. Another thing relative to the Broadcom model that's been really interesting, and which has allowed us to do the beyond code offerings in the mainframe space is education and services in most of any other computer company, software company, certainly my competitors in the mainframe space, runs as a horizontal, whereas I have that directly in the division. You actually hit on both things. It's focus on the customer, and then within the division, the direct control I have of all the mainframe skills that were scattered and fragmented, each trying to kind of focus on a little bit of a niche.
I can really focus them on driving value to the customer so that they're successful with our software solutions and make sure somebody isn't suboptimizing saying, "Oh, I wanna try to make $1 or $2 more dollars off of them taking one more class or doing this." Focus it on the need and the software deployment, align the education at the right time and the services to do that, and you drive efficiency.
Does somebody wanna answer if you missed out on anything?
You know what? I'll just say.
I don't feel like I did, so.
I'll just say this, the benefit of having Broadcom Software is you get to make bets in one area at the, you know, because you're generating so much success in another area, so it's a balancing act. It's diversification. I'm very happy with where we're at. Just remember that one of the cultural changes is as a standalone company, Symantec Corporation tried to serve everybody, where our mission is very focused. Yes, your answer that you had earlier is we absolutely stay focused and get the benefits from the same sales organization that's serving those accounts. So far so good.
Yeah. Go ahead, Serge Lucio, and I'll add one main comment.
If you think about the traditional enterprise software model, you have a huge cost of acquisition of new customers, right? The classic ratio is that you need to generate 3-4x pipeline for conversion. You spend a lot of time doing POCs on ultimately people who are going to yield an ASP around $100K for my side of the portfolio. You have that, which was a CA model, which was terrible. On top of it, you basically have people, you know, sales guys who are incentivized to basically just sign new logos. There was a complete bias at CA in terms of a sales conversation to really focus on signing new customers, again, at $100K a pop.
With the model we currently have, we're really focused just on these 600 customers. You know, I look at my churn. My churn has improved by, you know, like, a huge amount in just the last 3 years because, again, we're very focused on the 600 accounts, and we're not focused on acquiring new customers. We're really focused on kind of developing relationships and expanding at these accounts. That's really the significant shift.
There's also the reality in most of these software companies with a broad portfolio where they have sales teams that are only, you know, get commission or incentive, incentivized to sell one pillar of the stack. What you end up doing in that sales model is you overfund, and you actually have this action that is not only competing with the competition but is competing with your own company for the next dollar. In Broadcom, with the clarity we have around ESG, we're focused on the customer's problem first and figuring out, okay, what is the next best opportunity? What is the thing that's gonna drive the next best value to them for that next dollar procured by us? Then, yeah, we'll get around to the other stuff later, but it's not competing and expending effort against each other or conflicting.
Yeah. I'll close with this 'cause we're out of time. One thing I'm excited about, Ross, is that, you know, when you look at that model and we're able to drive this mid-single digit growth rate that we feel real comfortable about, that allows us to reinvest back, not just in R&D, but also in sales and marketing. Actually, when we think about the plan for next year, we're actually gonna grow the sales and marketing dollars year-over-year, and that's because we've got some leverage in the model, and that's supported by the growth rates we have. I think we've got it tuned in a way where, you know, we can drive mid-single digit recurring revenue growth. We can sustain the gross margins, you know, above 90%.
We can reinvest back in R&D and continue to increase that investment over time, in concert with some of the expansion and cross-selling that we're doing. Then we can engage and support the customers as we define them, and continue to actually be able to grow some investment there. You know, I think that's gonna give us a little leverage in the model. You know, we're running 70%+, and we'll get a little leverage from the revenue growth, but it does allow us to keep reinvesting and actually increasing the dollar spend on a year-over-year basis. Thanks, everybody, very much for attending. We appreciate your time today.
Thank you.
Thank you.
look forward to continuing to engage.