MoffettNathanson. Very happy to have with us David Obstler, who's CFO of Datadog. David, thanks for joining us. Really appreciate it.
Thanks a lot, Sterling, for having us.
Let's just jump in and hit some things head-on. There's been a lot of concern over macro headwinds, et cetera, we watched a lot of the public cloud hyperscaler revenue growth deceleration. You guys still put up greater than 30% growth i n the last quarter. How are you able to outgrow what the hyperscalers are doing?
Yeah. Well, we are correlating. We've always said that we are correlated to workloads that are put into the cloud. There are some differences. One, we are most aligned towards modern development workloads. The evidence of that would be workloads, the applications built with containers, et cetera. Those growth rates have been different than the clouds because they've been higher, because the clouds have a much broader set of workloads they're handling. Secondly, we have been expanding our platform and from the initial infrastructure to the 17 SKUs we have. Some of our products are synthetic testing or log ingestion or login indexing. As the platform has expanded, the growth rate has been higher than the growth rate of the hyperscalers. There's also market share with consolidation on the platform.
That's been a growth driver. In some cases, we are correlated, meaning if the hyperscalers are decelerating, we've tended to also decelerate, but we've tended to maintain a higher growth rate than the hyperscalers over time.
AWS in particular commented, you know, that in their call that the month of April saw further deceleration. O bviously, you reported after, so you saw a lot of that in and g ave the guidance that you gave. What gives you the confidence? What did you see in your business?
Well, first of all of our guidance always assumes more conservative assumptions than what we're seeing. In the guidance, we haven't changed. You know, you've known us since we went public. We've always given guidance, assuming more conservative drivers on both net retention as well as organics. That's one thing. Secondly, I think what we said on our call was that since the market changed, which was in the middle of the Q2 last year, we've seen the organic growth rate or net retention decline. We said in Q1 it was a little bit higher than in Q4, but not as high as Q2 and Q3. What we said about April was that it was too short a time, and we hadn't seen any difference than that broad trend, which was essentially lower than it had been before.
I think I don't know if they gave more insight than we did, but we said that April looked like the similar months in the affected period, meaning not up or down relative to what we've seen since the market changed in the Q2 of last year.
One of the other things that you called out was kind of the patterns you're seeing in some of your larger customers. You know, any more color there in terms of what was driving that behavior?
Yeah. Customers that had ramped very rapidly, either their business ramp, they were in sectors where they ramped or their use of the cloud ramp very significantly, have tended, and we said this starting in Q2 last year, to be more focused on cost containment. Because the spend was larger, there was more to work on than customers that were either younger and just ramping or were smaller. That's continued from the Q2 through what we said in our comments on the last quarter. In some cases, that optimization or that work is leveled off. We said we don't know if there'll be another wave, but that, you know, has already been optimized. In other cases, that work is still going on. That's what we saw.
The difference between a combination of the amount of spend and how quickly it grew, whether you were in certain industries, and affected how much you concentrated on cost management.
I was remiss in kicking off to remind people that there is a QR code up on the screen. If you do wanna submit a question, scan the QR code, and it'll come up to me here on the iPad. Let's drill into that cost containment or cost optimization. How does that either directly or indirectly impact the business? You know, I believe you also rolled out some tools to actually help customers in that process.
We did. Our business model is that is land and expand. Clients generally can use the platform. They can use the platform based on a commitment over, let's say, a year of capacity. We charge that from we meter. If they go over that, they have an on-demand. The way that this manifested itself was not the cross-sell and not the unit price, but on the speed of the number of units consumed. They were clients who were cost managing and optimizing level that off more than they had. What that means is, because our clients have not tended to over-commit, they've tended to stay short.
That means generally they're spending their commitment as they had before, but they are not as quickly getting to their next commitment, and that's affected the rate of growth, of the client, mainly in the number of units consumed.
That makes sense.
Yeah.
Have you seen, you know, a change in appetite in terms of the duration of that commitment?
Yeah, a little bit. Clients have, over time, as they become more enterprising and larger, have wanted to go multi-year a little more around the edges. We don't drive it that way. We don't incentivize to it. Clients have moved a little bit, but this is around the edges. Essentially, the average contract duration has moved out a little bit, but not materially relative to what it was before.
That's also you kinda matching that to cash flow. Your payments and collections also are similar in terms of that duration. It's not like you have a large multi-year collection.
We really don't. Very few of our contracts are multi-year. There are some times that we have individual bills, I know this is one of your questions, that create a departure of RPO or billings from revenue. Weighted average, that's noise in a sort of a year, and it tends to head back pretty much pro rata to revenues.
Yep.
Yeah.
Just I wanna come back, close the loop on the cost optimization. Is there a sense... I imagine it's difficult g iven you have so many customers. I s there any sense like what portion have kinda started or gone through the process? Or, y ou know, is there a way of saying, like in baseball analogy, are we in the fifth inning of this, the seventh inning stretch, or still in batting practice?
I mean, we're assuming our from what we've seen, we're assuming this will continue throughout the rest of the year. That's what's implicit in our guidance. That is largely based on the fact that we're not economists, but generally the Fed won't stop until what they're doing is going to have an effect on the economy and on inflation. In looking at past cycles, economic slowdown has continued past when they stopped tightening, so we're just assuming that's going to continue through the year. We're not smart enough to know. We may have that wrong. We think it'll end. We think we have a very long-term opportunity, and that clients can't optimize forever.
We also see that in the fact that our gross retention is staying very high, and that means clients aren't leaving us, clients are modulating the spend. You know, we think this will be over, but we were very careful on the call and otherwise, you know, not to make a call. We're not smart enough to make a call of an exact month.
Within that, are the optimizations just narrowing the speed of the expansion? Or, you know, do you have customers that are reverting some of the workloads back out of, you know, public cloud?
Not really, no. What we have is for the most part, we have clients flattening the growth of spend. Or in the case of a few examples we've given, their business has been substantially disrupted, meaning their volumes are less. In those cases, that's not moving, volumes out of the cloud, that's the volumes are declining. For the most part, most of our customers have tended to flatten more their growth than replace, we haven't really had that, or reduce.
I can imagine that, you know, the immediate thoughts are gonna go to financial services, crypto, some of those areas. Have you ever articulated, you know, what part of the business are, you know, associated with, with various industries?
Well, we said we're very well, very diversified, we don't have concentration, that we represent the digital economy. We said crypto is very, very low single digits. You know, not a lot of exposure. We said that consumer discretionary, which was, you know, e-commerce, food delivery, that type, we said that was very low double digits in exposure. We said at the time of COVID, we said that travel was around 10%. We said that we don't have any dominant sector, financial services is one of our largest sectors, which is a good thing because almost all of it is things like insurance and credit cards and we have some payments companies. It, you know, it's tended to be from just the customer names, it tend to be pretty stable.
We don't have regional banks. It probably has most to do with what they do in their business rather than us. We don't have, you know, we have a lot of boring insurance companies and things like that. That's good. That makes you feel good at this time. Yeah, it's one of the larger, but it's been very stable. again, except for crypto, if you take crypto out, we don't have a lot of crypto.
Yep.
Mm-hmm.
No.
Yeah.
That makes sense.
The other sectors that matter are, tech, SaaS software, gaming. We're really the digital economy. I would say if financial services was 75% of the digital economy, it isn't. It's, you know, maybe 20% or something.
Yep. yet as you're alluding to t he tech, the follow-on question that I get is, you know, e xposure by size of company. You know, 'cause its usually people asking what about your exposure to small tech.
Definitely.
Start-up tech. Yeah.
We're about 1/3 enterprise, which is 5,000 employees or over. We're about 1/3 middle, midcap or midsize. We are midsize. Datadog's midsize, 5,000 employees. I don't know if you consider Datadog at you know, $27 billion market cap to be a, you know, an SMB, but I don't think so. That's. We have 1/3 that is 1,000 employees or under. We haven't said 100 employees or less, but I'd say we're pretty diversified across that. We haven't seen a cycle so far of death of company in any material. We don't know. Maybe that'll happen, but that hasn't happened.
Another thing that's because the gross retention in this SMB or 1,000 companies or lower has been stable, hasn't changed through all of this and has been in the mid-90s, that indicates to us that as you're trying to save costs, there's certain things that you have to have. If you have a digital business, you're most likely gonna have a monitoring. You're probably gonna do a lot of other things before you cut there.
Yep.
Yeah.
With, you know, you made the comment, optimizations are finite. You can't, you know.
You can't. Yeah.
You're not gonna cut, cut to zero. As it k ind of plays out, is there something that you see as a catalyst to re-accelerate demand?
Well, one thing that is interesting is that the cycle of the caffeinated cycle, or COVID, or whatever you wanna call it, we said on our call broke in May of last year. When you think about it, when your comparisons are versus that period, and you're at a different level of consumption, you're going to be decelerating until the period you're in matches the period that you're comparing it to.
What we said was that Q1 was slightly less than Q2 and Q3 last year. If that was to continue, then it just mathematically, there would be some slighter deceleration. You're lapping those compares from a different environment. That means the rate of decel is likely to go down, but then the question is, do you accel? Then you have to start seeing the weighted average spending of customers grow at a higher rate than the period of a year ago. We said on the call, we haven't seen that yet, and that's what we're watching for. If that happens, then there's a little less cost pressure because essentially most of our revenues in the near term are created by customers we have already.
Most of our revenue growth in the future is by customers that we're getting. We said that's been fairly stable. That's the good news. We've been getting about the same amount of business, but it's the existing customers that have continued to exercise that cost constraint.
On those new customers, w hat's the land experience like now versus maybe a year ago? Are you landing larger, smaller, and what's the profile?
It's very similar. Our Q1 was very similar to the land, the same land we had a year ago. Tend to land with two products. We tend to land in a huge distribution between sort of landing and replacing what's there. You might have some multimillion-dollar lands all the way down to really small. Remember, we're land and expand, so most of it is on the smaller end. The average land is gone up slightly over time, but essentially has gone up like this. I would say the landing has been very, very similar in the Q1 we just experienced versus the Q1 we had before the change of environment.
That's both on number of modules, as you said, a s well as kind of the use cases.
Mm-hmm. It is.
Um-
The spread. Some traditional industries. I think if you read and listened to our earnings and you heard that, you saw a variety of industries, you saw cloud native, you saw expansion, you saw takeouts, you saw. That's been going on for, you know, so quite some time and, is still what's going on in the land cycle.
That early expansion phase, you know, given the comment. It's more the larger customers have slowed.
Yeah.
Landing is about the same.
Right.
Is the expansion about the same as well?
Expansion is higher than. Large, larger already expanded have the lowest expansion rate.
Right.
The next would be in terms of the size would be the smaller expanding, and then higher than that would be the new landers. In the first year. They are still growing higher than the existing customer. We've always had the initial land and expand cohort be faster than, you know, a more mature cohort. All those cohorts are lower than they were at the top of the caffeinating period.
Okay.
All of them.
Okay. again, when you say faster, on a percentage, not a dollar terms, because the figure-
Percentage.
Yeah.
Yeah. Essentially, if we were I'm making something up. If the average was a 70% growth rate for those that landed one year, the comparison of the spend when they landed a year ago and today, that might be 50% today, which is higher than those other cohorts, but lower than what it was. Cost consciousness and the business effect is pervasive. It's, you know, it's everywhere. It's just the relative intensity of it is different in these cohorts. Yeah.
Has the competitive intensity changed at all, given the environment?
It hasn't. We still have, you know, winning, we believe, increasing market share in new business and we believe in the areas where we are strongest, which is modern dev development, client-facing applications that are high priority, meaning they can't go down. That's what we've always won at. It hasn't changed. We haven't seen a movement, and there's been a lot of writing about moving to insource or open source or. It hasn't really changed. We still think long-term, the movement's towards the package solution. It's for some time been net balance of movement towards the platform, meaning consolidation on the platform. The competitive environment has not changed very much in since the more difficult economic times have occurred.
I imagine you see different competitors depending on that third, you know, in terms of size of company.
No, we see different competitors based on the business activity. You could have a modern dev critical application in the US government or the largest enterprise where they would choose a Datadog or in an SMB. Who conducts that activity hasn't largely been determined. It is what the activity and the function is that's determined whether Datadog wins.
Is the buying process, you know, different? Meaning, in other words, are these all RFPs where you're working down to a shortlist, or because of your position, especially with the hyperscalers, you're actually winning some business without necessarily seeing that RFP process?
Yeah. A lot of it, a lot of it is greenfield and doesn't go through RFP. One thing that I think has intensified in this economic environment is more scrutiny, so more procurement groups being involved. That's manifests itself mainly on existing customers. I think you will see, I would say more, whether they mean it or not, I could choose this, right? The gross retention indicates they're not moving, but this is the way that procurement departments are operating more aggressively in this economic environment of more cost management.
Besides the cost optimization, d o you find that the procurement process, it's lengthening but not stopping? I hearken back, if we remember the Great Financial Crisis.
Yeah.
We got to the point where it's very quickly, "Hey, we just don't wanna do this project at all. Let's revisit this a year from now." You know, v ersus, you know, are we at the stage where it's just lengthening sales cycle, but things are still, you know, coming through?
No. I don't think we even said the sales cycle is lengthen. It could be. Since the new logo production in both, dollars and amount has been very stable, you know, over the last five, six quarters, we haven't seen much of a change, and we've seen that the projects that are important and are prioritized continue on. We always have, you know, at the end of a buy in maybe some size, the back and forth, whether it be security or price or whatever, that's buying. That's happening. I don't know. We haven't seen the pipeline conversion from what we see in qualified pipeline entering the quarter change that much over the fast five or six quarters. Maybe we have, you know, but it's hard to tell.
It's really hard to tell when you look at Salesforce data and you analyze it. The fact that we have had about the same number of amount of new business so far indicates that that's continuing versus the growth rate of existing customers which had that change.
Does that mean that including those existing customers, that it's the top of funnel activity that's changed, that the conversion rates are similar?
No. It's the existing customers that are already using the product and essentially are up for their new commitment who are over that period of time saying, "Whoop, somebody should look at this and see if we're using all of it." That's what's been happening to us. The new business has been steady and solid. Like I said, the first quarter had more logos gross than the first quarter last year, and it had about the same amount of business. The Q1 is always lower than the Q4 every year. The only conclusion we can have, maybe it's a false conclusion, is that that buying of new business is stable and hasn't changed very much. Yes, there are other things that could account for that. One, maybe we're distancing ourselves in product, so we're winning more.
Maybe we're getting better at selling. Maybe we do have a larger sales team. Those are things that, you know, it's hard to know all that, right? The amount of new business and the breakdown of the new business that's been created, meaning they weren't a customer before, is very similar to what it was a year ago. It was in the Q1 .
You mentioned, though-
I know others have said something different. We're just reporting back just what happened to us. Yeah.
No, we appreciate that cause that gives us a chance to, you know, to do that comparison 'c ause we are seeing different experiences with different companies.
Yes. I think there are difference. I think the fact that, I think a land and expand type of motion, probably, is able to continue more. Like, when everything is top-down, you gotta do this in multiple millions for three years. We don't do that that much, so We're not a data observant, but I would assume that's a harder thing to do in an economic environment than land and expand and start with $10,000, $50,000, I would assume.
Yeah.
That's my understanding of human nature. Again, we don't do that. We're not a top-down, so maybe that's indicative of the fact that our land, the land and expand is easier to keep chugging in more difficult economic times. Could be.
Yep, s maller chunks, I think, are easier.
Yeah. Could be.
You'd mentioned the larger Salesforce.
Help investing at capacity. The quota capacity will be increasing at a higher rate for this year than the amount of headcount in the year.
You'd mentioned on the retention but easier to hire, I would imagine, as well.
It is easier. Yes, it is. I would say that market, it is. I, you know, to get the right people, it's still not, you know. We're swimming around in people, and tech and good tech people in certain areas, are still, you know, challenging, but it's a much more benign labor market than it was last year.
Have you changed, like, the mix of how much of the business is coming through partners like the hyperscalers versus what you're doing direct?
It's always. We don't care. We essentially wanna deliver it in any way the client wants. We're a heavy player on their marketplaces. The main advantage of going through the marketplaces is that the clients can use their hyperscaler purchase commitments. That's the main advantage of it. That's been increasing over time, but we essentially, we're indifferent. We basically, however you wanna buy is fine, and so are the hyperscalers pretty much. You know, that's been increasing as a percentage. We also have slowly increased the percentage from the systems integrators. Again, we're not the kind of a sale where it's not like a channel sale where we don't even touch it. We touch it together, we're always working with them in co-selling.
The system integrator is a little bit tough given how fast you can actually deploy the solution.
Exactly, yeah.
You've said on a number of occasions that, listen, the best way to judge Datadog is revenue. It's not billings, i t's not RPO. Why is that?
It's that, and then it's the co-computation of that to ARR, and they're very linked. There's a slight, you know, adjustment you have to make based on linearity, right? Which we've gone through. The reason for that is that we don't engineer the company for when the bills go out. We are lucky enough to be cash flow positive and to have been for a long time. For us, we're not trading off paying, like in a big lump sum up front for the structure, pricing, and value of the deal.
Mm.
The billing is an output. Because we don't engineer the company that way, it isn't. The little micro things that make it farther away aren't that helpful. What we do is every time they get out of whack, every time there was a big bill that went out or it was a different calendar day, and if you look back through our earnings, we pro forma it out either way. If it's above, we go, "Don't look at that as this." If we go. We're not being selective, we're always. The reason is that the driver of revenue and production for Datadog is not about when the bill went out, it's about the client's use of a commitment, and that's the reason.
With that being the case, you put yourselves in the shoes of an investor that's trying to understand kinda directionally if the business is getting better or getting worse o utside of revenue, is there any other things that you would be looking at?
Yeah, the linear is so we've, you know, of ARR. Basically, that's what we said. There was a factor in Q4 where we set all the time. We're between 34 and 36. What that means is that the run rate, exit run rate of the business is higher than 1/3, you know, then multiplying the revenue times 4, right? We basically say when we say it's normal linearity, what we mean is it's around 35. You can use that. When we say it's not normal, it might have gone down to 34.2 and maybe a little above. We give little tips like that.
Yeah.
That's the thing. I think that the other things, billing and RPO, if you look at the average of that, and you basically take our pro forma, then you can get something into it. It all gets sort of back. You see, if you do the pro forma of everything that we did on the call, it all is around 30 some, you know, low 30s. I think there's something in it as long as you remove all the things that are not related to the recognition of revenues going forward.
Yep.
Yeah.
Yep.
Yeah.
On the split of the business, y ou know, the cornerstone o f the company was built on infrastructure. It's been around the longest. You talked about how t he APM and log, you know, suite has gotten to $1 billion, y ou know, rate. What does that tell us around the rate and pace of, y ou know, the different, you know, franchises, i nfrastructure versus?
First of all, we have a platform, and actually we often don't even control this. Basically, they have all the functionality. They're looking at using the observability platform, and we wanna make it that way, and that's one of the things that's happened. That's one thing. Two is we've always said that if you, if you are fully ramped in buying the other two, you're gonna be on a spend between two and three times infrastructure. Guess what? Right now, we just showed it's pretty easy math, it's two times, right? On a weighted average, and that's including the fact that there are a lot of customers that are still not using the other two.
Yeah.
If they all use the other two, then it goes more towards the three. What that means mathematically is that the rate of growth of infrastructure has been higher than the hyperscalers. Still is. It's been, but this has been for a long time, lower than the growth rate of APM and logs as penetration of infrastructure companies who then take on APM and logs, which we didn't have five years ago happened. That's still the case. If you go back and look at all of what we said, you'll see that it's been going like that, meaning if this is infrastructure, it's been going. I think Yuka said there are some little things in your math. I think your rate of growth of infrastructure is lower than it really is.
Maybe, maybe it's just some little details, APM and logs have been higher than that. I think that they're getting I think that APM and logs most probably are growing still a premium, a significant premium, may not be as high as the premium was when we had no penetration of APM and logs on infrastructure base, pretty much by definition.
Yeah. Yep.
Yes.
Log small numbers.
Log small numbers. Exactly. Exactly, yeah.
You know, along those lines, you had mentioned the 17 SKUs a re out there. What are some of the programs or things that d oing to help drive that adoption e specially in this environment?
Well, we have a whole department of customer success whose main metric is none other than driving that adoption. Essentially, we have a commercial sales team for the non-enterprise that lands. As we said, most are landing with two modules. Then we have a large department of account managers and a large department of technical account managers that do nothing other than help clients discover and use the additional products. We invest in that more and more given the breadth of the platform. Then we have in enterprise, essentially the targets and the way it's structured is there are certain people in enterprise who are spending more time on hunting and new logos, and there's others that are spending more time on upselling the product suite and getting to new divisions.
A very large percentage of the go-to-market organization does and always has been about upselling the various parts of the platform. It isn't that we have to get, you know, a new sales, because in a platform sale, they can also self-discover. There's another sales team out there that is really important, and that's the customer who's actually using it in a frictionless way, and that's really important for Datadog.
I think even prior to the IPO, you had those customer success teams.
That we did. Definitely, we did.
Yeah.
Yeah.
Across the customer landscape.
Yeah. It's the same. Basically, whereas when we went public, you had infrastructure, and then they were working on the infrastructure, and then they were landing APM and logs. Now they're landing... Sure, there's 17 SKUs, but there are families of this, so they're working on landing the families of the different SKUs.
T his conference, you can't go any sessions without talking AI. H ow do you think about... Let's say first, how are you using AI to actually change the way that you run a business?
It's good. I think there's three or four ways that we're we're looking at AI. It's very early. We are essentially very early in experimenting with everything from the marketing department generating collateral to developers. When, I mean very early, I think there's probably been a real big excitement. Commercial organizations have to look at their practices, their security, their governance, their privacy. When I mean early, meaning we have some users of it.
Yeah.
You know, we're essentially plan to increase that over time. We think there will be inefficiencies for us internally that may be, as you know, copilots, as they say, in helping everybody do their job better, or it might be over time, ways of creating more menial tasks that are being done automatically. That's how we're using it. Secondly, I think we always have had the type of customer that's an AI customer is a good customer for Datadog because they're basically a cloud-native analytics data company. We highlighted one in our speech, in our prepared remarks, but we have many others that had, for a long time in AI, been Datadog customers. We think they're perfect Datadog customers 'cause of what they do, how they provide their service.
You have the fact that pretty much in the history of Datadog, any time there's been a technology change that has meant you have to organize the development of software that runs over data, Datadog. All of these models are gonna require that. Meaning they're dumb unless they run off of organized data, right?
Right.
Training. It's too early, but we think there'll be lots of workloads and cases that are perfect for Datadog in our customer base. The last thing would be in our platform itself.
Yeah.
Which is to create greater utility to discover problems and use. We've always had machine learning and automation, and we've been on that journey for a long time, and we're optimistic that we're gonna be able to add parts to the platform. We have a number of things we're working on. We tend to announce things at DASH. You know, we're working on different things to see if we can, you know, find the things that will be most useful to customers and then have the time to get customer feedback to sort of be on that schedule.
Yeah. It would seem like, y ou know, the companies with the largest data sets a re perfectly aligned to benefit.
Yeah.
Given what you've done with logs, in particular, it would seem like putting it into the platform t o enable your customers to be able to Beyond just, like, is your application running, i t's more, can you actually put some of the utility into the log system itself?
Yeah, definitely. The announcement we made, which is an API integration to OpenAI, we have over 600 integrations. Our, you know, at the very core of the platform, we want all of the software that our clients are using to be able to be monitored and how that's interacting in a production environment. That was what that was. That's like a, you know, that's like one thing you do, but that's not embedded in the platform, and that's what I was speaking about in my previous remarks, about embedding the large language models within the platform.
Do some of these kinda newer areas bring an incremental level of investment? Where I'm going with this. You know, your gross margins have actually run, you know, above 80%, k ind of above the long-term target.
Yeah.
You know, do you leave those long-term targets 'cause you think that there is gonna be this incremental investment, you know, to handle some of this incorporated into the platform?
We leave the long-term targets to be able to make investments in the infrastructure platform, which we've had at different times, so that we can have inefficiencies that might take it down in the short term, but will enhance the cost-effectiveness of the platform long term. That's what happened in our, in our Husky, which is, you know, the data stores.
Yeah.
It could happen here. Everyone asks, "Wow, you've been running against that a long time." And we might. We might run at that level forever, who knows? We wanna have the ability to, essentially work on the platform. When you work on the platform, you have to leave the old stuff running 'cause you have existing clients. You have built-in inefficiencies. When you have a new data center. All of that is to permit us to do the right thing in scaling the platform over time.
You've been very diligent in terms of the margin improvement i n light of, you know, the balance between growth and profitability. One of the questions we're starting to ask more companies is, well, wait a minute, you know, let's say whether it's later this year into next year, whatever it is, when we start to see economic improvement. Do you think that you'll rethink again and say, "Hey, listen, if the opportunity is there," you'll take advantage and step on the gas and maybe give back some of the margins to capture more market and growth opportunity?
Yeah. Hard to know. It's really a trade-off between the scalability and economies of scale in the business, right? When you get customers and you retain them and grow them, it's more efficient than if you had to produce that new business every... You know, there's, you know, and do you know, there's lots of efficiency. It's really essentially that economy versus the investment. So far, we've been able to balance that. We pretty much, you know, have given some guidance for the rest of the year, and we take what we say seriously, we wanna balance that.
I could see a time when there'll be investments that'll be no-brainers, and we'll tell everybody, "This is what we're doing." I think we wanna leave that flexibility, but we also understand that, you know, we're an over $2 billion run rate company and that as software companies grow, they're maximizing the free cash flow compounding over the long term, which is a combination of the top line and then the profitability, and we're trying to balance those two things.
We talked about infrastructure, we talked about APM and logging. Y ou know, let's kind of dive into security a bit with the time that we've got.
Okay.
You know, in particular, where is that kind of security attached within the customer base? Let's start with that.
Yeah. We said, and repeat that it's being attached to the customer base. We're not selling security standalone, so far, without them having our DevOps. It is attaching successfully in places where security is relatively near DevOps, the functions or the ability of DevOps to use security is either happening or being permitted. Like we had in APM, the state of the development is that it satisfies some use cases, but not all use cases. It's being successfully attached to our existing customers, when we said we got to 5,000 customers. We still have quite a bit of build. We have to get to product parity versus some of the other competitors which are not there yet.
Then we have to then from there figure out whether the, o ur strategy is gonna be to lead with security into the CISO, which is more of a top-down sale, or to have it in, again, specializing in places in a more fulsome way, where it's attached more to DevOps. We've said all along, and as the questions come out in specialized sales team, et cetera, that we don't know the answer right now until we have the product fully built, we won't know the answer, and we're happy with the success in attaching it where we are, but know that we have to build more functionality in order to be a leader of this long, in long term.
For those that are less familiar with the security segment. When you talk about parity with competitors, who are those competitors that you think about?
Yeah. This, by the way, is not endpoint security. This is cloud security or cloud SIEM. It is monitoring the workloads in production environments or the applications in cloud. It is not monitoring the endpoints or laptops or others inside the company or firewall. The leaders in that today are Palo Alto through their Prisma product.
They, of course, are one of the larger security companies. They have made significant investments in cloud security as well. There are emerging companies with the names of Wiz and Orca and others like that. There are, as you shift further over towards the development side, this goes more into pre-production. You have companies like Snyk. Essentially, that's the playing field that you have today that is in this cloud security or cloud security app. The name that's being bandied around is CNAPP.
That's what, and that's what cloud-native application protection platform.
Application protection.
This is different than the endpoint market. It's also different than the network market. There'll be players like Palo Alto that go across it, and there'll be players that are specialists in one of them. The landscape is still evolving.
Yes.
Those of you that follow security, probably know this well, but you really have to plot that out in order to understand the landscape for security.
Palo Alto's also approached it much more aggressively through acquisition.
They have.
You've done a lot more through kind of internal innovation.
Yeah.
Is that gonna be kind of the continued focus?
We made two acquisitions. We made a relatively large acquisition for us for SQreen, which was application security. So far what we've done is these are more acquihires.
I think we'll continue the way that we always have, which is we find good teams that can integrate in. Don't forget, almost everything we rebuild on our platform. We're not looking for point solutions that are gonna stand alone. If we can do it and accelerate it in an effective way through, we'll continue to do that, and we've been doing that. But if we can't, we're building it organically. It's likely to be a combo, but in the end of the day, it'll be one unit all knitted together.
Makes sense.
Yeah.
With that, David, thank you so much for joining us. We really appreciate it.
Thanks a lot. Thank you very much, everybody. Thanks. Good questions. Enjoyed it.