Good morning. My name is Gleb Budman. I'm co-founder and CEO of Backblaze. My co-founders and I actually started Backblaze in 2007. We almost entirely bootstrapped it to IPO, so we raised less than $3 million in outside funding, up until just before the IPO itself. Our fundamental mission is to make storing data and using data and protecting data astonishingly easy. So you've seen a few of these before. We're not going to make forward-looking statements. So if there are three things that you could walk away with from today, it would be this. One, we built a really best-of-breed storage cloud. It's a very powerful, durable, performance, and importantly, a very affordable storage cloud. It's about one-fifth the price of the traditional cloud providers. The second is that, in addition to the storage cloud, we're actually building a really good business. So it has strong growth and efficiency.
Our B2 Cloud Storage business grew 47% this last quarter, and we've also turned EBITDA positive. So strong best-of-breed storage cloud, strong business. Then the last item is where things are headed. Obviously, there's a lot of talk about AI, a lot of things happening at it, hard to know exactly which way it's going to go. But one of the things that I think we all realize is data's being created, and data workflows are being created that require AI. Backblaze is really well positioned to help companies do their AI workflows. We have nearly twice as many companies doing AI workflows on Backblaze today as we did one year ago. So if you can walk away with three things, those would be the three things. Now, here are the two cloud services that we offer. The first is our cloud backup service.
Very easy to use, unlimited backup for your laptops and desktops. Businesses and individuals use it. They love it because it's just easy to use. It's unlimited. They can sleep peacefully at night. This was our original business that we started in 2007. Then the newer business that we did is our B2 Cloud Storage business. B2 Cloud Storage is for all of the other storage needs that you may have, right? So every single app, every single website that you use, anything you interact with uses storage. Historically, that storage might have been servers in a data center. Today, much of that data is stored in the cloud. We are one of the cloud providers that data is stored on. So that is the B2 infrastructure as a service. Talking about the metrics of these two businesses, today they're about equal size.
So about half the companies are Computer Backup. About half the company is our B2 Cloud Storage. Both of these are growing businesses. The B2 side is the side that is the fastest growth, bigger market, bigger opportunity. So we're spending most of our focus and time on the B2 opportunity. But both of these are growing businesses, and we're spending time on both. Now, one of the things about these two businesses is that they share a common platform. That underlying storage cloud is a shared platform for both. And it's an incredibly efficient platform. This is a chart that we published in 2009 when we initially launched, when we showed just how much more efficient and affordable our storage is than what was publicly available from other choices at the time.
We originally actually intended to use Amazon Web Services as the underlying storage for our first product, our Computer Backup product. But we realized we'd lose money on every customer and ended up designing and building our own storage cloud that both the Computer Backup service and the B2 Cloud Storage service are now built on. So in addition to efficiency, right, in addition to building a very efficient storage cloud, we've also driven a lot of innovation. These are just some of the things that we've launched over the last six months. In Q4, we launched something called ShardStash. ShardStash makes us up to 30% faster than the traditional cloud providers like AWS, Azure, and Google Cloud. So not only are we about 1/5 the price point, but we're up to 30% faster. So we really are focused on building this best-of-breed storage cloud.
These are just some of the innovations that are more recent. Now, in addition to being a best-of-breed storage cloud, as I mentioned, we're building a good business. These are year-over-year numbers. You can see we've accelerated revenue growth. We've improved gross margins. We've dramatically improved our EBITDA from a negative 12% to a positive 6% and dramatically reduced cash usage. We are building not only a great technology and product, but we're building a really good business. Then in addition to the business, today, right, those were Q1 results. If you look at where the market is headed, one, we have a secular tailwind, which is that the traditional cloud providers are increasingly focused on serving the largest customers, the Fortune 500, the federal government. They're doing that because it takes billion-dollar contracts to move the needle for them.
By doing that, they're making their products increasingly complex, increasingly expensive, increasingly difficult to use. And they're leaving behind the mid-market. So there's a void in the mid-market, companies under 1,000 employees. And it's not to say that Amazon doesn't sell to anybody under 1,000, that nobody under 1,000 ever buys anything from Azure. Of course, they do. But their products and their efforts are increasingly not serving that market well. And so we are focused on that market. And while those companies are not the biggest, right, they're companies under 1,000 employees cumulatively that is about a $50 billion market just in infrastructure as a service, just cloud storage just in the mid-market. So if you can get to them efficiently, there's a lot of opportunity there. The other market opportunity is the shift from Cloud 1.0 to Cloud 2.0.
Cloud 1.0, it's comprised of several large traditional cloud providers that were building walled gardens, trying to offer 200 different services to you as a customer and saying, "You have to stay inside of our walled garden." For many years, that was fine. They didn't have a lot of other choices, and these walled gardens were providing them value. That's great. Over the last number of years, and this is a trend that's starting and accelerating, is that there are an increasing number of companies focused on being best-of-breed providers for certain aspects of the cloud, certain aspects of these companies' tech stacks. If you think about one of these walled gardens, they have 200 different services. One of the areas is networking. In networking, you now have Cloudflare, Fastly, Akamai, Bunny, and others who have built large businesses focused on just the networking piece.
It's one of the 200 different things that the traditional walled gardens do. But they're now best-of-breed providers offering great services and building big businesses around that. You can see the same thing happening in compute, in communications. The GPU clouds are now building up. Companies like CoreWeave and Applied Digital and others that are building up GPU cloud-specific best-of-breed providers. These walled gardens still exist, but there are big market opportunities for best-of-breed providers. No one has done that with data storage yet. We believe data storage is actually probably the most important one of these because in order for all of this to succeed, you actually need the data to flow between these different providers. That is what Backblaze is doing. Backblaze is building that foundational data storage platform for Cloud 2.0. So how do we do that?
The first is we build a really robust data platform, which we've done, right? We've got over 3 exabytes. That's gigabyte, terabyte, petabyte, exabyte, in factors of 1,000. So one of the largest storage clouds out there. It's durable. It's performant. It's available. It's really affordable. The second is we provide free egress. So we allow your data to move back and forth between these providers without cost. The third is that we build this ecosystem of partnerships. So all these companies, the Cloudflares, the Fastly, the CoreWeaves, et cetera, we want to support customers using them. Whereas the traditional walled gardens want to say, "No, just use our services. Don't use any of those providers." And the last one is the innovation for the open cloud. So we are innovating to support this transition to Cloud 2.0 instead of to resist it.
One of the latest things that we've launched is Event Notifications. Event Notifications allow customers to automate the workflows to use these different providers. This is how we are positioning ourselves to be that foundational data platform for Cloud 2.0, which is a path that's happening, right? The mere existence of these large organizations like Cloudflare and Fastly and others and Twilio and Stripe and others is proof that this is a transition that's happening. We also did a survey to 400 IT decision makers a few months ago and asked the question, "Would you prefer to buy from an all-in-one, or would you prefer to buy best-of-breed?" More than half said they prefer to buy best-of-breed. Now, why is that big?
It's big because today, something like 90% of people are in the walled gardens, which says that the direction that customers want, more than half of the customers that are currently inside of these walled gardens, hundreds and hundreds of billions of dollars' worth, want to buy from best-of-breed providers. So that is a trend that we see happening and something that we believe we're well positioned to support. The other trend is obviously all of the investments in AI. The thing about AI is nobody knows exactly where it's going to go, right? Which companies are going to succeed with AI? What are they going to do? How is it all going to play out? But what we all know is that data's getting created, and the data workflows are happening and being built around these.
And so what we increasingly see is customers using Backblaze for this AI workflow. They have data that comes from users, from cameras, from systems. That data lands inside of Backblaze. They then use our free egress and our connectivity with our best-of-breed provider partners to send the data for analysis. It might be to CoreWeave. It might be to Applied Digital. It might be to some other provider who does analysis. It might even be to an AWS, Google, or Azure for you to use their GPU providers. But they send the data out because we partner, we make it free, we make it easy. They put the analysis results back in Backblaze. Then they send it out to the customers, other servers, other analysis systems. And that workflow is something we're increasingly seeing more frequently.
In the last year, like I said, we've had twice as many companies now using Backblaze for AI workflows as we had up to that point. So here's an example of one of these companies. They do motion capture to create models. So I think we've all seen the people sticking the green balls onto people and filming them to create their models, right, to do the motion capture. Then they can create superheroes and video games and things like that. Well, this company figured out how to do that without sticking all the green balls on, but by using AI on the video. So they have cameras. The video footage goes into Backblaze. They analyze it inside of one of these GPU clouds. The data comes back to us, and then it flows out to their customers.
So this foundational kind of underlying workflow is something that we believe more and more and more companies are going to be using with their AI processes and something that we are well set up for. So the other part of it is that things that we are doing now are twofold. One is we're investing in go-to-market. So we're scaling up the more going up into the upper mid-market with an expanded sales team. We're bringing on a new sales leader. And we're also scaling up our channel effort. So we're layering the different go-to-market motions to accelerate our ability to go after this market. And the other thing, like I talked about, is the innovation. So we've accelerated innovation. We've shipped a number of different things. Some of them are just things to add value for the customer.
Some are things that we charge money for, like Computer Backup Enterprise Control is an extra charge to customers that enables us also to move up market with bigger companies. And some are things like Powered By, which is the first one, which is a new go-to-market path. It's an OEM white-label type approach where companies can embed Backblaze into their offerings. So we're driving and layering go-to-market motions, and we're driving more innovation. So like I said, if you walk away with three things, we're building a best-of-breed storage cloud, which is ideal for the Cloud 2.0 world, which is a direction that's happening. We're building a strong business, which has both growth and efficiency. And we're positioned well to take advantage of the directions that AI is headed because we're building the underlying layer, the picks and shovels for AI. Thank you very much.
We'll just transition now to a fireside. Thanks, Gleb. If anyone has questions, please feel free. Yeah, go right ahead. Awesome.
Excellent presentation. Thank you. Cybersecurity is top of mind for everybody. So when people are looking at making a purchase decision, where I'm going to store it, how do they get comforted? You know, when they think, "Oh, I'm going to put it in a drawer," or, you know, they got to be safe. You know, they want to cover their ass no matter what. Even if there's a significant financial incentive to save money, they don't want it to come back to ponder. How do you give people comfort that their data is not, by virtue of being Backblazed, more vulnerable for it to be hacked?
Yeah, it's a good question. So cybersecurity is obviously on everybody's mind. You see it in the papers every single day. Every company, there's ransomware and hacked and everything and breaches are over. So I think in terms of how do we get people comfortable with it, I would say a couple of things. One is we've been doing this for 17 years. So we have a long, long history. In fact, longer than Azure and longer than Google in terms of providing this. So we have a long history of providing a secure platform for our customers. We also have about 500,000 customers in 175 countries around the world. And we have over 3 exabytes of customer data.
So we have a lot of history over time, over locations, over customer types, over customer numbers, and over data volumes to say, "We've been doing this and keeping the data secure." The other is that we do a lot to make sure that the data is secure. We have a Chief Information Security Officer who helped both government agencies and Snowflake create their security programs. He has a team of people that work with him and others throughout the organization. We also pay a third-party company to attempt to do penetration testing into our systems. We pay another company to run bug bounty programs to try to get white-hat hackers to hack our systems. We take security seriously. We do a lot to make sure the systems are secure. We also offer features and functionality for the customers.
So with our Computer Backup service, we encrypt 100% of the data on your machine before it ever leaves it. We send the data over an encrypted connection. We store the data encrypted. With our B2 Cloud Storage offering, we send all data over an encrypted connection, and we let the customer choose if they want to also keep the data encrypted on our systems and whether they want us to manage those keys or whether they would like to actually manage the keys themselves. So we do a very comprehensive end-to-end approach for cybersecurity.
One other question. Do you have partnerships? Is that a two-way street with partnerships? Are they sending business your way, or you're sending business their way? How does that work?
Yeah, so partnerships are important for us. We have different kinds. So we have technology partnerships, and we have channel partnerships. The channel partnership motion is newer. So historically, when we went public, 80% of all of the revenue came self-serve. So that meant literally no human or channel or anybody ever talked to those people. They just showed up, entered a credit card, and signed up. The 20% that was sales-assisted were customers that are about 20 times bigger. And so one of the things we said at IPO is that we want to invest behind that sales motion. So we've stood up a sales team. I talked about getting additional people in leadership. In the last year and a half or so, we've been investing behind a channel motion. The channel partnerships, we have been historically sending them leads and then working with them to close it.
Over time then, we were sending them leads, but they were actually being valuable in helping close those opportunities. And then more recently, they've actually been sending us leads. So the channel motion is still on the newer side, but we are starting to get business coming from them. And we highlighted some deals that were part of the channel deal. We had a $1 million upfront multi-year deal that we talked about before. That was a deal that was done with the channel together. That's on the channel side. On the technology partnership side, it's also both. So for example, we partner with Fastly. And Fastly is a CDN provider. Every single customer that Fastly has has to have data somewhere. If they keep that data with Amazon, Amazon makes it very expensive for the customer to send their data to Fastly.
They are actively marketing to stop using Fastly in favor of Amazon's competing product. So for Fastly, it's a very bad idea for their customers to be using Amazon or Azure or Google for all the same reasons. With Backblaze, not only do we make it free to send the data to Fastly, but we work well in partnering with them. So we send them opportunities. They send us opportunities. And it's the same way across many of our technology partnerships.
Any other questions? And if I could just come back to the cybersecurity point, I just think, and let me know if I'm talking out of turn here, but one of the things to consider, a lot of people think, like, the on-prem data center for your storage is, like, the most secure way to go. But there's an issue there because those appliances or those boxes have, like, a set amount of time, right? You might be retaining data on that box for two weeks. And so if you have a 14-day window where you're retaining data, the cybercriminals are getting smarter, right? And so they will sit dormant for 15 days. And then when they take down your network and crash your storage, when you go to call up that backed up data, you can't access it because they've been there. They know.
Your 14 days are wiped because they've been there the whole time. So I think that probably goes to show that cloud storage, well, there's probably then we could probably touch on that as far as how much you still need to maybe evangelize this market to go away from on-prem towards cloud. But I feel like that's one way that they continue to take down some of those barriers to adopt cloud storage. And I hope that helps as well.
So I'd like you to focus a little bit on the economics of the sales organization side of it. Not necessarily the KPIs, but how many sales personnel do you have on the B2 Cloud? How long does it take to onboard them, make their first sales, and be quoted? What type of quotas do they have? Where do the best recruits come from? Are you recruiting them from some of your competitors, or can you just onboard a few BDRs, and it's a very simple sell based on price or something like that?
Yeah, so the go-to-market motion around the sales. So, you know, like I said, when we went public, 80% of the business was self-serve. And so the 20% that was sales was very, very nascent. So the sales team at that point, and this was 2.5 years ago, the sales team at that point couldn't actually sell anything because they had no SKUs. They had no contracts. There were no Net 30. There were no POs. There were limited systems. There was nothing they could actually sell. The only thing they could do at that point was they could take a customer that raised their hand and said, "I'm interested in talking to somebody," answer some of their questions, and then guide them toward the self-serve path and say, "Here's where you enter a credit card." So at that point, it was very nascent.
Over the last 2.5 years, we've been building that function out. And what I will say is we have seen success in it, but I definitely think it's still an area of opportunity. And it's part of why we're bringing in sales leadership for that function. I think it's an area of opportunity to continue to optimize what that team does. So over the last 2.5 years, we've brought in a team. We now have an SDR team that does outbound emails and calls, a BDR team that takes inbound leads and does that early first path of qualification, an account executive team, including a split for ones that focus on larger deals, ones that focus on smaller deals, a customer success team after they've been onboarded, a solutions engineering team that helps them from a technical perspective.
We also have the partnership team inside of the sales organization. And so those are the different groups. We've also built out systems, Salesforce, Gong, you know, other tools to help the team do, split it up into regions, and basically kind of put into place the ability to do Net30 and contracts and all. So we've put in a lot of the infrastructure. I will say we have seen success with the team in terms of, you know, things that we've publicly talked about is we now have over 70 customers that are paying us over $50,000 each. We have a customer that paid us $1 million upfront on a multi-year contract and another customer that's paying us $1 million per year.
Now, compared to where we were at IPO, which was that the company's ARPU was $350, you know, every one of these deals is orders of magnitude larger than that. And so the sales team and the sales motion are, I think, working, but there's still a tremendous amount of optimization in exactly the kinds of things you're talking about, right? So I think that there's optimization in terms of ramp speed and optimization in terms of efficiency. There's a lot of places where we can get more leverage on that. We just brought on an account executive group, not large, but it's a team that we hired up in Q1. And that group is focused on that upper mid-market. So historically, again, you know, the ARPU was $350, but now we're having customers paying us $50,000, $100,000, $200,000. And so they're focused on those kinds of deals.
So we don't, you know, we don't share all the details publicly of, you know, the ramp speed, the pipelines, the conversion rates, the commission plans, and everything else. But what I will say is we've come a tremendously long way from where we were at IPO. And I think there's a tremendous amount of opportunity for a new sales leader to drive more output out of the current scheme and process.
A follow-on. I know you have a very big presence on social media with blog. Can you just give me some idea of, like, how big that is, how many monthly users, how big your marketing team is? And are any of the, you know, Are you getting leads from that transitioning into the B2 Cloud?
Yeah, so up until IPO, because we had raised less than $3 million into the company, the go-to-market motion was we were very, very focused on efficiency. So one of the most efficient channels, probably the most efficient channel that we found throughout those years, was content marketing. We published a blog. We layered around that blog efforts around PR, social media, and engagement with the community. The blog itself has 2 million people a year that read it. This is a storage-focused blog, right? So we publish things like hard drive reliability statistics, ways that you can networking traffic patterns and what we've learned from them. You know, it's a blog. It's not a Hollywood blog about, you know, the latest pop star, right? So it's a technology-focused, storage-focused blog. 2 million people a year read it.
From there, obviously, not every single one of them is a potential customer. But because it is targeted towards storage enthusiasts, it is a brand-building approach to the company. And a fair amount of the inbound leads that we get, people mention, "Oh, and I've seen your blog," or, "And someone told me about your blog," or, "And I love your drive stats," or something of that sort. So it is part of the actual driving factor into B2. One of the other interesting things for us is we started the blog originally when we only had Computer Backup as a business. The linkage between the blog that we had and a Computer Backup customer was a little further afield, right? Because we were talking about things like how do you build storage and things like that and a Computer Backup customer that's just trying to use the service.
The linkage between the blog and the B2 service are actually closer because the people using B2 are more likely to be IT people, application developers, and others. And so I think we actually have a closer linkage between the content that we're producing and that. You talked about the team. So originally, for years, obviously.
You're retargeting about the same people?
Yeah, so we've actually started doing some of the retargeting. Historically, it was mostly around just put out great content, talk to the media about it, get awareness around it. You know, we publish these quarterly hard drive reliability statistics. You know, hundreds of thousands of people often read those individual blog posts. It often ends up in the media. It ends up in sites like Reddit and Hacker News and others. But we've put on some retargeting pixels and are starting to do some retargeting with it. And I've seen some success, but that's early, early for us. What was the other thing? Oh, in terms of the team, so one of the things that we did was we actually decided that this is a function.
We actually hired a managing editor for the blog away from a publishing house, and then hired some content folks within that organization that work across our organization to engage with our engineers, to engage with our cloud operations people, and others to bring interesting topics that we can tee up onto the blog.
Great questions. Any others out there? OK. I know we were talking about the north of 70 customers now that are above $50,000 in spend, right? So one of the things and I wonder if you get this first, do you get this as a question from investors? And then secondly, have you seen this in the business? But is there a concern, as those customers get larger, let's say $100,000, $200,000 spend, are they starting to entertain maybe the "graduating" to those hyperscalers? I know we're talking about the significant price differential that you have. But is there a certain level of scale where that becomes a concern as far as moving on market or to one of those hyperscalers? Excuse me.
So I'd say we haven't seen it. Mostly, we're getting customers from the traditional cloud providers. And part of that is because, in some ways, if you think about it, if you're a customer that spends $100 at a traditional cloud provider, yes, we're a fifth the price. But that means you're going to save $80 bucks. If you're a customer that's spending $1 million at a traditional cloud provider, you have the opportunity to save hundreds of thousands of dollars. So, you know, one of the other companies that we talked about publicly, I think, on the last quarter, they were on one of the traditional cloud providers. They switched to us. They're now saving $800,000 a year by having switched to us. The customer that's paying us $1 million a year was previously on a traditional cloud provider and switched to us. Right?
I think sometimes we lose customers because they've been acquired by a bigger company. So one of those large organizations has an existing infrastructure and existing relationship and everything. And they say, well, you have to use ours, right? So we've sometimes lost customers because of that. But I don't know that I've seen any customers that have simply scaled up and then decided to leave. It's usually we're getting customers who they start with free credits on one of the traditional cloud providers. Then once those free credits start expiring, they start realizing how much it's going to cost them to stay on it. Then they actually graduate to us.
Got it. Got it. Let's take a step back. I think a lot of people probably look at the size of the hyperscalers. Let's take AWS as an example, not billions of dollars. There's almost like a head-scratch moment, like, wait, how are they doing this at such a significantly cheaper price for customers, right? And maybe you can just elaborate, like, what is the differential that you guys have there, whether it's tech to architecture, that allows you to provide that value to customers at such a significant delta?
Yeah, I mean, it's obviously it's a thing that people go a lot of times like, wait, how is that possible? And I'll say there's a couple things. One is we're focused on mid-market. So because we're focused on mid-market, we don't need all the complexity that they end up putting in for the Fortune 500 and the federal government that adds costs to their infrastructure. The other is because we focus on storage, we get to just be best of breed at that, which both means for 17 years, we have focused on this singular problem. They have a lot of people, a lot of money, but they also are trying to do 200 different things. And it's hard to be best of breed in all 200 things. So, you know, I think there's a couple things in that way.
So one is we're focusing on a slightly different market, which doesn't need things that are adding costs to their infrastructure. We've focused on this problem singularly for all of these years. I think we've innovated in various ways. So, you know, I give the example of ShardStash, which is how we've innovated on the performance side. Because one of the things that people talk about is, oh, if it was cheap, then it must be slow. And we're like, no, but it's cheap and fast. And the example of that is all of the traditional cloud providers do this double hop with a caching layer where they have SSDs. The data goes to the SSDs. Then it hops from there to the hard drives. And that requires a double hop, which makes it slower. And it requires a really big, expensive SSD layer in front.
What we've done with ShardStash is our team realized that it was the actual acknowledgment from the hard drive that is the part that is the slowest. They have the data get written to both the SSD and the hard drive at the same time in parallel. That way, there's only a single hop, not a double hop, so it's faster. It only needs to live on the SSD until the acknowledgment from the hard drive. It's also a smaller SSD layer. It's both cheaper and faster than the traditional way that everybody else has done it. Our team has been innovating on the actual technology platform.
Understood. And I think that's a great example of the R&D innovation you guys are driving at as well. Maybe from a competition standpoint, how many of these customers that are coming to you, is it their first time using cloud versus, let's say, defections from those hyperscalers to Backblaze? Like, how does that play out for you?
So I'm not sure I can fully quantify how much come from which. But we do see a lot of both. So we see a lot of customers leaving the traditional clouds and coming to us. I think I've shared on almost every single earnings call, I've shared examples of customers that have done that. I think in the Q4, when we went through and talked about a customer that switched from AWS, a customer that switched from Azure, a customer that switched from Google, and kind of did the little mini case studies on each of those. So that's a common path. We also do see a lot of customers that have on-premise equipment and have been looking at cloud for a long time and saying, you know, I would love to get the benefits of cloud. I'd love to have the scalability of it.
I'd love to not have to deal with having my own on-premise. But they've looked at AWS and said, it's too complicated. And it's too expensive. And so I'm just going to stick with my on-prem stuff that I already know how to do. And for many of those, we are their first cloud. And so a lot of those use cases are backup and archive and sync-type use cases, media workflow use cases. In media production, a lot of the organizations used to have their media teams co-located together. And during COVID, they all had to go home. And they couldn't access all of these high-performance systems that were all in one room together. And so they had to figure out, how do they work now? And so the media workflows are now moving to the cloud.
For many of them, we are their first cloud because it's how they're able to do their workflow.
Got it. And we probably have time for one more. I know I have one on my side, but just want to open it up to the audience again in case you guys are sitting on any questions. OK. Last question from my side was going to be around the AI opportunity. And where I'm going with this is just based on there's been, like, this awareness out there in the market that for me to appropriately leverage AI, I need to have the data. I need to have it stored. I need to have it, in some way, maybe even normalize the data types, the data volumes. Are you seeing organizations rethink their data strategy in the context of AI? Maybe they need to either store a greater volume or different data types. But I'd just be curious what your view is on that.
Yeah, I think I would say the first part of it is I think all the organizations are trying to figure it out. And many of them are in that early stage of, I'm not exactly sure where this is going and what to do. They just see that there's potential opportunity. I think one of the things that they are also noticing, some of the ones that are kind of a little more leaning into the AI opportunity, is that because it's moving so fast, the services they want to use are not necessarily in one place, right? So one quarter, AWS has an announcement. They've got some great new thing in AI. The next quarter, Google does. The quarter after that, Azure does. And the quarter after that, it's CoreWeave and GMI Cloud and Applied Digital and everything else.
They're looking and saying, you know, whereas in the past, being inside of a walled garden felt fine because I need some compute. I need some network. I need some database. It's fine. This stuff is just kind of there. But now, if I'm inside of AWS and Google launches the thing that I want to use, now I'm stuck. If I'm inside of Azure and Amazon launches the thing I want to use, I'm stuck. They're starting to realize that they actually want to use a lot of the innovation that's happening outside of those traditional cloud providers. They want to be able to use their data with other places. That's some of what we're seeing with the data workflows.
So long story short, just given how nascent that opportunity is, the stack feels like it's fragmented, at the very least. Backblaze does give the customers at least that flexibility to plug and play and evolve as the market evolves.
Exactly.
Awesome. Awesome. All right. Thank you very much.