Thank you, everybody. Thank you, Sid. It's so good to see you. It's been a while.
Yeah, thanks for having me.
Absolutely. You look, you look great. You sound good. This is the third day of the conference. I think the audience has heard me say this so many times: the third day is the best day. Day three is the best. We end with a bang, and you look at the quality of companies. Don't tell that to the company management that's present—that presented on day one and day two, this is the best lineup. Okay. So we're really happy to have Sid Sijbrandij, founder and CEO of GitLab, join us this afternoon. I believe this is your first Goldman Sachs conference.
It is.
Yeah. Last year we had some schedule conflict, and this is the first opportunity to host you. So welcome, a big welcome to the Goldman Sachs Communications and Technology Conference.
Yeah, thanks for having me.
Absolutely. Absolutely. So, since we've not done this before, what are your goals for the company in the next five years? What do you want the company to look like? What are your aspirations, your biggest dreams, your audacious goals?
Yeah. We see a future where we're gonna be best in class in every part of the Dev cycle, SDLC. So being able to replace every point solution that customers are currently using-
Mm-hmm.
with something better, but also be able to host the data and the models that are essential to tomorrow's applications.
Mm-hmm.
We wanna be a single software platform for all of R&D.
And what is that? So the point solutions that are not doing a good job today, how do you visualize the friction-free in this perfect world, where somebody's on GitLab, what does that process look—what does their day look like relative to the day-to-day?
So if you weren't using GitLab, you'd be hopping from Jira to GitHub, to Jenkins, to Checkmarx, to Honeycomb. You'd be hopping between all these different applications.
Mm-hmm.
We see that if people move to GitLab, they get a 7x improvement in cycle time.
Mm-hmm.
They're able to go through the process seven times faster-
Mm-hmm
which is an amazing win.
Mm-hmm.
That's due to a single interface, a single data store-
Mm-hmm
really streamlining that process.
Mm-hmm. And has that value proposition continued to stay resonant in the face of the competition, also trying to maybe copy what GitLab made a name for itself with? Or how, how much is the difference between that?
Yeah, I think we have a big head start, and we have the most comprehensive platform. If you look at security, only GitLab has dynamic application security testing, container scanning, API security, fuzz testing. Look at operations, only GitLab has feature flag, incident management-
Mm-hmm
... Service Desk functionality.
Mm-hmm.
We're able to give better data-
Mm-hmm
over the, on the overall process.
Mm-hmm.
We have-
Mm-hmm
Portfolio management, OKR management, DORA metrics, value stream management, and that's what helps companies go faster through the entire process.
Mm-hmm. Mm. And if this vision comes true, what... How, how do we think about the savings, process-wise, 7, 7 times faster? Completely get it. As, as you become more institutionalized, how the companies that have adopted GitLab at scale, that have thousands and thousands of people, what, what kind of savings do they get by deploying them?
Yeah. A Forrester study shows that they have a 427% return, return on investment over 3 years, and that's from being able to consolidate spend-
Mm-hmm
... not having the integration cost between all those different apps.
Mm-hmm.
There are people being able to do more in their day-
Mm-hmm
and reducing that cycle time.
Got it. You should charge more if it's more than 27% return on investment.
You're gonna ask a question about our pricing future. I anticipate.
Yeah, Gili will definitely get into pricing, but I'm curious, the big return on investment and the... I know you were early on to flag the macroeconomic concerns, but just to get that part of the story out, I mean, there's a secular aspect of it, the cyclical aspect. The cyclical aspect, everybody is going through what they're going through. Where are we? Are we done with the bottoming of all the metrics, the net expansion rate coming under pressure because of attrition, lack of expansion within your customer base? Where are we in that cycle?
Yeah, we've announced earnings on Tuesday-
Yep
... and we said we see a stabilization. Q2 was very similar to Q1.
Mm-hmm.
All of these metrics are now stabilizing.
Mm-hmm.
The gross retention number for GitLab, it's a trailing metric over 12 months, but we're getting to the point, next quarter will be 1 year in, and then-
Yep
That should stabilize as well.
Got it. So as you, I know that the January quarter was a very odd one because the first couple of months were okay, and then third month you saw a big decline in net expansion rate. Then when you guys reported—I mean, I don't mean to get too much into the numbers. I know you're the technologist, and I wanna hear more about the product, but in the interest of understanding the trajectory, is it right to say that the net expansion rate, unlike the Q4, where it got worse as the quarter ended, did it get better as this quarter ended?
Yeah, we haven't seen that either. I think since that January, there's been a new normal.
Yeah.
It used to be aspirational buying.
Yeah.
We think we're gonna do this, we're gonna buy all the licenses in one big deal.
Yeah.
Now it's really, companies are taking it as it goes-
Mm-hmm
-because the future's unpredictable.
Yeah.
That's the new normal, just buy what we need.
Okay.
And that's the trend is now the new normal for us, too.
Got it. Got it. Got it. People sort of overbought, got overenthusiastic, it sounds like.
I think because companies were growing-
Yeah
... it made sense to buy the extra licenses upfront and negotiate-
Yeah.
rather than being forced
Yeah
in the year to renegotiate and not have a good-
Yeah
negotiation position. But as companies have toned down their hiring, that dynamic has shifted.
Yeah. Your net expansion rate is still very solid, to be sure, right? It used to be very high, 143 or so, and it's down to 128?
Four.
Twenty-four.
Absolutely.
Still very good. I mean, we know companies whose net expansion rates have gone to 100 or maybe below 100, so it's still good. So which means that your end market, these software developers, it's just they're not being hired at a fast enough rate. It's not that they're being laid off or, or having a-- It's not like selling to salespeople, and there's fewer salespeople in the tech industry than a year back. So that's not the dynamic we're talking about here at all.
Yeah, I-
Much healthier.
Every company still needs to become a software company-
Yeah
and they still need the people to do that.
Yeah.
It's just that the very fast hiring-
Yeah
That's passed us now.
Yeah. Yeah. Excellent. Excellent. So as you talk to customers, and I've asked this question of every other CEO and so we'll have gone through probably, you know, 21, 22 CEOs and five CFOs, and I'll share with you what, what I've heard so far. But as you talk to customers, what are they saying about calendar 2024?
I think everyone's cautious.
Yeah.
We're looking forward to the next year. We'll see some of the price increase effects kick in. We've got a new product called GitLab Dedicated-
Mm-hmm
... which is a single-tenant SaaS offering.
Mm-hmm.
That's super interesting for customers that want to have their own thing on their own private cloud-
Mm-hmm
but still want us to run the product.
Mm-hmm.
We'll see the effects of AI kicking more. We have 10 features now available to customers, but we'll be, we'll be selling those.
Mm-hmm.
We're looking forward toward next year, but if I could predict the macroeconomy, I probably wouldn't be a software CEO.
I have some news for you. We have a macroeconomist who has actually been quite good, Jan Hatzius. He has been calling for a soft landing for quite some time. I think it looks like we are having a soft landing, and I'm sure you guys have heard this many times before. We have co-opted his term. We're calling it a software landing. It's a GS proprietary thing. It's software landing. We've got to start using that, software landing. His reasoning goes like this: So we've been through the worst of rate increases, inflation relatively under control, so maybe a couple more rate increases. Even if that happens, it should not really jolt the economy.
You're seeing the labor market coming to the right level of balance, and the GDP is still 2%, and the supply chains are nursing back to health overall. So if you kind of play the scenario out, the past 18 months, we've been through one of the worst times possibly that could be inflicted upon us, and we still are alive. And your net expansion rate is 124, when rates run from 0 to 525 basis points, that's like a pretty damn good achieve-- I know that it's down, but I think we should somewhat be relieved that it was not that bad. It was bad. It was not that bad.
But then, then the next 12 months, next 15 months, we see some stability in the economic environment and that everybody's not running for covers and saying, "You know, we're gonna have a recession." We've been behaving like it's gonna be a recession for the past 2 years, really. But if you take that off, what does GitLab look like? What are the things that, that, that you see that could come unblocked from the customer spending perspective?
Well, the biggest thing that happens for us is the recognition of our category. In this quarter, Gartner and Forrester released a quadrant for our category, DevOps platforms.
Mm-hmm.
Gartner called us out as one of the leaders, with rating us the absolute highest on the ability to execute.
Mm-hmm.
In the Forrester one, we were the only leader.
Mm-hmm.
We've been trying for years to get a category created. It's created now, and we see this market moving from point solutions to this platform. This is a $40 billion market.
Wow, okay.
The macro can do a lot of things, and we'll certainly go up and down with it.
Yeah, yeah.
But this secular trend of going from point solutions to platforms, that's what we're really excited about.
Got it. Got it. So you cannot grow unless you hire, but are you thinking about stepping up the pace of hiring? And if you are, what are the things that you need to see from the standpoint of demand signals and KPIs, forward-looking indicators, that you can say, "You know what? I'm gonna start hiring"?
Yeah, we wanna grow, but do it responsibly.
Mm-hmm.
We've shared the expectation that next year will be cash flow breakeven. This quarter, we grew 38%.
Mm-hmm.
Year-over-year, we had a 23% margin improvement on our non-GAAP metrics.
Mm-hmm.
So, really being able to get additional operational efficiency. That being said, we're hiring, especially in sales-
Mm-hmm
-and R&D.
Okay. Are you stepping up the rate or about the same as the previous quarter? Is there a change in your hiring, is what I'm...
I think it's relatively stable. I think in R&D, where it's growing a bit faster.
Okay. Okay, great. So, when we look at the competitive environment, there is a quick tendency to conclude that GitHub Copilot has had tremendous impact, which clearly has. How do you look at that code generation single trick, which is an effective trick, does that change the competitive landscape for you guys? Have you seen any changes at all in evaluation, et cetera? I wanna hear the answer to that, and then I have another question about your own code generation feature, which we've been expecting for quite some time.
Cool. Yeah, we'll talk about that next. I think code generation is important.
Mm-hmm.
We'll talk about our feature there first. But I think to give some context, we just released the State of DevOps Report on Tuesday, and often people in the DevSecOps process, about half are developers, but they spend only 25% of their time coding.
Mm-hmm.
If you make that more effective, that's helpful.
Mm-hmm.
But it's also super helpful to help them with the other 75% of their time.
Mm-hmm.
You've gotta make the security and operations people also more effective-
Mm-hmm.
Otherwise, you're gonna get bottlenecks .
Right. Yeah.
It's not just generating more code. For example, we have a feature, an AI feature called Suggested Reviewers.
Mm-hmm.
It suggests the best reviewer for your code, because most of the time that it takes to get something out to market, it's not time spent coding behind the keyboard, but waiting to find the right person to review your code.
Mm-hmm. Mm-hmm.
We now have 10 AI features-
Mm-hmm
... usable by customers today, 3 times more than the competition.
Mm-hmm.
They help throughout that life cycle, and that's super important to get the overall efficiency up.
Mm-hmm. And the code generation, is something that Google's been working on, its coding. Is that something that you could find partnership opportunities with?
Yeah. We're partnering with Google on that.
Tell us more about that.
Yeah. We're using their coding model-
Yeah
... on Google Vertex.
Mm-hmm.
So that's been a great partnership.
Mm-hmm.
Google Next was just... You had Thomas Kurian here.
That's right. Yeah, just before, yeah.
Great, great partnership. They named us Partner of the Year again-
Congrats
... for the third time in a row. And they also, we also jointly announced that they'll integrate GitLab into their Google Cloud Console, so super excited about that.
When is that coming out, the integration?
We haven't given a timeline yet.
Okay. Well, we're all waiting with like bated breath, because that's important to restore some balance of forces in the market. You got one big company, Microsoft, that is getting a lot of attention, which deservedly so, but then this is the partnership that the other side of it has not been seen yet, the competitive response.
Yeah, I think as Microsoft-
How exciting could it be, Sid, with Google helping you guys with?
I think we're super excited about our partnerships, and it's not Google. For example, AWS just released an integration for GitLab with their CodePipeline product.
Okay.
So I think as GitHub has more and more Azure-only features.
Mm-hmm
... the other hyperscalers are getting closer to us, which is great.
Mm-hmm. Mm-hmm. So it's not like there is pent-up demand for code generations; that is only 25% of what the developer does, right?
Yeah, and it's an important part.
Yeah.
It's something where AI can help tremendously.
Yeah.
We're super excited. We have Code Suggestions out in beta, and-
Yeah
... we'll make it generally available later this year.
Mm-hmm. Tell us about the other AI features that you have that the competition does not have, and what is it that Wall Street needs to know, and people like me and Gili need to know, that there's, this is the AI story of GitLab that has not been told. What is that AI story?
Yeah. So there's 10. I'm not gonna go into all of them.
No. Yeah.
We talked about Code Suggestions. We talked about suggested reviewers.
Yeah.
Another one is Explain This Vulnerability, because the software is really good at saying, "Oh, these are the vulnerabilities we found," but then what's the next step?
Yeah.
AI can help you there, give you context-
Mm-hmm
... and see whether it applies to you.
Mm-hmm. And, the price... So Gili's gonna talk to you about the price increase, so I mean, punt to Gili now. So-
Yeah, actually-
That's a good segment.
... before we kind of get to price increases, first of all, thanks, Sid, for taking the time and being with us today. I want to actually ask you around generative AI, if you see a particular stage in the software development life cycle that you think that this will have outsized opportunity in, and, you know, GitLab definitely caters and tries to add that value in the, at lo-- and the, like, further part of the software development life cycle. Is that where you're going to see that differentiated solution?
I think we're in a unique position because we have a more comprehensive product. We have a single data store, and at GitLab, we can do more, and I think that the quantity of AI features is an example of that. Another AI feature, for example, is to summarize an issue. We just announced during earnings that at a large multinational bank, you move thousands of users from Jira to GitLab in order to have them as part of that life cycle, and there, too, AI can make a major impact.
Yeah, and so that does bring us to pricing. And you guys released GitLab 16, you guys released GitLab Duo, and you also recently did a price increase. Can you tell us a little bit about your overall strategy for pricing, and then also how that can manifest itself with a lot of the new generative AI tools?
Yeah. In April, we did a price increase for GitLab Premium from $19 to $29, and it's been five years since we changed the pricing. In that time, more than 400 new features came out-
Mm-hmm
... that add a lot of value and allowed our customers to replace more point solutions.
Mm-hmm.
So, we wanna make sure we always generate more value than we capture. But it was the time to raise the price, and that's gonna have an effect over a long time because contracts gotta come up for renewal.
Right
... typically 12 months or more. Then there's the ratable nature of our revenue, and there's legacy pricing for the existing customers.
... And when you, in your last earnings call, you also talked about compliance and security, and how those are becoming greater parts of the story, and what people are looking for when it comes to GitLab. So how are you, you know, as more people want more features, and you really are releasing them, should we be thinking of it as a factor of how many releases GitLab has? Or what's the overall your mindset when it comes to maybe leveraging pricing or maybe putting a certain feature behind a different paywall, whether it be premium, ultimate or something else?
Yeah. So most of the paid security features are in GitLab Ultimate, our product for $99 per user per month. And, if you look at what we offer, we offer the most comprehensive security offering in the market. We're able to replace products like Snyk, Checkmarx, Synopsys, Black Duck with it.
Yep.
It goes all the way from static and dynamic scanning, container scanning, API, fuzz testing, secret detection. It's, it's a superior product to everything on the market, but it also allows us to have a, a compliance functionality that is unlike anything else. I think the only alternative for GitLab would be to build the compliance yourself, but with GitLab, it's easy to prove, like, on every environment, with every project, with every change to that project, I've met all these objectives. I've done all the scans. I've reviewed all the reports that came out. I've every piece of code was reviewed by two people. It's really hard to do that. It's really hard to make that yourself, and GitLab is the only product that offers it out of the box.
Understood. Understood. You know, I'm drawing... Through this conversation, I'm actually drawing a parallel to another company that we had up here, and they mentioned how they're seeing a mind shift change within customers and the conversations that they're having, from open source to managed services such as what GitLab can offer. And I know you guys had to remove that, a seat capacity or, like, the unlimited free version in that sense, and that's been driving a lot of that conversion, and you're expecting that to be a bigger growth driver next year. Are you-- do you agree with that overall mindset within the industry and what you're seeing there?
You know, what we did is we on GitLab.com, where we host it for you, we instituted a seat limit of five.
Mm-hmm.
If you have more users, you need to switch to a paid plan, and that's to make our business sustainable. We were paying for the hosting costs-
Yeah
... and we want to make sure that that gets paid. Our open source version of GitLab still has unlimited users, but then the users, the companies themselves pay for the hosting.
Understood. And when we think about, you just mentioned cost, and my mind went back to Generative AI around, you know, how to think through the cost associated with that and the computing cost. And I know that you mentioned your partnerships with both Google and AWS, as well as others. And how is that helping you with the, not just go-to-market, but also with some of that training and operation, and inferencing costs that are related to those types of services?
You know, we wanna leverage our partnerships with the hyperscalers to do that. We don't want to become experts in building large language models. There should probably be only a few companies in the world doing that. So it's great that you have things like, Google Vertex, Amazon Bedrock, in order to kind of do that for you, where you can focus on maybe fine-tuning, but for sure, prompt engineering and integrating it into your interface.
Yeah, that's very helpful.
So I said, sorry to make you move back and forth, but, I am curious to get your views, if you have one, on Vertex AI.
I think it's great. We're big users of it.
Really?
I think our promise to customers is that their privacy and their IP protection is paramount. We promise we'll never use their data to train other people, and Google Vertex helps us deliver on that, where it stays within our private cloud.
Okay. And, have you had a look at Gemini, the LLM that are gonna be coming out with?
It's, it's very interesting what's happening, like, rest of the day, Falcon 180B came out, Gemini is coming out, Claude 2 is coming out. It's, it's just mind-blowing-
Wow!
how fast it's going.
You're such a measured man, you never say mind-blowing, so, but you say mind-blowing, it's got to be really-
Yeah, the innovation in AI is the speed is really great, and it's great to be able to kind of rely on what other peoples are making and then applying it to the product.
So the code generation feature, once you get access to it, are you able to... like, GitHub Copilot is whatever, X dollars, right? Are you able to charge a separate SKU or charge a premium?
Yeah. We announced in earnings that it will be a separate SKU-
Yeah
It will be priced at $9 at the beginning.
That was sort of the expectation that was set before. You'd already talked about it, right? So, how comfortable are you that that price will stick, that people will pay, or maybe you can ask for more?
Maybe, maybe we can ask for more. Maybe, maybe the costs are-
Yeah
Higher. Like, it's very early.
Yeah.
Like, things like inference costs are changing really, really-
Yeah
- rapidly.
Yeah.
So it's hard to get that exactly right.
Yeah.
The value creation is also generates a ton of code, but what is the value you can charge for? And it's, that's, that's all gonna be very interesting.
Yeah. And so it comes out later this quarter, is it, the code generation?
Later this year.
Later this year. Okay. So all right.
... But it's already available in beta for people wanting it.
Okay. Got it. Got it, got it. Okay. The GitLab Dedicated, let's talk about that. What kind of opportunity does it open up for you?
Yeah, so we have self-managed, where you're the only one on that GitLab installation with your own database and everything else. We've got GitLab.com, where we run it for you, we make sure it's up to date.
Mm-hmm.
This is the best of those two worlds.
Mm-hmm.
In your VPC, it's only you-
Mm-hmm.
But we manage and update it.
Mm-hmm.
We're super excited about that offering.
Mm-hmm.
It's, it's now available on AWS, and we anticipate rolling it out to GCP and Azure in the future.
Mm-hmm. And, what kind of customers can you attract with this that you couldn't attract with the, the regular offering?
Yeah, it's in the previous last earnings, we talked about NatWest-
Mm-hmm.
Moving to it.
Mm-hmm.
And it's customers that have these big regulatory requirements, but also want to get rid of the burden of operating GitLab themselves.
Mm-hmm.
As GitLab gets bigger, it's more and more complex to operate, so they no longer want that responsibility. They want to make sure it's always up to date, they don't have to think about it.
Mm-hmm.
So it's for us, it's both converting some self-managed customers to this, but it's also attracting new companies. Last earnings, we talked about a competitive win against GitHub with a customer who is starting with GitLab, and they're able to do so because of dedicated.
Got it. Anything to update us with regards to competition from Jira, as they take a more holistic view of software development life cycle?
What do you mean by that?
So we had Anu Bharadwaj earlier today, and she's talking more about a holistic SDLC approach in the Jira. They already have Bitbucket, right? So they realize that this whole cycle, life cycle that you have been talking about is a tremendous productivity savings as well. Have you seen that manifest in the competitive encounters at all?
I think what Jira does is planning. What Bitbucket does is creating software.
Yeah.
GitLab does do those two things, but it also secures the software and also operates the software.
Yeah.
So it has the complete life cycle, Atlassian cannot offer that.
Yeah.
That's why customers are switching to us. During earnings, we talked about a multinational bank-
Mm-hmm.
moved thousands of users from Jira to GitLab to have everyone on the same DevSecOps platform, and to benefit from these, for example, value stream metrics.
Mm-hmm
that we're able to better give if you do everything within GitLab.
That's amazing. Yeah. They talked about, I think must have been those kinds of customer losses that might have been a catalyst for them, so they're talking about the holistic view of the SDLC.
Yeah, I think that's, that's the right vision, and I think we have the product to deliver on that.
Yeah. And you've been talking about it for many years that I've known you, since when the company was small. So,
I had a quick question-
Yes, please.
For you around just managing costs and, investments, right? We've been talking a lot about, in this conversation, around the breadth and the depth of the GitLab platform, and especially with the product profitability improvements that you've been reporting over the last several quarters and actually many years. How, how are you thinking about them on a go-forward basis, between balancing the various different life cycle, steps in the life cycle, as well as making sure that you're going after new opportunities and go-to-market and everything else?
Yeah. So we told the street they should expect us, in the new fiscal year, to be cash flow breakeven.
Mm-hmm.
We wanna invest responsibly. So the vast majority of our R&D investment is going into what's getting heavy usage, source code control, CI, the testing.
Mm-hmm.
And we're investing in revenue generators, security and compliance. A lot of the new initiatives, like GitLab's Service Desk, the Mobile Ops functionality, the observability stack, we'll initially do a smaller investment.
Mm-hmm.
But we allow our customers to co-create it with us. In the last August release of GitLab, we saw 237 contributions from the wider community. So it's that co-creation, where we don't have to do all the R&D by ourselves, but we work together with our customers to do it. That's why we're able to deliver such a broad platform.
Mm-hmm.
In the two minutes that we have, I want to open up to our clients to see if you have any questions. If you have a question, feel free to raise your hand; we'll bring a mic over. And I have a question while people are deciding. You hired a new CRO, Chris Weber, from UiPath. Prior to that, he was at Microsoft. Can you tell us a little bit about what the mandate for Chris is?
Yeah. We're super excited to have him on board. 20 years of experience at Microsoft. He ran a few billion-dollar sales organization, and he's bringing that expertise. So we're looking forward to scaling with him in a very seamless transition. There is. It's evolution, not revolution.
Mm-hmm.
We'll continue the things we did before.
Mm-hmm.
Focus on the enterprise, it's always been the majority of our revenue, and the focus on the partnerships with the hyperscalers.
Mm-hmm.
Chris is now seven weeks in. Spent a lot of time listening to our customers and the team.
Mm-hmm.
So super happy.
Any other changes in the C-level or one step below, like, to go with Chris? Anything on the go-to-market side that you've... Anybody, new talent, any new talent you've hired?
No, the hire we made before Chris is Josh.
Mm-hmm.
He's our CISO, and security is becoming a bigger and bigger market for us, so he's an equal level, equal member of the executive team, reporting directly to me.
Mm-hmm.
Because we know security is paramount for our customers. We deliver the functionality. People expect GitLab to be secure, so we elevated that function.
Mm-hmm. Got it. Excellent. We'll give you back your minute.
Cool.
Unless there are any final questions. Quick questions? I don't see any questions. But Sid, thank you so much. We hope you had a productive day today, and it was great to see you and catch up with you. And thanks for taking all our questions.
Yeah, thanks for having me.
Absolutely.