GTLB - GitLab S1 filing

As someone with zero tech background, I’m seeking input from the board on GitLab.

GitLab filed their S1 on Friday. https://www.sec.gov/Archives/edgar/data/0001653482/000162828… I’ve heard of GitHub, owned by Microsoft, which I understand is a competitor. It seems a lot of people do like/prefer GitLab though.

Here’s what I’ve learned so far:

GitLab was incorporated in 2014. It is cofounder led by CEO Sytse Sijbrandij. The company has very positive glassdoor ratings: 4.6 stars, 274 reviews with 97% approve of CEO.

It has 1350 employees in 65 countries and is a 100% remote-only workforce. GitLab has pioneered “The DevOps Platform”, a single application that “brings together development, operations, IT, security, and business teams to deliver desired business outcomes. Having all teams on a single application with a single interface represents a step change in how organizations plan, build, secure, and deliver software.” They aim to reduce software development cycle times from “weeks to minutes.”

Ok, to me that sounds like they should be mission critical to a company. They should be a very sticky platform. The numbers (see further below) mostly looked good to me.

This is a SaaS company. The company believes the total addressable market is $40 billion.
GitLab uses an open core business model.
A free tier with lots of features to encourage use of their platform, solicit contributions, and serve as lead generation for paid customers.
Two paid subscription tiers allow access to ‘more features’ relevant to ‘managers, directors, and executives.’
A typical subscription contract is 1 to 3 years.
As usual, this company like many others, boasts of a virtuous flywheel cycle: more community contributions lead to more features, leading to more users, leading back to more contributions.
They have a land and expand strategy that starts with developers and expands to executive buyers.

GitLab certainly appears to be a hypergrowth company. Their run rate revenue based on last quarter ending 7/31/21 is 233M.


                    FY20    FY21
Total FY revenue    81.2    152.2
FY growth YoY                87%

GitLab has very high gross margins:


               FY20  FY21  
Gross margins  88%   88%  

They define “base customers” as those generating $5000 or more in annual recurring revenue, which are growing:


                1/31/20  7/31/20  1/31/21  7/31/21
Base customers    1662    2126      274     3632
Growth rate YoY:           65%               71%

They are also growing 100K and 1 million ARR customers fast:


                    1/31/20   7/31/20   1/31/21   7/31/21
$100K ARR customers   173       219      283       383
Growth rate YoY                          64%       75%

                          1/31/20    7/31/20    1/31/21    7/31/21
$1 million ARR customers    11         15         20          27
Growth rate YoY                                   82%         80%

Low churn (note, their fiscal year end is January 31):


                                     FY20    FY21
Dollar based gross retention rate    97%     97%

However, I have a concern about recent revenue growth…


3 months ending     4/30/20   7/31/20   10/31/20   1/31/21   4/30/21   7/31/21
Revenue (millions)  29.515    34.362    42.152     46.147    49.930    58.127
QoQ rev growth                16.4%     22.7%      9.5%      8.2%      16.4%
YoY quarterly rev growth                92%        73%       69%       69%

QoQ growth appears all over the place, but it looks like they claim there is seasonality: “we typically enter into a higher percentage of subscription agreements with new customers and renewals with existing customers in the third and fourth quarters of each year.”

YoY quarterly growth, however, has decelerated over the last four quarters…but they boast a whopping dollar based net retention rate of 152% as of July 31, 2021.


        FY20    FY21    6mo ending 7/31/20   6 mo ending 7/31/21
DBNRR   179%    148%         153%                  152%

Improving op cash flow margins:


                              FY20     FY21     6 mo ending 7/31/21
Operating cash flow margin   (74.1%)  (48.4%)    (35.8%)

They are being very aggressive with sales/marketing, having spent more than their entire fiscal year revenues on it (99.225M in FY20 and 154.1M in FY21 on sales/marketing expense, including SBC). Sales/marketing expense increased 55% YoY, while revenue increased 87% YoY.

Now, I’ve looked at some online comments (https://news.ycombinator.com/item?id=28568101) on GitLab versus GitHub, and many seem very positive:
“Gitlab is one of the pioneers of “remote-first” and “building in public” to the extent of sometimes even live-streaming CEO meetings and sales pitches"

"Yes, Gitlab has been leading the way for a while now, consistently introducing new features that Github eventually copies. It’s a real testament to the power of competition. I first started using Gitlab due to free private repositories, which Github eventually added. Gitlab had free built-in CI/CD first, and Github eventually followed. I’m still on Gitlab these days despite Github catching up, and still enjoying features like organizations with subfolders that Github lacks.

But some negative comments:
“Yeah I absolutely support Gitlab and love seeing new projects use Gitlab over Github.
But to be fair, they have a massive backlog of issues to fix. Basic issues too, like variables not expanding correctly in CI jobs, or Google not being able to index projects on gitlab.com unless there’s another page already linking to it.

"I’ve been using Gitlab.com and Gitlab on-prem since 2013 and over the years I’ve found many of these bugs that I feel should be top priority instead of new features.”

“I really wanted to like gitlab, but have had so many reliability issues. The UI thinking the source branch doesn’t exist, CI jobs not running for hours etc”

“I haven’t personally had a compelling enough reason to move from GitHub to GitLab, so I mostly just use it to mirror a couple of my repos there in case people prefer to browse various open source projects via GitLab.”

What are the thoughts on GitLab, particularly regarding revenue growth decelerating recently but great margins, DBNRR…and anyone who has experience with their product, or comments on GitLab vs GitHub would be helpful (are there other competitors besides GitHub?)

We’ll see their IPO pricing soon. They were valued at $6 billion last year, and I’m sure they’ll come out of the gate much richer than that. I’ll be definitely keeping them on watch (they will list on Nasdaq as ticker GTLB)

59 Likes

Do you remember Jfrog (FROG)? Bert Hochfeld has recommended it a few times, and Saul had taken a small position. Saul sold it early on saying Jfrog sells a tool for developers’ use and hence is not infinitely scalable. That’s exactly what Gitlab does too! They sell code repositories and build tools for developers to develop their code and roll it out. (I sold Jfrog in 60s, it’s in 30s now)

Companies like Cloudflare, Cloustrike, MongoDb, Snowflake, Confluent grow as data grows, and data grows exponentially. Companies like Jfrog and Gitlab will grow as head counts of developers grow, which grows slowly.

Like you mentioned, when Github was bought by Microsoft, everyone flocked to Gitlab, which also proves that the product is not that sticky.

Lastly, along with Microsoft’s Github, their other competitor is? ?Atlassian? (TEAM) which has a repository-product (bitbucket), build tool (Bamboo)? and multiple additional tools for developers under one umbrella.

57 Likes

Thanks for your insight. I have no idea what JFrog is. (And I didn’t know this board existed before June). So it seems even we suppose GitLab were the best product among its competitors, it would not scale in the long run due to the nature of the platform/developer headcount relationship?

3 Likes

JW,

My developers use Gitlab and really like it. We had internal supporters who looked at several of the main options (GitHub and some on prem stuff) and decided Gitlab was the best.

I see no compelling reason to switch and as we add developers we will add spend.

This is only for companies that write software so it is a bit more limited TAM than some companies (eg CRWD). But there are of course many many companies that do.

Rob

4 Likes

I have worked as a software engineer for over 10 years. I have used both GitHub and GitLab within the last few years.

For non-software engineers, I think the most important thing to know is that both GitHub and GitLab are built upon Git, an open source technology. When talking purely about Git and its features, it is relatively simple to switch from one provider to another, or even manage it yourself.

Beyond Git, both providers have added additional features that are not built directly into Git. These features increase the stickiness of their offerings. I haven’t done an in depth comparison of the features, but my understanding is that they are pretty similar to each other. The most important feature is the ability to review code. Most if not all of the other features can be obtained using other tools. For example, Atlassian’s JIRA would replace GitLab’s Planning feature, and JFrog’s Artifactory would replace GitLab’s Package feature.

For a new software company starting out, I think there is a reasonable chance that they could choose either GitHub or GitLab for their complete feature suite. For an existing company which already has software development tooling set up, I believe it is much more difficult to make the decision to swap out individual tools for GitLab/Hub features. Maybe there is some cost savings in moving to a single provider, but engineering time is expensive and migrating tooling is risky, so I think it is a hard sell.

This is all anecdotal. I haven’t looked in depth at the S1 yet.

9 Likes

Companies like Cloudflare, Cloustrike, MongoDb, Snowflake, Confluent grow as data grows, and data grows exponentially. Companies like Jfrog and Gitlab will grow as head counts of developers grow, which grows slowly.

This is an interesting quote but in my opinion, only partially true and with an important caveat. As Datadog’s CEO mentioned in the last earnings call:

Two, when you think long term, the – it’s almost a given that there will need to be a different way of charging for some of the – for capturing some of the value provided to customers that can’t just be attached to the straight volumes of data that are being exchanged because those volume of data are exploding exponentially while our customers’ revenues are not going to explode exponentially.

In other words: Yes, data may be exploding but that doesn’t mean that the revenue will show the same level of growth. The customers still need to be able to pay the bill.

20 Likes

Hi johnwayne,

GitLab is, hands-down, in my opinion, the absolute best tool for what it does. However, it has a LOT of competition, the two largest of which are GitHub (owned by MS now) and Atlassian’s Bitbucket. It’s important to understand that there are some differences between GitLab and the other two as well. GitLab is a platform of very well integrated tools for the software development lifecycle, similar to Atlassian’s offerings.

Atlassian made their name with their flagship Agile Methodology management software Jira, and then branched out to add tools like Confluence (a Wiki) and Bitbucket (a Git server) (Atlassian also had other version control systems as well prior to Git taking over the world, but I digress. The point being, that these are modular pieces to the Atlassian offerings).

GitHub has made its name as a Git repository server, and has only recently added other related and integrated tools.

JFrog made its name as an artifact repository (a place to store the fully compiled and built pieces of software) and has since added on other related and integrated tools.

Each platform has a core strength and now significant overlap with the others. The core competency for one is not usually the core competency for the others. But they all overlap to some extent. The basic tools required for a software development shop are:

  • Version control server (git)
  • Build engine (Jenkins, CircleCI, “pipelines”, etc.)
  • Artifact Repository (Nexus, Yum, etc.)
  • Feature/Bug tracking engine (e.g. Atlassian’s Jira)
  • Documentation tool (Wiki - Atlassian’s Confluence)

All of the above have some or all of these tools. Gitlab was the first commecial offering to integrate all of these concepts/tools into one seamless package offering for developers. Since then, everyone else has tried to do the same. All of them are playing on the “DevOps hype” (full disclosure, I’m now what they call a “devops engineer” - the hype is all BS :slight_smile:

GitLab’s strength and moat, in my opinion, is it’s pipeline implementation. The “standard” is Jenkins, which is, at this point, an antiquated “build system”, which has had all sorts of plug-ins developed for it so that, in theory, it can do anything and everything for everyone. IMO, it’s garbage. But, it’s also the industry standard, and “free” in the open-source, unsupported manner which costs you a bundle in the form of paying employees to manage and maintain this infrastructure.

Jenkins’ “pipelines” are all written in a language called Groovy, which is cumbersome, difficult to learn, and mostly not used for anything other than special purpose things like Jenkins pipelines (not to say that Groovy isn’t used anywhere else, but that it’s a Domain Specific Language, in that, people use it within very specific applications like Jenkins. It’s based on Java, which belies Jenkins’ background as a build tool designed by and for Java developers. So of course they’d pick a language like Groovy that only Java developers would know… But I digress again.)

GitLab took its lead on pipelines from another open source project called CircleCI, which implemented it’s pipelines using simple YAML files; plain-text, no special language to learn, just list the steps you want the system to execute, and it away it goes!

Isolated build systems like Jenkins and CircleCI need to poll remote systems like isolated build servers to know when to trigger a build.

GitLab has it all built in. The definition of the pipeline exists in the Git repository. When there’s activity in the git repository, it just fires off the build.

Atlassian’s pipelines are similar to GitLab’s, but not nearly as robust. JFrog has pipelines following in the same fashion (no idea how they compare to GitLab or Bitbucket pipeline. But they’re YAML, so automatically better than Jenkins!)

For Feature/Bug tracking, the industry standard is Atlassian’s Jira. It’s the absolute worst product for the job, except for all the rest of them! GitLab has something built in, but it’s crap. GitHub is similar to GitLab, their issue tracker is crap.

They all have wiki’s. Atlassian’s Confluence is (currently) the industry standard.

Of all these companies, JFrog is the only one which has a unique-ish offering (and yet there are competitors here too!) in it’s Artifactory artifact repository product. There are others, but they currently offer more types of artifact repositories in a single, integrated package than anyone else.

Currently, the industry standard for “team ware” is Atlassian. They hit the trifecta of issue tracker, git repo, and wiki earlier than anyone else. And when all their software moved into a cloud-based SaaS offering it just got better from an integration/ease-of-use perspective.

There isn’t another issue tracker out there that comes close to Jira. And if you have Jira and Confluence, which are very well integrated with each other, you may as well go with Bitbucket which is as well integrated.

This means getting GitLab into an established company where Atlassian is the incumbent will be extremely difficult. My company uses Jenkins (because it’s “free”!) along with Bitbucket. Upon my arrival I tried to bring in GitLab, but lost that battle for many reasons (mostly cost). So I’m slowly trying to move us away from Jenkins to Bitbucket pipelines. Not nearly as good as GitLab, but way better than Jenkins. We also use JFrog because Artifactory is better than anything else as an artifact repo. No matter how you slice it though, most shops will end up using several of these tools in combination. We use Jenkins, Atlassian, and JFrog. Others will use GitLab + Jira/Confluence + JFrog. Or GitLab + Jira. Or GitLab + Bitbucket. etc.

Why spell all this out? Because as others have pointed out, these are all tools for use by developers. And we developers have strong feelings about our tools, and when something better comes along, we will jump to the next thing as fast as we can! (e.g. Today we use git. In the past, the “standard” has been ClearCase, CVS, Subversion, Perforce, etc.)

As a general rule, I don’t invest in “infrastructure plays” (i.e. networking or hardware companies, etc.) These companies are nothing more than “infrastructure for developers”. I lump Slack into this group as well.

Sure, they may be SaaS companies and have recurring revenue, but generally, they have no moat, huge competition, and fall out of favor very quickly unless they can adapt, and aren’t nearly as sticky as some may think (it’s trivial to move your source code from one VC system to another, even easier moving between providers using the same VC engine like Git). The other aspect is that growth does not tend to be rapid. They all license “per-head”. Therefore, they need to expand entirely by adding whole new accounts. Sure, this year’s revenue is baked into next year’s, but it’s not likely to expand much. Headcount tends to stay pretty static in most companies unless the company itself is in a rapid expansion mode. But even if they are, engineering is not usually where that happens. Engineering goes through its rapid growth very early on in the company’s growth. After that expansion is almost always in management and on the business side (sales, marketing, account mgmt, etc.). A company with 100 developers this year is likely to have about the same head count for a long time.

Anyway, this has gone on entirely too long, and I apologize for my long-windedness bordering on irrelevence.


Paul - A DevOps guy who has played with all of the toys.

38 Likes

All good points about competition, and there is a dizzying amount of tools out there with affinities and loyalties to contend with. One of GitLab’s selling points is that you can adopt their solution but it doesn’t require a complete rip and replace of existing tools right away - it’s possible to adopt components over time and retiring the redundant offering when it makes sense. While some competitors have similar offerings under their brand umbrella, they aren’t completely integrated products which is exactly what GitLab was built to be; plus it plays well with others.

FWIW, GitLab is quite ‘open’ - roadmap, competition comparison, etc. all publicly published on their site (i.e. https://about.gitlab.com/devops-tools/github-vs-gitlab/). If there’s an issue with info posted there they will change it quickly.

My VAR company was the first public sector GitLab partner and made the first public sector sale back in ~2016. They are now in the majority of DoD software factories; some mature and some are growing rapidly. Likely a small percentage of the overall sales from a vertical perspective, but they are strong in US public sector.

When MS acquired GitHub, there was a significant exodus from the repository because the user community didn’t have faith in MS ability to support the platform going forward.

Finally, GitLab is getting to developers early on. The are providing free, unlimited licenses of their top-tier functionality (SaaS or self-managed) to educational institutions. That will build a user base and a demand going forward.

3 Likes

Regarding GTLB, MSFT earnings call today had relevant details about its competing GitHub:

Today’s Q3 2022 MSFT earnings call:
"Now to developer tools. From Azure DevOps and GitHub to Visual Studio, we have the most comprehensive and loved developer SaaS service. Increasingly, every new developer project starts with our tools…And GitHub usage is increasing among both independent developers and startups as well as the world’s most established enterprises.
90% of the Fortune 100 use GitHub. In fact, Mercedes-Benz, for example, is using GitHub Enterprise to provide a unified development platform for more than 20,000 employees to build, ship and maintain software."

From Q1 2022 MSFT earnings call:
"GitHub is now home to 73 million developers up 2 times since our acquisition three years ago.
More and more businesses are choosing GitHub Enterprise to provide their developer teams the most advanced platform to build, ship, and maintain software. This quarter alone, we introduced more than 70 enterprise features, 84% of the Fortune 100 use GitHub, and we are seeing growing usage from digital-native companies and the world’s most established firms, from Pinterest to Procter and Gamble, from Stripe to Societe Generale."

While I am sure GTLB should continue to grow fast in the near term, in the longer term I remain concerned by competition from MSFT, especially whenever more of the low-hanging fruit developers/enterprises are taken up

14 Likes