GitLab introduces user limits for Free users on SaaS

Blog post: Upcoming changes to user limits on Free tier of GitLab SaaS

We are always exploring ways to become more efficient as a company. To ensure we can continue to offer the free tier to small teams, we are limiting the number of users per namespace on the free tier to 5 users. These changes allow us to redirect our investment to maintain and improve the quality of service on GitLab SaaS and to invest in research and development for product innovation of the GitLab DevOps Platform.

This detailed frequently asked questions document outlines the upcoming changes and how you can manage the changes to your free GitLab SaaS account.

In order to publish your feedback regarding this change, please add a reply to this thread.

Thank you very much for taking the time to provide us with your perspective. Please allow our Product Team some time to review the input you add here.

Please note that this forum recognizes GitLab’s Code of Conduct.

Are there plans to notify impacted users about this change? I found about this from hacker news but this post is dated Jan 18th. I can’t find an email notifying me.

This is a great shame. I have been advocating the use of GitLab for small (<5 devs, <40 reporters) non-commercial hobby projects in the space I’m active in (video game modding) for years, and have personally tried to ensure that my CI/CD tooling supports GitHub and GitLab equally (GitHub - cwtools/cwtools-action: Run CWTools on your Clausewitz mod PDXScript code in parallel to your builds thanks to GitHub Actions.).

These projects tend to mainly consist of young, novice, programmers, who like playing games but have no experience with software development, git, and open source licenses. They’re rarely comfortable using public projects due to the slightly murky copyright issues and the simple desire not to have people steal their work (licenses don’t mean much in the non-commercial world…).

I have always encouraged teams to use GitLab, previously because it had free private repos, and then to try and balance the dominance of GitHub in outreach. Many of these amateurs go on to study computer science and get jobs in the industry, and I figured the tooling they are first exposed might help them push employers to avoid the full Microsoft stack.

I’ve even gone as far as trying to self-host GitLab runners to try to encourage adoption of GitLab.

While these teams might end up with dozens of contributors over the years, they will have literally zero funding (as it is generally prohibited by publishers), despite contributing significantly to society (Have you ever played video games with mods?). One of the projects I help manage here has had hundreds of thousands of players.

You don’t provide community programs for this use-case. They’re not open source, they’re not educational (you don’t even support non-profit research!). And there is no way that a group of teenagers are going to find $200+/m to use GitLab.

Instead, they’ll have to use GitHub.

Hi @tboby,

Thank you for your feedback! Our Product Team will be reviewing all the replies to this thread and so we appreciate you taking the time to share your thoughts.

The announcement went out today . There’s also a blog post with more information: Upcoming changes to user limits on Free tier of GitLab SaaS | GitLab. I’ve also updated the date on this post, thank you!

Hah, OK, I was a little surprised I hadn’t heard of it sooner, but that makes more sense!

Really? This change is worth for the 2% users? Ditching the Starter tier wasn’t enough? (Sorry, I had to)

I mostly agree with @tboby here. I was encouriging GitLab over alternatives, because I love the features.

Free tier was a great starting point for small/mid sized companies which business is not in IT, but still needed a way to handle IaC or some kind of automation. Maybe some of them eventually grew and switched to paid tier, some of them haven’t. But the Free tier was the first step for them to know the platform and get experience long-term. Sorry, but 30-days trial isn’t enough to try all the features, there are a lot of them. You can’t try and experience it properly when you have other day-to-day tasks to do.

Maybe you have your analysis and polls, maybe you know what you are doing. I am thinking that users will look for other free alternatives (hello GitHub) and eventually grow with those.

It would be interesting to see how many of the customers on Paid tiers actually started with Free tier for at least couple months.

Yes, money is in the Enterprise business. It seems that’s your path now. I am OK with it. It’s just make me … sad.

I said something similar on HN but I figure I’ll repeat it here as you’ve said your Product Team are monitoring feedback

Firstly I’m disappointed Gitlab didn’t send out an email about this. I found out through a friend mentioning it, and I think they’d seen it mentioned on HN. Hiding something this serious in a blog post isn’t a good look from where I’m sat.

I’m heading up a team of volunteers who handle the internal software and infra for a non-profit fan-run event. With the industry we’re in (if you can call it an industry), we have what’s probably best described as an ‘excess of time and a shortage of money’. To give an example of the sums we’re talking - the money available to run the event for a year is around $10,000 after hire fees for the venue. From that we have to pay for everything to run the event, plus some to pay the bills until next year.

Our team is 10 people, so Gitlab is asking us to find $19 x 10 x 12 months = almost $2300 per year (actually probably pretty close to that with VAT added). You’re asking us for nearly a quarter of our total takings, for one product, for ten people.

If I took that to our Secretary, he’d tell me to go see a doctor and get my head examined.

The next thing he’d ask is “what other options are there” – and I’d have to tell him that Github is $48 per user per year (so already about a quarter of the price per seat), does everything we need, and would cost us $480 per year. I could probably get approval for that.

This is the same price as the Starter plan which was axed last year.

The SaaS and self-hosted packages being the same price is also a bit perplexing to me. I appreciate you’ll have additional support costs for installation, but those go away after the product is installed. The costs for hosting and CI minutes would be my responsibility. When Atlassian offered self-hosted products, they were at slightly reduced prices compared to SaaS (or perhaps SaaS cost more to reflect the hosting costs). It’s common practice in the industry, or at least seems to be.

If you gave me the option of buying 5 more seats on the Free tier at maybe a few dollars per user per month, I’d gladly take you up on that with that.
I’d also be happy with paying a few dollars per user if that got me a few extra features (Epics and Promote Issue To Epic would be nice).
You could even tell me that comes with zero CI runner minutes – I’d happily set up a VPS to serve as a CI runner.

But asking for four times the cost of Github for a very similar product and a whole pile of features I won’t use is a very big ask, even as much as I like Gitlab as a product.

I’d probably have a very similar outcome if I asked for Gitlab in a larger business. “Hey, why do you want Gitlab for $19/user/month when Github does almost the same thing for only $4?”

I get that you need to bring money in, but it feels like you don’t want the custom of small enterprises. Seems like the mantra of ‘today’s acorn is tomorrow’s oak’ has been forgotten.

For me this means, I’ll have to move my free software somewhere else. Sending > 20 applications for the Open Source Program once a year is definitely not an option.

Maybe you could offer a $40/user/year tier and give only everything in the FREE tier minus the 5 user per namespace limitation? $228/user/year is pretty rough jump

Hey there, I am not intimately familiar with your use case but if your projects are within the same group then you would only need to submit 1 application which would be renewed annually as long as you continue to meet the eligibility requirements of the program.

Hi @tboby,

We appreciate your feedback. Thank you for taking the time to share your perspective.

Regarding your question about our plans to notify impacted users: In addition to today’s blogpost, all impacted namespaces will be notified in-app of the coming changes to user limits for the free SaaS offering in the coming weeks. Those who still have yet to respond to the in-app notification will also receive an email notification. We have an MR created to add this clarification to the FAQ as well.

This is way too expensive, you’ll need to bring back the $5 starter plan. GitHub and Atlassian offerings are much cheaper now.
It’s too easy for a small business to go over 5 users when you include testers and part time/casual workers. Sounds like if people are wanting to keep using Gitlab in this category, self hosted CE is the way to go.

Hi @philpem,

We appreciate your feedback. Thank you for taking the time to share your perspective.

We understand the desire for an email to inform you of the coming changes to user limits. In addition to today’s blog post announcement, namespaces impacted by this change will receive an in-product notification in the coming weeks. Those who do not engage with the in-app notification will also receive an email notification approximately 4 weeks prior to the effective date.

Hi @balonik,

Thanks for the feedback. We thought long and hard about this and as costs to maintain the free product continue to grow we did our best to find a balance of continuing to offer our free SaaS offering while still working to improve the economics of our business.

Hi @awulf
Thank you for sharing your feedback. We understand this is a concern for you.

Be assured it was not an easy decision. We thought long and hard about this and as costs to maintain the free product continue to grow we did our best to find a balance of continuing to offer our free SaaS offering while still working to improve the economics of our business.

May we recommend you to contact our team so they can guide you on your options?

Hi @chrisgo - Thanks for your input. We will share this feedback with our pricing team.

Thanks for the fast response!

My use case are game jams but I think it similarly applies to hackathons or student projects. Multiple times a year, the team changes (though it may have some overlap).

can you make the cost analysis public that lead to this decision? I cannot see it mentioned in the blog post.

Hi @winniehell we appreciate your interest to work with us on this.

While we try to make many things public because transparency is one of our values, some things, including financial information like this, can’t be made public because we are a public company. Here are a list of things we consider internal, some of which is limited access internally too.

Hi @winniehell - if all these projects are within the same group and all the projects are public open source, you only need one application which will be renewed annually - as long as you meet the requirements. We are in process of automating the onboarding process for the Open Source program to support users.

Hello,
We are a small startup started in may 2020, with limited fundings and honestly, I don’t see any other choices then moving out to gitlab.
With 8 devs, we are barely above the limit.
Also, we are not eligible to the startup program since we raised more than 3m$ in seed, however, we are an hardware and infrastructure business, and most of this money goes to raw materials (which has already drastically increased), chips, shipping, and field operations. I understand that for a SaaS business, 3m is a lot, for effectively building physical stuff, it’s another story. We do everything in house, with devs which are not only devs, but also goes on the street to fix our stuff, du customer support, etc. We are far from making money (totaled billed is around 3keur now)
It’s a shame, since we committed a lot of time having stuff running on gitlab on our own runners. We don’t need anything from the premium tier, but I see no other option. And quite frankly, if we do the move, we will never come back.

Regards