Hacker News Reader

Comment by gpm | Hacker News Reader

> I know of absolutely no service which failed because it couldn't scale. First focus on building what people love. If people love your product, they will put up with the growing pains (e.g. Twitter used to be down a lot!).

They haven't failed (yet), but I think gitlab.com would be a lot bigger if they scaled faster. Lots of people have been rejected using it because it was too slow.

View on HN

Would it be as big if they had stopped to consider all scaling issues at the beginning?

You could argue it would be bigger. Instead of working around Travis to connect to repositories, the repositories are there. That's the biggest sell for me, anyway.

In 2015 we only had the people to focus on one thing. We opted to focus on the downloadable product. This grew our revenue much faster than GitLab.com. We started to focus on .Com in 2016 but in hindsight not aggressive enough. Now it is one of our top three priorities.

(Note: I agree w/ the OP + commenter whole-heartedly)

The other counter-example I can think of is Myspace who frequently had tons of server errors and page load errors alongside failing to iterate their product rapidly.

gitlab.com is not their core business. It's mostly a demo.

That's not how people perceive it, though.

How people perceive it is not relevant if it actually is a side business

How people perceive it may be impacting your main business.

Only they would have the data to be able to determine that, if anyone would have that data at all.

In 2015 we saw it as our demo. In 2017 we see it as one of our two main products and are improving performance accordingly. 95%+ of our revenue is still coming from licenses instead of GitLab.com. But over time this week change.

Reply on news.ycombinator.com