Dylan Ratcliffe

I have not planned for scalability from day one, mostly deliberately, because we need to move quickly, we need to get feedback quickly, and we need to make decisions quickly.

And I would much rather outgrow five or ten different processes than have one process that will serve us for the next ten years, but for the first three years will slow us down.

Because if we get slowed down in the first three years, we won't make the next ten.

So we have definitely scaled inefficiently and then thrown away processes and tools and meetings and all kinds of stuff has been created to solve a problem and then realized that we've outgrown it and thrown it away.

And we recently, some of the big examples recently were we moved from GitHub to Linear.

We did that actually in a week on one of those team events.

We got to a team event and someone said, hey, I intend to try Linear.

And by the end of the week, we had completely migrated to that.

We got rid of stand-ups, which was a weird thing to get rid of.

It's such a normal part of your day in the morning.

You get up and do stand-up.

We got rid of that because it just wasn't working for us.

The team was getting further distributed in terms of time zones, and Linear gave us...

a bunch of abilities with their pulse and stuff like that to not need that anymore.

So we are constantly changing the way we do things to scale for the next three to six months.

But I'm not trying to overinvest in process and things like that to scale for years and years because we need to move quickly.

It's definitely a 50-50 split between the technical results of what the product actually does and the growth in the people who have built it.

From a technical perspective, the simulation model, it is amazing and I'm really proud to be looking at it and it's something that we invented.

Nobody is doing currently, nobody was doing before, and it kicks ass.

We've been doing tests internally where it's able to correctly predict the impact of a change 10 layers deep.