In terms of organizing teams and, let’s say, you have this experimentation culture – Bezos and Amazon pioneered this approach, with the two-pizza teams – how do you look at organizing teams, to move effectively and efficiently, through decisions and productize ideas quickly?

There are a couple of things there. One is, at one extreme, we could say, we’re going to have teams and the teams can do whatever the heck they want. They can run things however they want; we don’t care, as long as it’s a success. The other extreme would be, we’re going to, forcefully, be very rigid and say, every team has to work in this way. Figuring out what the balance of that is, is a tricky thing. Ideally, you would at least be learning from successes and failures. If a team tries something and it’s a success, they would share that way of working with others and others would adopt it. Or if they did something in a certain way and found that it was a total disaster for them, then they would share that and other people would learn from those experiences.

Figuring out how you take the best of things and the worst of things and don’t do those things, is crucial. Then the other thing is the whole scaling factor. Of course, a lot of things are much easier when it’s smaller. When you have one or two teams and those teams are small, that’s one thing. Now do it with 1,000 people, 5,000 people; that’s a whole different story. That’s where tools and platforms can be very helpful. They allow us to amplify and to handle data and so forth, at scale. It becomes pretty obvious, to most small companies that, at some point, even though they might have tried to avoid it, that they need some kind of tool, to track what is going on. Even if it’s just spreadsheets, you’ve got to have something. You can just be having it all stored in your head and you’ve got to have some organization around it. It might even just be Google Drive, but you have to have tools and you have to evolve those tools.

One of the things that was powerful at Shopify, with the teams that I was working with, was that we built a lot of internal tools, to run the company. That’s what a lot of other companies, today, are doing. What’s also interesting is that there are a lot of companies out there, who build tools like that and they offer services. They will come to an organization and say, hey, we’ve got something that we think is going to transform your organization and will help you run your organization. Sometimes, those things can be useful and sometimes, not. If they’re just basic infrastructures, like Google Docs and those kinds of things, sure; most people need those. But often, what I’ve seen, is that companies that are trying to figure out how they run their projects and how they run their teams, often end up building their own stuff, internally, because it’s something that meshes very well and fully supports the way that they work and their culture.

Investing in that and be willing to build your own tools, without going crazy is important. That’s the other problem, some organizations will say, we’re going to build our own tool and we’re going to have our own this, that and the other. They spend forever doing and they go overboard with it and that’s not a good thing. That often ends up being a failure.

Sign up to read the full interview and hundreds more.