When to kill your product

Product axiom: you should maximize your chances of getting lucky.

Corollary: you should minimize the time you’re spending on branches that aren’t going to work.

How do you do that? You kill branches that aren’t going to work.

How do you know when a branch is not going to work?

Right. That’s a good question to dig in to.

To start with, need to have three structural pieces in place. You need to know:

  1. What kind of success am I going for? This is your end-of-the-line goal. Not your theory about a means to a goal. Eg: build a best-selling mobile game.
  2. How much success do you need? What’s the bar? Eg: top 10 in US App Store.
  3. What indicators are you looking for at this stage that would tell you if you’re on the right track or not? Eg: by June 1, 45% D1 retention in Canada.

If you don’t have these three pieces in place, and you’re deciding whether to continue or kill your product, you’ve got nothing to go on but gut. Gut alone isn’t a good way to steer a product.

If you have these three things in place, then you can apply your gut and your brain, and the guts and brains of all your teammates.

Killing a product isn’t easy. It’s your baby. Or if not your baby, it’s probably somebody‘s baby. So be kind with people’s emotions.

For more, see Supercell’s Jonathan Dower at GDC 2016.