A friend of mine is starting a business around personalised yoga instruction on Zoom. 

He called me up and asked if retention is something they should be thinking about this early in the game.

My answer was YES. 

On the one hand, working on retention makes more sense when you have at least a year of behavioural data. Without data, it's hard to reverse engineer what’s working and what’s not.

On the other hand, using analytics to optimise things is exactly that, an optimisation. When it comes to retention, the real work comes from how you've designed your product in the first place.

Good design, from a user retention perspective, all boils down to your primary loop. 

Every habit loop involves a trigger, a response and a reward. 

The problem your product was designed to solve is your trigger. The thing people do in your app to deal with the problem is their response, and delivering on your promise is their reward.

The basic idea here is that if your app does what it claims to do, then people will reach for it when they have the problem it was designed to solve. It's a loop, so the more times you help someone solve their problem, the more readily they will reach for you the next time.

You can reinforce a primary loop with loads of supplemental loops, notifications up the wazoo, new feature releases and flashy ad campaigns...but none of it's going to matter if your core loop is broken. 

Your problem must exist, your core action needs to be something people can actually do, and you have to deliver on your promise. 

This can all come together by accident but it’s a lot safer if you’re thinking about it from day one.