Refer a friend is a great idea because it ensures your app will go viral and other unverified claims.
But to simply drop the refer a friend feature anywhere into the user experience usually proves to be problematic, as you’ve probably guessed.
The above image shows how Uber asks you to refer a friend just after you’ve called for a ride. This is not ideal because right now the user should be focusing on getting to a reasonable area to be seen and finding the driver.
This second example is from DoorDash, it makes so much more sense to ask about referring a friend after you’ve finished with the part of the experience that requires your strict attention, especially at this point where you’ve already answered whether you like the service or not.
All in all, refer a friend is not necessarily an app performance booster. In fact, nothing is always a performance booster. It depends on when it is presented to the user. Asking someone to rate your app in the App Store as they open the app is perhaps the worst possible time to do so (although it is a very common mistake). Personally, this is the time when I’m trying to get passed all the red tape and into the app. I would also be willing to bet that a lot of users share this objective at this point in the experience. But neither of us will know this for sure until we ask our users.
And I think it’s about time we stopped guessing. It’s not just about what we include, but why (of course) and (more importantly, I’d argue) when.