Airbnb chose nights booked because it showed that guests were finding places to stay and hosts were getting business. It also measured usage without the number going up just because prices increased. However, a team could game this by making cancellations harder. The number might look better, but would customers actually be happier? People could feel trapped in bookings and be less willing to use Airbnb again.
Facebook’s early growth team focused on seven friends in ten days because users who made connections early were more likely to keep using Facebook. The number was a signal that a user had found enough people to make the platform worth returning to. A team could push users to add people they barely knew just to reach seven. However, that could fill their feeds with posts they don’t care about and make Facebook less enjoyable.
Slack focused on teams that sent 2,000 messages because it suggested that they had actually tried using Slack together. Stewart Butterfield reported that 93% of teams reaching that point were still using it at the time. A team could encourage unnecessary messages to get users past that threshold, but more messages could just mean more interruptions and time wasted.
I think Goodhart’s Law explains the problem here: once the number becomes the goal, people can improve it without improving what it represents. As a PM, I would also look at customer feedback and whether people kept using the product.
