For Airbnb, nights booked makes sense because it shows that people are actually using the platform, rather than just looking at listings or signing up. It also captures activity on both sides of the marketplace. The problem is that if Airbnb only cared about increasing nights booked, they could push heavy discounts or encourage longer stays even when they aren’t the best option for guests or hosts.
Facebook’s early growth team focused on getting users to add 7 friends in their first 10 days because they found that people who did this were more likely to keep using Facebook. I think this is interesting because the metric isn’t measuring engagement directly, but something that makes future engagement more likely. The risk is that Facebook could start pushing people to add random friends just to hit that number. Having 7 connections doesn’t mean much if they aren’t people you actually want to interact with.
Slack looked at teams sending 2,000 messages because reaching that point was a strong sign that a team had started using Slack regularly. But I think this metric is particularly easy to game. If the goal becomes sending more messages, teams might end up with unnecessary conversations, automated updates, and more notifications. That would actually make Slack worse at helping people communicate efficiently.
What stood out to me is that all three companies were trying to measure whether users were getting real value, but none of these numbers measure that perfectly. I think Goodhart’s Law is especially relevant here because once a metric becomes a target, teams can end up changing user behavior to improve the number instead of improving the product itself.
