Slack

Slack's retention signal was an early message count, not usage or NPS. Mark Roberge's retention indicator for Slack was a message count reached early, a leading indicator after which churn becomes unlikely. Rob Snyder uses it as the model for finding the point after which customers are "weird to churn."

Snyder's advice is to assume every buyer will churn until some event X happens, then design onboarding as a bullet train to X. He suggests delivering to your first customers by hand and repeatedly asking whether they would churn now, to find the moment they say they cannot live without it. For his restaurant retention software, X was a notification in the first two weeks that let an owner prevent someone quitting.

Lenny's Podcast▶ 1:21:56

Slack grew from a small internal tool at a game company. It came out of an internal tool built at a game company after its founder's big MMO idea was humbled. Mark Pincus says Slack may have been only proven and better, with nothing new.

Lenny's Podcast

Hackability and hard-won logic make Slack a team's operating system. Cat Wu calls Slack Anthropic's core operating system for real-time updates, and says its hackability makes it easy to build Slack bots.

Max Schoening notes that the decision logic behind even one Slack notification comes from real scale and years of customer knowledge, which is not worth rebuilding.

Lenny's Podcast

Hackability and hard-won logic make Slack a team's operating system. Cat Wu calls Slack Anthropic's core operating system for real-time updates, and says its hackability makes it easy to build Slack bots.

Max Schoening notes that the decision logic behind even one Slack notification comes from real scale and years of customer knowledge, which is not worth rebuilding.

Lenny's Podcast