The hierarchy of engagement | Sarah Tavel (Benchmark, Greylock, Pinterest)
Executive Summary
Sarah Tavel explains core action, retention, self-perpetuation, cohorts and focused onboarding. Selected third-party transcript coverage excludes the marketplace hierarchy, financial and investment claims, sponsor reads and later career discussion. Company examples are her historical accounts, unverified here. Untimed editorial topics are not video chapters.
Key Takeaways
Define the core action before measuring growth
Tavel distinguishes downloads and sign-ups from the action that shows someone understands a product's value. Her examples include pinning, friending and writing a note. Choose the action that represents the product's use rather than counting every visit as equally meaningful.
Make use improve the next experience
Her second level asks whether the product gets better as someone uses it. In her Pinterest account, pinning informed recommendations and built a useful collection. The retention question is about value accumulating for the user, not simply adding another notification.
Look for value that accumulates
Tavel describes saved material and a persistent identity as reasons people keep using a product. The product becomes a place with something useful to return to. These examples explain her retention lens; they are not a recommendation to obstruct export or make departure deliberately difficult.
Connect user activity to useful loops
Her third level asks how activity feeds back into growth, re-engagement or other people's experience. She describes sharing, collaboration and recommendations as possible routes. A loop needs a real product mechanism, not the assumption that a larger audience automatically makes the product better.
Check whether a loop gets worse at scale
Tavel describes real-time products where following more people created too many notifications. The intended flywheel could become noise that users ignored. Inspect the user experience as participation grows, because more activity can weaken a loop rather than improve it.
Read retention through cohorts and the core action
She suggests looking at groups who joined around the same time and whether they keep returning and performing the core action. An overall growth graph can hide users disappearing from older groups. Her cohort lens separates acquiring users from learning whether those users stay.
Keep onboarding focused on the first useful action
Tavel warns against copying a mature product's full set of features before new users understand the basics. Too many choices can distract from the action that matters. Guide a new user toward a clear first experience before expanding the product around them.
The Conversation Breakdown
Define the core action before measuring growth
⏱ Untimed topicTavel distinguishes downloads and sign-ups from the action that shows someone understands a product's value. Her examples include pinning, friending and writing a note. Choose the action that represents the product's use rather than counting every visit as equally meaningful.
Make use improve the next experience
⏱ Untimed topicHer second level asks whether the product gets better as someone uses it. In her Pinterest account, pinning informed recommendations and built a useful collection. The retention question is about value accumulating for the user, not simply adding another notification.
Look for value that accumulates
⏱ Untimed topicTavel describes saved material and a persistent identity as reasons people keep using a product. The product becomes a place with something useful to return to. These examples explain her retention lens; they are not a recommendation to obstruct export or make departure deliberately difficult.
Connect user activity to useful loops
⏱ Untimed topicHer third level asks how activity feeds back into growth, re-engagement or other people's experience. She describes sharing, collaboration and recommendations as possible routes. A loop needs a real product mechanism, not the assumption that a larger audience automatically makes the product better.
Check whether a loop gets worse at scale
⏱ Untimed topicTavel describes real-time products where following more people created too many notifications. The intended flywheel could become noise that users ignored. Inspect the user experience as participation grows, because more activity can weaken a loop rather than improve it.
Read retention through cohorts and the core action
⏱ Untimed topicShe suggests looking at groups who joined around the same time and whether they keep returning and performing the core action. An overall growth graph can hide users disappearing from older groups. Her cohort lens separates acquiring users from learning whether those users stay.
Keep onboarding focused on the first useful action
⏱ Untimed topicTavel warns against copying a mature product's full set of features before new users understand the basics. Too many choices can distract from the action that matters. Guide a new user toward a clear first experience before expanding the product around them.
Core Frameworks & Ideas
Hierarchy of engagement
Core action establishes use; retention asks whether use creates lasting value; self-perpetuation asks how activity feeds useful loops.
Core-action cohorts
Compare groups who joined together by whether they continue to perform the action that matters.
Loop stress test
Ask whether more activity improves the experience or creates friction and notification noise.
Actionable Next Steps
- ✓Write a plain-language definition of the core action.
- ✓Inspect what becomes more useful with repeated use.
- ✓Compare retention across joining cohorts.
- ✓Trace one loop and identify where added activity could create noise.
Notable Quotes
“When you perform that action, that means you're an engaged user.”
“The network effect is the strongest thing that you can do.”
“The second thing is just focus.”
The One Thing to Remember
Measure the action that represents value, then ask whether repeated use makes the experience better.