Finding hidden growth opportunities in your product | Albert Cheng (Duolingo, Grammarly, Chess.com)

· Source ↗

Summary based on the YouTube transcript and episode description. Prompt input used 79979 of 95544 transcript characters.

Albert Cheng (Duolingo, Grammarly, Chess.com) explains how sampling paid features to free users nearly doubled Grammarly’s upgrade rate overnight.

  • Showing sampled paid suggestions to Grammarly free users nearly doubled upgrade rates; free users previously only saw spelling/grammar fixes, missing the product’s full value.
  • Chess.com flipped post-game review to highlight brilliant moves instead of blunders; result: game reviews up 25%, subscriptions up 20%, retention meaningfully higher.
  • 80% of Chess.com users who review games do so after a win, not a loss — opposite of original design assumption.
  • In mature consumer products, resurrected (churned-then-returned) users drive roughly 80% of growth, not new user acquisition.
  • Duolingo tracked in-app screenshots to find viral hotspots, then invested illustrators and animators specifically in those moments to amplify organic sharing.
  • Chess.com targets ~1,000 experiments per year; Albert uses AI-powered text-to-SQL Slack bots to surface cross-experiment patterns and answer ad hoc data questions without analyst bottlenecks.
  • High agency — fast clock speed, moves without prompting, already tested your product before the interview — matters more than domain experience, especially as AI forces unlearning of prior habits.

2025-10-05 · Watch on YouTube

Related coverage