Data-driven Product Management

Transcript

1. Lean In (to your metrics) Product Manager, Pivotal Labs Graham Siener @gsiener Data-driven Products
2. Do you improve your product every time you push new features?
3. Do you improve your product every time you push new features? Can you prove it?
4. Does anybody like what you’re building?
5. Does anybody like what you’re building? Who are they? What do they have in common?
6. How do you prioritize your product’s roadmap?
7. How do you prioritize your product’s roadmap? Does your team agree?
8. (Customer Development) Steve Blank Build Measure Learn
9. (Metrics for Pirates) Dave McClure Acquisition Activation Retention Referral Revenue
10. • Features have hypotheses • Hypotheses are easily/quickly validated • Validation is quantifiable • Validation doesn’t always require code Goals
11. Every feature should have a hypothesis that you can test
12. Trials aren’t converting because the app is too confusing. Hypothesis
13. Experiment: Plan Abandonment Emails
15. Run the test out of band Create a re-engagement opportunity Bonus[es]
16. Engagement is low because users can’t remember their login information. Hypothesis
17. Experiment: Auto-login emails
18. Baseline: Email click through
19. Metrics also help you say no to features
20. Pretty good Not so good
21. Funnels show you barriers to entry
22. It was a good idea at the time!
23. Homepage context
24. Skip Nice-to-Haves
25. Ignore the Good Areas
26. Dig Into the Bad
27. Metrics: Boss Mode
28. Why should your boss[es] care?
29. Why should your boss[es] care? They won’t unless it solves their problems.
30. Highlight victories Get buy in on the right metrics to track Show time/$ saved
31. Best Practices
32. Metrics should have feature stories, too
33. As a product owner, I want to see which guests register for an account, so that I can find the parts of the site that helped them convert.
34. Validation should not be a blocker for acceptance.
35. Validation should not be a blocker for acceptance. Track validation outside of Tracker. Kanban anyone?
36. Use consistent events and identifiers across platforms
37. Send all your metrics to one place (This is hard!)
38. Use KISSmetrics or MixPanel if you value your time Google Analytics otherwise
39. Use A/B testing and feature flipping sparingly!
40. Let’s go spelunking
42. Daily reminders help people enter their time ...on time Hypothesis
43. iPhone: daily push reminder Android: no reminder
44. Android
45. Android
46. iPhone
47. Recommendations: Try it for Android? Get more data (and put it in one place)
48. Start Today
49. Q/A