Too many people think of acquisition as the main lever for growth when in reality retention is king. Dan Wolchonok who leads up part of the Growth Team at HubSpot walks through how HubSpot measures growth, as well as how they work to experiment with ways to improve retention to ensure a long lasting business or product unit.
12. Dan Wolchonok - @danwolch - http://danwolch.com
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
GOOD
BAD
RETENTION GOALS
13. Dan Wolchonok - @danwolch - http://danwolch.com
RETENTION GOALS
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
THE IMPORTANT
PART IS THAT IT
FLATTENS OUT
14. Dan Wolchonok - @danwolch - http://danwolch.com
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
ANALYZING RETENTION
Week 1 Retention - Can you get
your users to use your product
more than once?
15. Dan Wolchonok - @danwolch - http://danwolch.com
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
ANALYZING RETENTION
Mid-term retention - Can you
establish any pattern of usage?
16. Dan Wolchonok - @danwolch - http://danwolch.com
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
ANALYZING RETENTION
Long-term retention - How
does your product become an
indispensable tool?
17. Dan Wolchonok - @danwolch - http://danwolch.com
IMPROVING RETENTION
Long-term retention - How
does your product become an
indispensable tool?
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
18. Dan Wolchonok - @danwolch - http://danwolch.com
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
IMPROVING RETENTION
New Features
24. Dan Wolchonok - @danwolch - http://danwolch.com
THE PROBLEMS
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
March April May
25. Dan Wolchonok - @danwolch - http://danwolch.com
THE PROBLEMS
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
March April May
PROBLEM 1: COHORT
RETENTION WORSE IN
LATER MONTHS
26. Dan Wolchonok - @danwolch - http://danwolch.com
THE PROBLEMS
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
March April May
PROBLEM 2: RETENTION
DOESN’T PLATEAU
28. Dan Wolchonok - @danwolch - http://danwolch.com
0
0.2
0.4
0.6
0.8
1 ACTION 3 ACTIONS 5 ACTIONS 7 ACTIONS 9 ACTIONS
“Drive By” Usage in Week One
WHY WOULD USERS ONLY
TAKE A SINGLE ACTION?
42. Dan Wolchonok - @danwolch - http://danwolch.com
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
May June
WEEK 1 IMPROVEMENTS
CASCADED OVER TIME
WEEK 1 IMPROVEMENTS
43. Dan Wolchonok - @danwolch - http://danwolch.com
THE PROBLEMS
0
25
50
75
100
WEEK 0 WEEK 2 WEEK 4 WEEK 6 WEEK 8 WEEK 10
March April May
PROBLEM 2: RETENTION
DOESN’T PLATEAU
44. Dan Wolchonok - @danwolch - http://danwolch.com
FIGHTING LONG
TERM CHURN
45. BREAK DOWN THE FEEDBACK
BY WHETHER IT’S ACTIONABLE
Brian Balfour
46. Dan Wolchonok - @danwolch - http://danwolch.com
QUALITATIVE RESULTS
FOR LONG TERM CHURN
Reason % Possible action Actionable?
Left Company 30%
Capture Personal Email, Recapture
Email Campaign Yes
“Stopped Working” 15% Fix Operating System / UI Issues, Bugs Yes
Cost Confusion 10% Iterate on copy within the app Yes
New Computer 10% Recapture email campaign Yes
Competitor 5% Needed a feature of a competitor No
47. Dan Wolchonok - @danwolch - http://danwolch.com
CUT DOWN ON
NOISE, COGNITIVE
ENERGY REQUIRED
BY OUR APP
ACTIONABLE LONG TERM CHURN
48. Dan Wolchonok - @danwolch - http://danwolch.com
RESURRECTING USERS
49. Dan Wolchonok - @danwolch - http://danwolch.com
INVESTING IN OUTLOOK
50. RETENTION TAKEAWAYS
▸ Soak time
▸ Weekly / Monthly cohorts take awhile to mature
▸ Cohort sizes
▸ You need large cohort sizes to run many experiments
▸ Retention problems sneak up on you
▸ Your top level metrics may be going up
51. RETENTION TAKEAWAYS
▸ Ownership
▸ We set up a NUX (new user experience) team that was
directly responsible for on boarding and W1 retention
▸ Alignment of goals
▸ Ensure that other teams are excited about optimizing for
growth metrics and are willing to make sacrifices to help
improve retention