Retain or Die. App Retention notes from #Growthclinic meetup
In San Francisco meetups can be extremely high quality and Amplitude had a great meetup last night — the theme was “Retain or die” and so the focus was both web and mobile audience engagement and retention with very strong panels and great questions from the audience.
Below is a dump of notes that Linda and I thought were insightful, if you need anything explained more deeply drop me a note — we keep a bunch of best practices in ourDashboard and I’ll probably add a few more from here, there are so many nuggets below that many deserve a deeper look but I wanted to post this before the day got to busy.
For the record my favourite insights(tweeted) are :
- Retention curve shape comes from your product functionality. Curve height comes from improving activation or 2nd action
- Square have models for “most likely to churn” and the best time to call.
- Big wins come from building a culture of iteration — Fareed from Instacart
- Running experiments on retention is like playing battleship. Fire off until you get a hit then hammer on that — Balfour quoting Shaun Clowes from Atlassian
The panelists were: Naomi Ionita from Invoice2Go, Anne Young from Square, Brian Balfour of HubSpot, Julia Lipton of Rise, Fareed Mosavat of Instacart
First thing you would do when you walk into a new role what do you do?
- is this actionable or not. Is it a fit problem or acquiring wrong users
- Slice up retention vertical
- Segment by user source or other attributes
- Attack low hanging fruit like fix bugs that kill retention
How do you turn someone into a good customer
- Talk to customer
- Rise: Eg drop off in first 14 days. We didn’t match user with right coach, so gave them and option to change and this improved retention.
- Early retention is the best place to start
- Start at activation or second time as point of biggest drop off
- Fix bugs that impact retention are quick wins
Activation — biggest bang for buck as it is here that they are forming their opinion
- Understand how to shift the curve up
- Curve height influenced by first experience
- Focus on first experienced
- Understand your frequency rate
- Does your product continually add value over the lifecycle of the user. If they achieve their goals what next? How does your product need to change?
- Be clear about the customer you want and br brave and fire a large percentage of customers that not getting value
- For example — Rise Got rid of weekly plan
The biggest retention win?
- Focus on the persona to improve retention and align your whole company
- It’s scary but It’s worth it. Take the hit now
- If not improving numbers then no use iterating. Walk away
Retention learnings?
- Retention curve shape comes from your product functionality. Curve height comes from improving activation or 2nd action
- The more messages you you send less impact, be strategic.
- Understand primary purpose of a growth Chanel. For example Uber needs messaging to provide service and let users know when a ride is arriving. They wouldn’t want to over use this as it means their messages would be ignored.
- Get rid of the channels not working be aware of channel burn
- Don’t remove too much of the friction to sign up as if it’s too easy they won’t come back.
- Free trialers can be the worst customers, weekly payers for Rise were the least committed customers — weekly payment reflected their lack of long term commitment to their health.
- Getting people to commit is valuable retention tool
- Don’t give free stuff make it valuable
- Protect customer from themselves eg. Twitter sign up is about following and followers, get more feed with more people start tp degrade the experience overtime. Basis of twitter growth plan
- Eg linked don’t really know connections — networks are out of control
- As yourself s- do these engagement loops increase or degrade the experience?
- Must Protect users from themselves
- Battle ship analogy (see above)
- Not about failing fast about learning fast
- Challenge yourself for a hypotheses before testing and have an outcome in mind
- OK, we hate the lags in Appstore. So “Test and learn” on web or mobile web is faster (also Android)
Is it worth focusing on Churning clients to re engage?
- Ask yourself Is this the most impactful ?? They have already informed an opinion so number you will resurrect is small
- Ask yourself At what point did they fall of and how long dormant
- Try turning off a channel and see what happens
originally posted on StreetHawk Blog.
