How to Fuck Up Retention
Feature selection with Stanford dropouts Muck and Zusk
Consider a fictional startup, VidAI, founded by Stanford dropouts Muck and Zusk. Their two-sided AIGC platform aims to achieve TikTok-level engagement (~2.5 hours/day) while cutting content production time to 1 minute for a hit video. Muck and Zusk believe this app will finally fuel their Evan Spiegel dreams. Like the Clubhouse founders, they’ve launched a number of failed consumer social apps and need this one to hit, else they’ll run out of funding. This time, they need to stop fucking up retention.
Forget about metric goals
According to experts, consumer social retention after six months is good at ~25% and great at ~45%. Muck and Zusk previously thought their apps were fine at 10% six month retention, but they’re not gonna be those hand-wavey bros anymore. First goal — 50% D1 retention and 45% D30 retention by end of Q3. They know that’s the first step to go from a 🫠 retention graph to a 🙂 one. Now that’s a SMART goal!
Feed your ego rather than talking to users
Users —> Product
Muck and Zusk determined that their user persona is 11-18 teenagers who are creative, but somewhat lazy, and active on other social platforms. In previous launches, they spent their time talking to random VCs for an ego boost and perfecting UI details because they let their designer drive product. This time, they’ve learned to put on their degen teenager hats by joining Discord servers, posting on Reddit forums, and spamming Insta DMs. They segment users based on the following criteria:
For VidAI to work, they have to incentivize stickiness on both sides of their user persona. Addicted viewers don’t stick around without talented creators. That means segmenting users into cohorts — not just by engagement behavior — but also by usage frequency. Namely — churned, retained, and advocating users.
To parse efficiently, Muck and Zusk track user feedback in a Notion table, filtered by type of user, form of feedback, user flow, and target metric.
Haphazardly select features
Product —> Company
To hit their retention goals, Muck and Zusk realize they can’t just listen to users. Especially teenagers who can’t predict what they want half the time, besides something cool!
Focused on their usage frequency cohorts, Muck and Zusk ask themselves the following questions to kickstart feature development:
How often will the feature be used? Does it give users super powers?
How many users is this feature relevant to?
How hard is it for us to build this feature?
How confident are we in our answers to the previous questions?
Using this RICE framework, they are able to prioritize the user feature requests they collected. They then take another pass through the features and make sure each addition strengthens their position in the market. Again, they know they can’t just listen to the teenagers, else they’d end up building a one hit finger on the app product that has 0% D30 retention.
Treat users like investors
Company —> Users
Download the newest metaverse AI experience with interoperable NFT avatars! Muck touted, in an ad for a defunct app. They’d raised millions from investors off this pitch — surely it would work for users, right? Turns out teenagers are burnt out from metaverses, and don’t care about NFTs.
This time around, they employed a regular update schedule; every Friday they announced a slew of new features and sent a sample AI-generated video to inspire creators. The teens felt like the app was alive — that they were early adopters of the next TikTok, that the devs shipped at a fast rhythm on beat with their problems, that they were more than happy to stick around.
By avoiding their previous mistakes, Muck and Zusk may get to their TikTok-killer goals. Learn from their fuck-ups so you don’t end up in a cage fight with your churning users.
Katie, please check your email with the subject of "insight in gaming industry"
thank you.