Coconote · B2C · AI note-taker for students
Double onboarding from 5 to about 13 screens in one big change
Trial conversion up 16%; one-or-two-screen tests had only given small wins
Workedonboardingexperimentation
What they did
After iterating one or two onboarding screens at a time (small wins, hard to make big progress), they made one big jump: from ~5 to ~13 screens ending in the paywall. Added screens include login, notification permissions and mostly personalization (what the user is studying). Login was moved to the last screen, after the paywall.
What happened
Initial trial conversion rose ~16%
This shipped together with other changes, so the result can't be credited to this alone.
In their words
The Superwall Podcast · Brett Bauman & Zack Hargett - COPY this ai app's $6.7m/yr marketing strategy“we were very iteratively experimenting in onboarding, like one or two screens at a time, testing the uplift, moving things around, um, and it just felt It felt like you would get a lot of small wins, but difficult to make big progress.”
“we at least doubled the length of our onboarding to, I think, like 13 screens. And that really really helped with our initial trial conversion. Um, I think it went up 16%”
“a lot of it is making the app feel more personalized for the user going through it, um understanding what they're studying”
Related topics
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from Coconote
All 31 Coconote tactics- When people cancel, offer 7 more free days instead of a discountSaved 25% of cancellers; beat a 30% discount and a 3-month pause
- Double onboarding to ~15 screens with more questions and social proof+16% trial starts
- Ask people to log in after the paywall, not on the first screenBiggest onboarding win; removed ~10% drop-off
- Charge students $99/year and push the annual plan~80% chose annual
- Launch a viral 'PDF to brain rot video' side site to funnel users~41M views on one video, but no paying users; shut down
- Raise the annual price from $99 to $129More users and more revenue at the same time
More from this episode
- Coconote: Build a novelty free tool (text to brainrot video) and push it on TikTok100-200M views and 3,000 trials in days, but only ~$25K revenue
- Coconote: Post short videos that show the app solving a real student problemSeveral ~20M-view videos, ~99% organic, converting far better than viral novelty content
- Coconote: Mix a known game (2048) with studying and promote a catchy URL in videosOne video reached ~2M views; no conversion numbers
- Coconote: Build hidden app screens only creators use, made to look good on cameraUsed in the 18M-view video; no separate numbers
- Coconote: Keep a small team of 10-12 part-time creators, led by a proven oneLead creator drove hundreds of millions of views; founders say it worked
- Coconote: Recruit creators who aren't influencers via Handshake, DMs, email and referralsA Spanish tutor with no content experience became a top creator in 9 months
- Coconote: Let people pay before creating an account; move login after the paywallFounder calls it significant; no separate number
- Coconote: Let users record a whole lecture before asking them to payCalled a big win; no number given
- Coconote: Put a little ad money behind organic videos that already went viralNot much success; ~99% of views stayed organic
- Coconote: Pitch the app as great for ADHD in videos and the first App Store screenshotUsers say it helps their ADHD; no numbers
- Coconote: When trial users cancel, offer them 7 more free daysKeeps 27% of cancellers in their trial; a smaller share later pay
- Coconote: Hire an investment bank to run the sale so you keep buildingSold to Quizlet ~18 months after launch; bank ran it end to end
- Coconote: Hire engineers who can also do backend or a second platformSmall engineering team ships a lot; with 50%+ EBITDA margins
- Crew: Post creator-style TikToks and Reels to drive app signupsHundreds of thousands of signups for an earlier app
- Coconote: Track each creator's results with a dedicated tool (viral.app)Only one of 5-6 tools tried that worked for them