At a glance
- App category
- Sleep sounds
- Industry
- WellTech
- Channels
- in-app onboarding. No paid media in scope
- Period
- 2022 -
- Markets
- Worldwide, United States primarily
- Key results
- conversion to purchase +46.08% and ARPU +52.82%, from an A/B test of a rebuilt onboarding flow
The challenge
Deep Sleep Sounds wanted more conversions to purchase and higher ARPU, and the request arrived without a diagnosis attached: the team knew the numbers they wanted to move, not which part of the journey was holding them down.
That is the normal starting point, and it is why the first piece of work was not a redesign. A sleep app has a short, unusual window to convince: people open it tired, at night, intending to fall asleep rather than to evaluate a product. Whatever the app has to say for itself has to land inside that window.
What we did
1. Audited the product and the design before touching either. We went through the user flows end to end to find where intent was being lost. The onboarding carried the most weight and the least work, so that is where we concentrated.
Reasoning and learning: the audit is what makes the redesign cheap. Rebuilding the flow that is not the constraint costs the same as rebuilding the one that is.
2. Looked specifically at how personalised the onboarding was. The existing flow treated every new user the same way. For a sleep app that is a missed opening, because people arrive with different reasons: one cannot fall asleep, another wakes at three, another needs noise to cover a street outside. The same screens answer none of them specifically.
3. Rebuilt the onboarding flow from scratch, around personalisation. Applica designed the new flow and its visual design, then supported the implementation. The extra onboarding questions were the mechanism: they made the flow feel addressed to the person going through it, and they gave the app the information to behave that way afterwards.
Reasoning and learning: a longer onboarding is not automatically a worse one. Each extra screen costs attention and has to buy something back, and questions that visibly change what the user is shown buy more than they cost.
4. Tested it rather than shipped it. Deep Sleep Sounds ran an A/B test of the new flow against the old one on live traffic, on our recommendation, so the result is a comparison rather than a before and after.
Results
Against the previous onboarding, the new flow lifted conversion to purchase 46.08% and ARPU 52.82%. Both figures come from the same A/B test, with the extended, personalised flow as the variant.
+ 46%
conversions to purchase
+ 53%
to ARPU
Client feedback
Applica Agency is more knowledgeable in this area than most other providers we've used in the past.

Michael Brandon
CEO of Slumber Group
Key learnings
Personalisation earns the extra screens. The longer flow won because the questions changed what the user saw next, which is the only version of personalisation a user can feel.
Find the constraint before redesigning anything. The audit put the work in onboarding rather than in the paywall, and that choice is what the result rests on.
A/B test the flow, not just the screen. Onboarding is a sequence, and sequences compound. Testing it as one unit is what made the size of the win legible.
Conclusion
Deep Sleep Sounds did not need more traffic or a new offer. It needed its first two minutes to speak to the person going through them. Rebuilding onboarding around that lifted conversion to purchase 46% and ARPU 53% in a controlled test.
More on our conversion rate optimization and health and fitness app marketing.





