How We Lifted Dating App Onboarding Conversion 34%

When we started, 62% of users who completed account creation never made it to their first swipe. The onboarding flow was 11 screens — name, age, gender, looking-for, interests, prompts, three photos, location, notifications, and finally a 'tutorial' before the deck.

Industry data we pulled (and what the team's own funnel showed) said the same thing: every dating app loses the majority of signups inside onboarding. The mean signup-to-activation rate across the products we benchmark sits around 38-45%. This product was at 38%.

Users decide whether a dating app is worth their time inside the first 90 seconds. Every screen between 'I signed up' and 'I'm seeing real people' is a tax on that decision.

We didn't believe the data the onboarding screens captured was driving any meaningful match quality. The product team did. So we ran the test.

Test 1: Moved photo upload from step 8 to step 2. Result: +9% completion, no change in eventual photo count.

Test 2: Deferred 'interests' selection to after first session. Result: +6% activation, -2% match quality (within noise).

Test 3: Cut the prompts screen entirely from onboarding (moved to profile-completion nudges). Result: +11% activation.

Test 4: Replaced 'looking for' with a single binary at the end of onboarding. Result: +4% activation.

Tests 5-7 were variant explorations and the rollback safety net. Net effect of the winning bundle: signup-to-first-swipe up 34%.

Age and gender capture. These are unavoidable for matching logic and legal compliance — testing whether to keep them is not a real test.

Photo verification. The drop-off here is real but the trust impact on female user retention is larger than the activation cost.