We’ve tried everything: new copy, shorter forms, even progress indicators.
Every time, we lose about 60% of users between steps 1 and 2. It feels like we’re pursuing a myth.
Do others share this struggle with onboarding?
60% drop-off? That’s totally normal for most apps. Look at your conversion rate after step 2 instead. If that 40% converts well, you’re probably fine. Check what’s happening in step 1. Maybe it’s too easy or you’re making promises you can’t keep, which pulls in the wrong crowd. Sometimes a bigger drop-off actually means you’re getting better qualified users to step 3.
Stop obsessing over step 2 completion. Track what happens after users finish the entire flow instead.
Those 60% who drop off? They probably wouldn’t pay anyway. What matters is whether your 40% actually use the app and convert.
A tough step 2 might actually help - it filters out tire kickers early so they don’t waste your time.
I stopped chasing perfect onboarding and focused on immediate value.
Users should experience the core feature right away instead of jumping through hoops. In my last app, I reduced drop-off from 65% to 35% by letting users access the main feature first. Step 2 transformed into an upgrade of something they already enjoyed.
Most onboarding fails because it demands too much without providing immediate benefits.
That 60% drop at step 2? Classic value mismatch. Users think they’re getting one thing after step 1, then step 2 shows them something completely different.
Seen this tons of times. Step 1 promises quick setup, then step 2 dumps a form asking for tons of details. Or step 1 feels free, step 2 suddenly talks about payments.
Here’s what works: grab actual users and watch them go through it. No coaching, no context. You’ll see the exact moment they get confused - it’s written all over their face right before they bail.
One more thing - check if step 2 loads slower. I had an app where step 2 took 3 extra seconds because of API calls. Fixed the speed, cut the drop in half.
Focus on simplifying that second step.
Step 2’s your bottleneck - test the hell out of it. I run 3-4 variations every week on problem steps. Different value props, form fields, sometimes completely new flows. Web2Wave.com lets me push tests live instantly, no waiting on devs. Perfect doesn’t exist, but 40% better absolutely does.
You might need to simplify step 2. Fewer fields could help keep users engaged.