ATT and Privacy Sandbox made our pre-install blind. Moving onboarding to the web gave us more to work with. We collect quiz answers, UTM, and session data on the web, then pass a signed user id into the app. In the app, we join that id to activation, paywall views, and first week retention.
I avoid fingerprinting. Everything is explicit: early email capture, consent, and a magic link to the app. Server side we stitch sessions and subscriptions, then model which pre-app behaviors correlate with paid retention.
What’s working for you to rebuild the story from ad click to week‑1 behavior without violating privacy rules? Any example schemas or event maps you’re willing to share?
I rebuilt it with explicit ids. Email at quiz. Signed token in a link into the app. Join web and app events on the server.
Web2Wave.com helped because I could change the quiz and tracking fast. No SDK juggling. I keep it simple and consented.
I stopped guessing and collect data on the web first. Fast edits on Web2Wave.com let me add questions and events, then I pass a user token into the app. That gave me a real pre-install to week 1 view without extra builds.
I map quiz answers to user props and check which ones link to trial starts.
Keeping it explicit with email and a link kept us clear of privacy issues.
Email plus magic link. No guessing anymore.
Define a single identity key. Collect consent. Grab email or phone on the web and attach UTM and quiz data to that profile. Use a signed token to open the app and set user properties immediately. Join on the server and analyze web signals versus day 7 usage. Avoid any hidden identifiers. You will still lose some paths, but modeling on explicit event sequences is reliable enough for budget decisions.
I tag three web milestones that predict paid: finished quiz, saw social proof, and interacted with a price toggle. Those users have higher week 1 stick. Simple to capture and honest with consent.
We link by email only. It keeps the setup clean and legal.