Docs

Passkey Login Funnel Tracking for Mobile Apps

Passkey enrolment is now a real login path, not an experiment — track activation, usage, and error rate to see if it's actually replacing passwords.

Engagement

FIDO’s 2026 State of Passkeys report puts consumer awareness at 90%, with 75% of people enabling a passkey on at least one account — and the companies that rolled passkeys out well are reporting startling numbers: Uber found over 90% of its passkey enrolments came from timely in-line nudges rather than account-settings pages, and eBay saw a 102% jump in adoption after switching to auto-triggered biometric prompts. Passkeys have moved from a security nice-to-have to a login path that measurably changes conversion. Teams still measuring login success as a single blended rate are missing the fact that passkey and password logins now behave like two different funnels with two different failure modes.

The risk in not tracking this separately is that a genuinely broken passkey flow — a device that doesn’t sync via iCloud Keychain or Google Password Manager correctly, a fallback that silently drops users back to a password field they’ve forgotten — gets absorbed into an overall login success rate that still looks fine because most users are still on passwords. By the time passkey adoption is high enough for a broken flow to visibly move the aggregate number, you’ve already lost weeks of enrolments and forced an unknown number of users into a password reset flow they didn’t need.

Data Points to Track

  • Passkey prompt shown vs. passkey created, to calculate activation rate separately from whether the prompt was even surfaced to the user
  • Login method used per session (passkey, password, password_with_2fa, social), so the login success rate can be broken down by method rather than blended into one figure
  • Passkey login error rate and error type (device sync failure, biometric mismatch, platform authenticator unavailable), since a rising error rate on a specific device or OS version is the earliest signal of a broken integration
  • Prompt placement and trigger (account settings, in-line at login, post-signup popup), matching the pattern that Uber, eBay, and TikTok all found the placement — not the feature itself — drives adoption
  • Fallback rate from passkey attempt to password entry, to catch users who tried and abandoned the passkey path rather than never attempting it

Setup Steps

  1. Instrument passkey enrolment as a funnel of its own — prompt shown, prompt accepted, passkey created, first passkey login completed — rather than a single boolean flag on the user record.
  2. Tag every login event with the method actually used, not just whether the login succeeded, so passkey and password success rates can be compared directly.
  3. Capture structured error codes from the platform authenticator API rather than a generic “login failed” event, since passkey failures need a different fix than password failures.
  4. A/B test prompt placement using the funnel data, following the pattern of in-line nudges outperforming settings-page prompts, and measure activation rate per placement rather than assuming one placement works for all user segments.
  5. Alert on passkey error rate by device or OS version, since a spike isolated to one platform combination usually means a sync or platform authenticator regression, not a broad UX problem.

Actionable Insights

If passkey activation is high but the passkey login rate stays low, users are creating passkeys but not actually using them to log back in — check whether the passkey is being offered as the default option on return visits or whether users are defaulting back to a saved password out of habit. A rising fallback rate from passkey attempt to password entry on a specific device model is a strong signal of a sync integration bug, not user preference, and should be treated with the same urgency as a crash spike. And once passkey login share crosses roughly a quarter of daily logins — the level that tends to correspond to a healthy 35–40% enrolment rate — it’s worth re-evaluating whether password reset volume and support load are dropping in line with that, since that’s the actual return on the investment.

Expert help

Need help tracking this in your app?

Our team sets up analytics pipelines for mobile and web teams every day. Talk to us and get your first events flowing in under an hour.

Talk to an expert