Passkey Enrollment Rate

Passkey Enrollment Rate (PER) measures how often users who are offered a chance to create a passkey actually complete enrollment. It is the gateway metric for passkey adoption because enrollment is the moment users get access to faster sign in and stronger account protection.

We measure Passkey Enrollment Rate across the enrollment funnel that starts when a user is shown an enrollment offer, and ends when the passkey is successfully registered to the user account. Measurement boundary: from Passkey Offer shown to Passkey enrolled within the chosen time window.

flowchart LR
    O((Offer
    Passkey))
    V((Offer
    Viewed))
    S((Start
    Enrollment))
    C((Create
    Credential))
    R((Register
    Server))
    E((Passkey
    Enrolled))

    O --> V --> S --> C --> R --> E

    O -. "PASSKEY ENROLLMENT RATE" .-> E

Get the Authentication Analytics Whitepaper

Analyze your Authentication Performance with Real Numbers

10 KPIs that connect authentication performance to revenue. Track adoption, friction & conversion impact.

Download the Authentication Analytics Whitepaper

Work email *

Name *

Company *

Job title *

Get Whitepaper

We calculate Passkey Enrollment Rate per unique user. Each user is counted at most once in the numerator and denominator for the selected window, for example ever enrolled or enrolled within 30 days of first offer.

Enrolled Users is the number of unique users who completed passkey registration. Users offered Passkey Opportunity is the number of unique users who were actually shown an offer that could lead to enrollment.

Count a user when Passkey Enrollment completed is logged and the credential is registered to the account. If a user enrolls on multiple devices, we still count one enrolled user. Do not count starts, client side creation without server registration or failed attempts.

Count a user when Passkey Offer shown is logged, meaning the user could realistically proceed. Do not count users who were only eligible in theory, users on unsupported devices or users who never saw the offer due to rendering or delivery issues.

We use Passkey Enrollment Rate to decide where to invest in prompting, UX and eligibility controls so more users actually enroll.

We can improve the following business outcomes:

• Higher successful sign ins Diagnose low enrollment in high value cohorts, change prompt timing and copy, validate via cohort lift in enrollment within 14 to 30 days
• Lower user drop off during enrollment Diagnose drop between Passkey Offer Shown and Passkey Enrollment Started, change offer design and perceived value messaging, validate with higher start rate and stable completion rate
• Lower support contacts Diagnose failures after Passkey Enrollment Started, change client error handling and recovery messaging, validate with fewer technical_error failures and fewer repeated starts per user
• Better operational cost Diagnose repeated retries and timeouts, change performance and retry strategy for registration calls, validate with lower median enrollment duration and fewer duplicated events
• Lower fraud and abuse exposure Diagnose low enrollment in risky segments, change where we prioritize enrollment offers, validate by raising enrollment in those segments without increasing suspicious enrollment failures

• Intent bias: Users who see the offer are not random, especially if we only prompt after certain behaviors, so the rate can look healthier than true population readiness
• Missing telemetry: If Passkey Offer Shown is not logged consistently across clients, the denominator shrinks and the rate inflates
• Mix shift: A release that changes traffic toward more mobile users, more returning users, or more regions with higher device support can move the KPI without any UX improvement
• Window confusion: Ever enrolled and enrolled within 30 days answer different questions, mixing them breaks trend interpretation

At minimum, we report Passkey Enrollment Rate by prompt context, device type, platform and browser, new versus existing users, and by week of first offer so we can separate product changes from cohort mix changes.