Skip to main content

Sign-in Factor Share of Successes

Each sign-in factor’s share of successful second factors, including passkey share.

Each sign-in factor’s share of the second factors that succeeded over a selected reporting period. The fido2 row is the passkey share: of the people who cleared a second factor at sign-in, how many did it with a passkey.

This is the measure to use for passkey uptake. It compares like with like, a completed second factor against a completed second factor. The share in Sign-in MFA Activity by Factor does not: it counts an SMS code as soon as it is sent, entered or not, but a passkey only once the prompt reaches an outcome, so it understates passkeys.

It covers the factors the public can sign in with: sms_otp, voice_otp, and fido2. email_otp is a sign-up check, not a sign-in factor.

Calculation

Restrict to the sign-in factors and to result = 'success'. Sum count by mfa_type, then divide each factor’s successes by all sign-in successes.

For a trend, do the same per Monday-to-Sunday week, and leave off any week the reporting window cuts short: a part week has too few days to compare.

By factor:

SELECT
    mfa_type,
    SUM(count) AS successes,
    CAST(SUM(count) AS DOUBLE)
        / NULLIF(SUM(SUM(count)) OVER (), 0) AS share_of_successes
FROM ibm_verify_events_transformed.events_mfa_activity
-- set the reporting window; see the timespans note in cookbook/notes.qmd
WHERE CAST(from_date AS DATE) BETWEEN DATE '2026-01-01' AND DATE '2026-05-28'
  -- the factors the public can sign in with; email_otp is a sign-up check
  AND mfa_type IN ('sms_otp', 'voice_otp', 'fido2')
  AND result = 'success'
GROUP BY mfa_type
ORDER BY successes DESC;

Passkey share by week:

SELECT
    date_trunc('week', CAST(from_date AS DATE)) AS week_start,
    SUM(count) FILTER (WHERE mfa_type = 'fido2') AS passkey_successes,
    SUM(count) AS successes,
    CAST(SUM(count) FILTER (WHERE mfa_type = 'fido2') AS DOUBLE)
        / NULLIF(SUM(count), 0) AS passkey_share
FROM ibm_verify_events_transformed.events_mfa_activity
-- whole Monday-to-Sunday weeks only: start on a Monday, end on a Sunday
WHERE CAST(from_date AS DATE) BETWEEN DATE '2026-01-05' AND DATE '2026-05-24'
  AND mfa_type IN ('sms_otp', 'voice_otp', 'fido2')
  AND result = 'success'
-- date_trunc('week', ...) anchors each bin to its Monday
GROUP BY date_trunc('week', CAST(from_date AS DATE))
ORDER BY week_start;

By factor:

tbl(con, in_schema("ibm_verify_events_transformed", "events_mfa_activity")) |>
  # set the reporting window; see the timespans note in cookbook/notes.qmd
  filter(
    as.Date(from_date) >= as.Date("2026-01-01"),
    as.Date(from_date) <= as.Date("2026-05-28"),
    # the factors the public can sign in with; email_otp is a sign-up check
    mfa_type %in% c("sms_otp", "voice_otp", "fido2"),
    result == "success"
  ) |>
  group_by(mfa_type) |>
  summarise(successes = sum(count, na.rm = TRUE)) |>
  mutate(
    share_of_successes =
      as.double(successes) / sum(successes, na.rm = TRUE)
  ) |>
  arrange(desc(successes))

Passkey share by week:

tbl(con, in_schema("ibm_verify_events_transformed", "events_mfa_activity")) |>
  # whole Monday-to-Sunday weeks only: start on a Monday, end on a Sunday
  filter(
    as.Date(from_date) >= as.Date("2026-01-05"),
    as.Date(from_date) <= as.Date("2026-05-24"),
    mfa_type %in% c("sms_otp", "voice_otp", "fido2"),
    result == "success"
  ) |>
  # floor_date(..., "week") translates to Athena's date_trunc('week', ...),
  # which anchors each bin to its Monday
  mutate(week_start = floor_date(as.Date(from_date), "week")) |>
  group_by(week_start) |>
  summarise(
    passkey_successes =
      sum(if_else(mfa_type == "fido2", count, 0L), na.rm = TRUE),
    successes = sum(count, na.rm = TRUE),
    passkey_share =
      sum(if_else(mfa_type == "fido2", as.double(count), 0), na.rm = TRUE) /
        sum(count, na.rm = TRUE)
  ) |>
  arrange(week_start)

Interpretation notes

The shares are proportions between 0 and 1 and sum to 1 across the three factors. Passkeys (fido2) became generally available on 2026-07-06, so a window that starts earlier shows little or no passkey share. A share can rise because passkey use grew or because SMS use fell; read the share next to the successes behind it.

Alternative names

  • passkey share
  • passkey uptake
  • sign-in factor mix
  • share of completed second factors