What counts as activation, which events stand behind it.
The plan records the metric and the tracking it needs.
Growth
Check the activation number before the growth review.
When an activation event is renamed, the funnel can stop counting it. Check the event definition before treating the chart as a change in behavior.
Free means one audit: sign in, connect a source, no card.
code
account_createdanalytics
signup_completedCheck the rename before treating a lower count as a change in behavior.
What happened
The rename works. The funnel reads the old name.
A valid code change can still change what an activation report counts.
Same signup. Different name.
For example, a team renames signup_completed to account_created while splitting a signup handler. The new name fires correctly. The activation funnel still reads the old one.
Skene maps what your product collects and names the signals that need attention. A missing event stays visible in the evidence.
code
analytics
plan
setup_completed
Verified
Give engineering the finding behind the concern.
Skene reviews tracking changes in a pull request. This real, redacted review shows a missing event, a suggested fix and the later passing review. It demonstrates the review mechanism; it is not an activation result.
Engineering can review the suggested fix in the pull request. The later review records whether the tracking issue is resolved.
check
1missing event
08:14 UTC
verdict
Passedlater review
09:20 UTC
Make the next change measurable first.
Define activation and its reporting window before launch. Verification checks the required events and fields; the after-launch Check computes the result separately.
plan
Trial activationtarget4.0%
required evidence
setup_completedaccount_id
Read the window before you explain the number.
A saved result can be stale. Check its reporting window and when it was computed before comparing it with the number in the review.
verdict
current4.1%code + data
previous 3.2%7 days
Ready
Who owns whatYou decide Engineering does At the growth review
Reviews the finding on the line that needs work.
The review identifies a tracking regression for the team to inspect.
Review what changed before you explain the number.
Read the metric alongside the findings behind it. A change in the number alone does not prove what caused it.