Lifecycle Segmentation
Part of Designing customer onboarding journeys
Measuring onboarding completion by customer type
Define customer types at onboarding entry, set a useful completion result for each and report rates with clear units and observation windows.
Measure onboarding completion by defining each customer type at entry, specifying a useful result for that type and counting how many eligible people or accounts reach it within a stated window. Report types separately when their tasks or counting units differ; one overall rate can hide an unusable path.
Fix customer type at entry
Choose types that change the onboarding task. In a hypothetical workspace, a person creating a new workspace and someone invited into an existing one start in different situations. A pricing tier or organisation size belongs in the breakdown only if it changes the task or helps explain a meaningful difference.
Record the type when onboarding begins. Keep an “unknown at entry” group visible. If a person’s role changes later, retain their original entry type for that cohort and record the change separately.
Define a completion result for each type
Completion should describe a usable result, not the end of a checklist:
| Customer type at entry | Illustrative completion | Counting unit |
|---|---|---|
| Creates a workspace | First shared task completed | Workspace account |
| Invited into a workspace | Assigned task completed | Invited person |
| Needs assisted setup | Agreed setup outcome confirmed | Account |
The definitions are examples to validate against customer needs. A first shared task may not suit every workspace creator, and a support case marked closed may not prove assisted setup worked.
When types have different outcomes or units, show their rates side by side with clear labels, but do not rank them as though they measured an identical task.
Set the denominator and window
For each type, count the eligible units that entered during a defined period. The confirmed completion rate is the number of those units with a verified result inside the observation window divided by all eligible units in that entry cohort.
State whether cancelled accounts, duplicate signups and customers moved to an assisted path remain eligible. Set those rules before looking at the results.
Do not compare a recent cohort with an older one until both have had the same time to complete. Show an open window as pending.
If a completion event is late or missing, show that data uncertainty separately: a confirmed completion rate may understate actual completion, and an absent event does not prove failure.
A dashboard’s settings can alter the result. In Amplitude funnel analysis, segmentation by a user property and filters in the Segmentation Module apply to the first funnel step. A group-by filter in the Segmentation Module also applies only to the first event; Amplitude also allows a group-by filter for an event.
Amplitude also offers a configurable conversion window and a Unique Users metric. Check the entry step, property timing, unit and window before adopting a dashboard figure.
Interpret the differences
For each type, show eligible units, verified completions, the rate, and pending or uncertain records. Put counts beside percentages, especially for small groups.
Review customer histories and support cases before attributing a lower rate to the journey: another type may face an approval step, or its completion data may be incomplete.
If the completion definition changes, record a break in the series rather than presenting the new rate as an improvement.
Completion by type answers whether each group reached its defined result. Time to value and the effect of messaging require separate analyses.
