Definition problem
Registration measures completion of a registration flow. It does not establish that eligible supply appeared, demand saw it, or a useful outcome occurred.
Working definition:
the share of eligible sellers in a new cohort reaching a predefined useful-participation event within a stated window.
The substantive choices are eligible, useful participation, and window, not the division itself.
Event ladder
- registration;
- mandatory verification;
- first supply unit created;
- first qualified supply available for matching;
- first qualified exposure;
- first qualified response;
- first eligible match;
- first completed transaction or other terminal success;
- repeat activity in the next relevant period.
Not every marketplace needs every step. The ladder exists to prevent one event from silently substituting for another.
Selecting the activation event
Choose the event nearest to value while still supporting timely action. A first transaction is economically meaningful but may be late and demand-dependent. Listing creation is fast but may precede eligibility. A practical candidate is the earliest event that creates available supply and shows a stable relationship with a later outcome.
The team should influence the event, but a purely local event risks optimising form completion instead of the market. Pair it with a downstream market outcome such as qualified exposure, response, match, or completion.
Version eligibility rules and their effective dates. Otherwise a higher activation rate may merely reflect a weaker quality gate.
Formula and cohorts
Seller activation rate = sellers reaching the event in-window / eligible sellers entering the cohort
Keep sellers who do not reach the event in the denominator. Mark cohorts whose window is still open rather than mixing them with mature cohorts. For long cycles, publish time to activation as a distribution and keep censored observations visible.
Minimum metric system
- One primary seller activation rate.
- Transition diagnostics: time to first qualified listing, qualified exposure, qualified response, and time to first terminal outcome.
- Guardrails: listing-quality indicators, zero exposure share, unmatched supply share, transaction completion, and seller retention.
Guardrails reveal whether the primary measure improved while quality weakened or failure moved downstream.
Diagnostic matrix
| Observation | Working hypothesis | Next check |
|---|---|---|
| Created supply rises, qualified supply does not | preparation friction or quality | reasons for ineligibility by rule version |
| Qualified supply exists, exposure does not | wrong micro-market, availability, or allocation | candidate coverage and zero exposure |
| Exposure exists, response is weak | relevance, price, trust, or content | response and quality indicators by micro-market |
| Response exists, completion is weak | mutual acceptance, payment, or fulfilment | cancellation and failure taxonomy |
| First transaction occurs, repeat does not | one-off activation or weak seller economics | retention and repeat rate |
Each row is a hypothesis rather than a diagnosis.
Illustrative example
After a shorter form, registrations and created listings rise while qualified listings do not. Calling this an activation gain is premature. Eligibility, time to first qualified listing, exposure, and the later outcome must be checked first. No company or real value is represented.
Data contract before dashboard
The downloadable CSV specifies candidate events, properties, quality gates, exclusions, owners, and versions. It is a template, not a finished tracking plan.
Strongest counterargument
A common objection is that only the first transaction deserves to count as activation. It is a sound terminal test, but often a poor sole operating signal because it can be rare, late, and constrained by demand. An earlier proxy is acceptable only when a downstream outcome sits beside it. If that relationship disappears, replace the proxy.