Which acquisition channels bring users who activate?
Compare channels on the same eligible population, activation definition, and completion window. A channel with more visits can produce fewer useful customers if its visitors do not finish the product's first valuable task. Keep traffic, completed signups, activation, and settled payments distinct until the recorded identities and events connect them.
Use this guide when search, social, or campaigns bring people to your product but you cannot yet decide where to spend the next hour. The decision should serve your selected objective, whether that is qualified activation, paid conversion, retention, or another measured outcome.
Start with the evidence you have
Connect the sources relevant to the question through Data Sources. Web tracking supplies arrival and campaign evidence; PostHog can supply product events; Stripe supplies billing activity. Which joins are possible depends on recorded identity, privacy settings, resource scope, historical coverage, and the connector's declared capabilities.
Choose the existing completion event that expresses value in your product. Record its meaning and the entry population in the project's measurement definition. Keep optional setup steps out of a required funnel unless they are actually necessary for that population.
Make the channel comparison reproducible
Retain these items with the source reports or receipts:
| Item | Why it changes the answer |
|---|---|
| Entry event and dates | Visitors, new accounts, and trial starts are different populations |
| Channel definition | First arrival, latest arrival, and campaign credit answer different questions |
| Activation event | A button click is different from completing a useful task |
| Completion horizon | Recent entrants may still have time to activate |
| Identity coverage | Unlinked anonymous visits cannot become known customer journeys |
| Source scope and freshness | Separate sites, accounts, and partial imports cannot be silently combined |
| Payment lineage | A billing record needs an observed customer relationship to support conversion |
Preserve unclassified traffic. A missing referrer is not evidence that someone typed your URL; privacy tools and untagged links can remove referral context. An aggregate channel report also cannot establish every individual's path. See measurement methodology before calculating a conversion rate across sources.
Ask Scout to investigate the difference
Ask: “Which acquisition sources bring new users who reach our recorded activation outcome? Preserve my objective. Compare mature entrants with the same completion window, show eligible users and completions by source, and state the attribution method and unclassified coverage. Inspect one relevant landing page and one product step before proposing a change.”
Inspect the existing metric rows and Details references. Check whether a small channel is merely noisy, whether an apparent decline comes from an incomplete cohort, and whether a source change coincides with a different audience or offer. Have Scout distinguish a recorded observation from a proposed explanation.
A mismatch can occur before or after signup. A landing page may attract the wrong intent, while a setup flow may prevent suitable users from reaching value. Prepare one change tied to the evidence: a clearer page promise, a more useful answer, or a correction to a required product step. Inspect the real page or screen before drafting it.
Worked arithmetic with fictional data
Suppose two fictional channels each produce 100 recorded new accounts with a full completion window. Search has 30 activated accounts; a newsletter has 20. Their observed activation rates are 30% and 20%, a difference of 10 percentage points. That does not prove search caused better activation, establish revenue, or justify a budget move: audience, timing, and offer may differ, and the samples need an appropriate uncertainty assessment.
These figures illustrate the denominator only. Use your own source records for the baseline; do not treat them as customer results or target rates.
Measure a reviewed change
Record the changed page or flow and the date it became available. Compare new eligible entrants using the same event, identity scope, source definition, and completion horizon. Review the downstream outcome and applicable guardrails, such as retention, refunds, or acquisition cost. More traffic or more tasks completed is not automatically progress toward the objective.
The trial activation workflow helps investigate product friction. The Search Console workflow helps investigate search intent and clicks. Analytics access can support analysis and a draft; publishing and spending require their own authority.