At a glance
| Field | Detail |
|---|---|
| What | Decide whether cohort evidence is ready to inform an app budget change. |
| Who it matters to | Google Ads app buyers reviewing retention and revenue after acquisition. |
| Where | AppsFlyer or Adjust cohort reports, alongside Google Ads App campaign reporting. |
| When it applies | When comparing acquisition groups or investigating weak recent returns. |
| Action | Match cohort age, population, and revenue definitions; then review conversion delay. |
Google Ads cohort analysis should compare users at the same post-install age before it informs a budget change. Our suggested review pairs acquisition-date cohorts with Google's conversion-delay checks. The decision is whether the available evidence supports a budget review or whether the comparison needs more time or a corrected reporting definition.
Define the population before reading the curve
AppsFlyer's Cohort and retention dashboard groups users by acquisition date and supports campaign or media-source comparisons. It offers separate acquisition, retargeting, and unified views. Its retention metric measures app opening, while revenue measures reported revenue. These answer different questions.
For an acquisition review, we suggest recording the selected population and business outcome beside the report. Avoid treating an app-opening signal as evidence of purchase value. Keep the comparison focused on the outcome that prompted the budget question, and use retention to add context.
Adjust's Cohort analysis documentation adds an important population check: its widget defaults include installs and reattributions, click and impression attribution, and mixed attribution and network spend sources. Revenue metrics also require a revenue-type selection. Review those settings before interpreting a report as a view of newly acquired paying users.
Choose the report that answers the question
The approaches below serve different purposes. Their capabilities come from the linked documentation; the checks are our suggested review steps.
| Approach | Source-backed capability | Limitation | Suggested check |
|---|---|---|---|
| AppsFlyer complete-data view | Excludes unfinished cohort periods | Recent periods can be unavailable | Compare groups at a common completed age |
| AppsFlyer cumulative or on-day view | Shows accumulated results or activity within a selected period | The views describe different quantities | Label the aggregation beside each curve |
| Adjust revenue cohort widget | Supports revenue-type selection and population filters | Defaults include installs and reattributions | Confirm acquisition status and revenue scope |
| Google Ads conversion-lag reporting | Estimates how CPA or ROAS may change as conversions arrive | Forecast information is an estimate | Keep estimates distinct from observed cohort revenue |
For definitions and reporting-basis checks, use the measurement data-path review to document the event and report being compared.
Check maturity and the meaning of a period
Adjust's How cohorts work explains maturity in terms of whether users have completed the selected period. An immature cohort can have fewer eligible users at later ages. Its revenue metrics also distinguish accumulated revenue from revenue restricted to users who have completed the relevant period.
Our recommendation is to record maturity alongside the metric name. A curve can change because the included population changes; investigate that possibility before interpreting its shape as a change in buyer quality. If the chosen horizon is unfinished, retain the early read as provisional and schedule another review.
Period labels also need reconciliation. AppsFlyer bases daily, weekly, and monthly periods on calendar boundaries; Adjust defines daily periods relative to install or reattribution. Matching a label alone does not establish matching observation time. Check each report's clock before comparing exported curves.
Review the campaign's conversion delay separately
Google's conversion-lag documentation lists App campaigns for installs and engagement as eligible. It explains that delayed conversions can make cost per acquisition, or CPA, appear inflated and return on ad spend, or ROAS, appear deflated. Its forecasts estimate how those measures may change.
Treat this as a separate check from cohort maturity. Our suggested workflow asks whether users have completed the selected observation period and whether the campaign's conversion results are still developing. A forecast should remain labeled as a forecast when discussed alongside observed revenue.
Google's App campaign conversion-window guidance says conversion timing varies by action and buying cycle. It provides Days to conversions to help guide window selection, and describes a tradeoff: longer windows capture more conversions but can include less incremental organic conversions; shorter windows can miss incremental conversions.
We suggest recording the conversion action and window before escalating a performance discrepancy. Do not choose a window merely to make reports agree. A common budget-change threshold is not specified in the cited documentation; define the business decision and acceptable evidence before acting.
Make the budget decision reviewable
For a budget review, we suggest keeping a compact decision record: the campaign being assessed, acquisition dates, target event, selected cohort horizon, revenue scope, and remaining reporting gaps. State what would justify reconsidering the budget and what would trigger another measurement check. Keep the proposed action conditional on the matched comparison.
If retention and revenue point in different directions, review them as separate observations. Ask whether the intended outcome is repeat app use or monetization, and whether the revenue view covers the same users. For a revenue objective, require revenue evidence in the decision rule. For a recent shortfall, complete the maturity and delay checks before proposing a reallocation.
What to review
- Confirm whether the question concerns acquisition or returning users, and check the report's population filters.
- Match completed cohort age, period boundaries, and cumulative or on-day aggregation before comparing curves.
- Check revenue type and the spend source used in the comparison.
- Review Google Ads conversion delay and keep forecast values labeled separately.
- Document unresolved differences and the evidence needed at the next review before proposing a budget change.



