Customer Care · prepared review
Understand the history before choosing the next retention question.
Compare observed value, customer age and missing coverage in a prepared example. NOGL does not run a churn model or contact customers on this page.
Extract cutoff (exclusive, UTC)
2026-04-01 · UTC
Supplied net order value
30 / 90 days
Future estimate
Not supplied
Descriptive arithmetic only; not lifetime value or a forecast.
Three decisions, different evidence
Make the question explicit.
Use purchase history to describe a supplied group, a defined future estimate to discuss a horizon, and permission evidence before considering contact. One does not establish the others.
Supplied net order value
Descriptive arithmetic only; not lifetime value or a forecast.
Defined future estimate
A future estimate needs its own target, issue date, horizon, model reference and validation context.
Marketing permission
A complete denominator does not prove consent, future demand, causality or effectiveness.
Prepare a comparable review
Start with source, identity and time.
Keep the extract version, cutoff, currency and refund definition visible. Check whether each customer has reached the selected age since their first order before comparing group value.
01
Fictional source
Extract version · Extract cutoff (exclusive, UTC) · EUR
02
Coverage
Identity · Window completed at cutoff · Marketing permission
03
Review question
A complete denominator does not prove consent, future demand, causality or effectiveness.
Explore observed customer value
Choose a group and an observed age window.
These six fictional members form three separate examples. Changing 30 to 90 days changes the observation requirement; it does not forecast the missing days or alter the source records.
Preparing review controls
- Extract cutoff (exclusive, UTC)
- 2026-04-01T00:00:00Z
- fictional:cutoff
- Extract version
- retention-review-v1
- fictional:version
Net order value after refunds attributed to the original order window; excludes tax and shipping. EUR only. This definition is authored for this example.
Fictional member A1
- First order (UTC)
- 2026-01-01
- fictional:A1:firstOrder
- Window ends (exclusive, UTC)
- 2026-01-31
- Window completed at cutoff
- Yes
- Identity
- Matched in this example
- fictional:A1:identity
- Coverage
- Complete within the stated window
- fictional:A1:coverage
- Supplied net order value
- EUR 100.00
- fictional:A1:net30Cents
- Currency
- EUR
- fictional:A1:currency
- Marketing permission
- Not supplied
Fictional member A2
- First order (UTC)
- 2025-12-01
- fictional:A2:firstOrder
- Window ends (exclusive, UTC)
- 2025-12-31
- Window completed at cutoff
- Yes
- Identity
- Matched in this example
- fictional:A2:identity
- Coverage
- Complete within the stated window
- fictional:A2:coverage
- Supplied net order value
- EUR 200.00
- fictional:A2:net30Cents
- Currency
- EUR
- fictional:A2:currency
- Marketing permission
- Not supplied
Read the denominator before the average
An unknown period is not zero revenue.
A group average is shown only when both supplied members have a completed window, matched identity, complete stated coverage and a known value. Otherwise the group result stays unavailable; known rows remain visible.
Complete eligible members: 2 / 2
Window completed at cutoff: 2 / 2
- Average denominator
- 2
- Group net order value
- EUR 300.00
- Average per supplied member
- EUR 150.00
Descriptive arithmetic only; not lifetime value or a forecast.
Sum of both supplied window values ÷ 2 members, only when all review conditions pass.
Choose an observed age window
First 30 days
Average per supplied member: EUR 150.00
Descriptive arithmetic only; not lifetime value or a forecast.
First 90 days
Average per supplied member: EUR 200.00
Descriptive arithmetic only; not lifetime value or a forecast.
A complete denominator does not prove consent, future demand, causality or effectiveness.
Define what a future estimate would need
Keep observation, prediction and permission separate.
A future estimate needs its own target, issue date, horizon, model reference and validation context. None is supplied by this cohort example. Marketing permission is unknown for every member, so no audience, offer or message is prepared.
Questions before a retention decision
Is this a customer lifetime-value prediction?
No. The example describes supplied net order values in completed 30- or 90-day windows. No model runs and no lifetime is extrapolated.
Why can a newer group have no 90-day result?
The observation window has not finished at the supplied cutoff. A missing future period is not a zero-value period.
Can this create a retention audience or campaign?
No. Identity, permission, business policy and appropriate evidence require review before any separate operational action.
Take the next review step
Bring the missing questions to the responsible team.
Discuss the required inputs, inspect the existing individual horizon example, or review connection scope. This page prepares a conversation; it does not approve or launch retention work.