Compare acquisition spending with a defined customer cohort and a stated cost boundary. Align attribution, eligibility and outcome windows, separating immature cohorts from completed follow-up. Dividing this month’s spending by unrelated customers recorded this month can create a misleading acquisition measure.
Define the acquired customer and the cost boundary
Specify the event that makes a person or account acquired: first paid order, activated account or another agreed milestone. Deduplicate according to that unit and keep returning customers separate if the measure concerns new acquisition. A lead is not necessarily a customer. Record eligibility dates and how cancellations, test accounts and refunds affect the cohort.
List included costs, such as media, agency fees or campaign production, and explain allocation of shared work. A media-only measure can be useful when labelled; it should not be called fully loaded acquisition cost if other relevant spending is excluded. Retain the currency basis and cut-off. Estimated or accrued costs should be distinguishable from confirmed amounts so a later invoice does not look like an unexplained performance change.
Align spending, attribution and outcome timing
A campaign can generate customers after the spending month, and customers can interact with several channels. Define the attribution window and rule without presenting attribution as proof of causation. Last-touch, first-touch and experiment-based estimates answer different questions. Keep unattributed customers visible and avoid assigning them to a preferred channel merely to make the totals reconcile.
Choose a cohort based on the decision, then allow the relevant outcome window to mature. Acquisition cost per customer, early retention and realised contribution are different measures with different clocks. A new cohort's revenue after seven days cannot be compared directly with an older cohort's revenue after ninety. Show maturity and incomplete follow-up rather than treating all recorded outcomes as equally final.
Interpret the result alongside quality and uncertainty
Inspect customer fit, refunds, repeat use and the defined downstream outcome before choosing a spending response. A low acquisition cost can coexist with poor retention, while a high cost may reflect a different product or customer need. These possibilities require evidence, not a narrative invented to defend a channel. Compare like cohorts and explain material changes in offer, audience or measurement.
Document the remaining allocation and attribution uncertainty. Sensitivity analysis can show whether a plausible cost allocation changes the ranking, but it does not establish a channel's incremental effect. Assign data recovery or a prospective measurement plan where the decision depends on that effect. The output should support a bounded operating choice and a review date, without claiming guaranteed returns from repeating historical spending.
A fictional campaign cohort
A fictional scenario to explain the method, not a customer result or a market benchmark.
A fictional service assigns 6,000 currency units of media spending and 2,000 of campaign production to a defined spring campaign. Within its stated thirty-day attribution window, 80 eligible first-time paying accounts are recorded, including 20 that pay in the following calendar month. The cohort acquisition cost under this boundary is 8,000 divided by 80, or 100 per account.
A calendar-month shortcut would divide 8,000 by only the 60 accounts paying in the spending month and show approximately 133.33, despite excluding the later attributed accounts. The analyst labels the cohort rule, keeps the attribution assumption visible and separately shows unattributed accounts. At the review date, only 50 of the 80 accounts have completed ninety-day follow-up, so retention is reported for that mature subset with its count. The example does not prove that the campaign caused all 80 acquisitions or that future spending will acquire customers at the same cost. The review also distinguishes a later correction to campaign spending from a genuine change in customer outcomes. If a shared production cost is allocated differently, the revised figure receives a version note instead of being interpreted as new evidence that acquisition suddenly became less efficient.
TAKE THIS INTO YOUR NEXT CONVERSATION
A practical next-step checklist.
- Define the acquisition event and unique customer unit.
- State included costs and allocation rules.
- Align attribution and follow-up windows.
- Keep unmatched customers and immature outcomes visible.
Common questions.
Is attributed acquisition the same as incremental acquisition?
No. Attribution assigns credit under a rule; incremental effect requires a defensible counterfactual design.
Can I compare ninety-day outcomes for a one-month-old cohort?
No. Show its maturity and wait for comparable follow-up or use a clearly labelled model with assumptions.
Further reading & source context.
These primary references provide background. They are not evidence for the fictional example above and do not endorse Accuralytix.
Found an error or a missing qualification? Send an editorial correction.
Check my data readiness.
Use a practical tool to turn this idea into a clearer starting point.
Check my data readiness