Growth marketing analytics: acquisition to retained customers

Growth analytics for marketing compares acquired customer cohorts through activation, repeat use and retained revenue. Keep each customer's original acquisition source, measure outcomes at the same customer age, and reconcile the results to the activity and invoice exports. Acquisition volume alone cannot show whether a source brings lasting business.

Define the cohort before comparing sources

Start with a customer register containing a stable customer ID, acquisition date and one acquisition source. Join the activity export and net invoice register using that ID. Retain unmatched records for investigation rather than assigning them to the closest name.

Choose an inclusion rule and a return rule. Google's cohort exploration documentation distinguishes the condition that admits someone to a cohort from the subsequent behavior being measured. Apply that distinction to your business file: acquisition fixes membership; later activity determines the outcome.

For this example, a US subscription business groups customers acquired in June 2026 by source. Activation means completing the agreed onboarding action within 14 days of acquisition. Repeat use means a qualifying service action during days 15–30. A retained customer has positive net invoiced revenue during days 61–90. These are separate observations, so repeat use does not require an activation flag.

The reporting cutoff is September 30, 2026, giving every June customer a complete observation window. These illustrative definitions are management choices, not universal retention benchmarks. For a sales-led acquisition process, keep the preceding lead stages in B2B marketing analytics.

Build customer flags from the exports

Create a Cohort sheet with customer ID in A, source in B and acquisition date in C. Use D for activation, E for repeat use, F for retention and G for later net revenue. Use real spreadsheet dates, with event timestamps converted to the agreed local date first.

On Activity, put customer ID in A, activity date in B and agreed activity code in C. The activation flag in Cohort!D2 is:

=--(COUNTIFS(Activity!$A$2:$A$9999,A2,Activity!$C$2:$C$9999,"activated",Activity!$B$2:$B$9999,">="&C2,Activity!$B$2:$B$9999,"<="&C2+14)>0)

For the repeat flag in E2, count qualifying events inside the later window:

=--(COUNTIFS(Activity!$A$2:$A$9999,A2,Activity!$C$2:$C$9999,"repeat",Activity!$B$2:$B$9999,">="&C2+15,Activity!$B$2:$B$9999,"<="&C2+30)>0)

Google Sheets documents COUNTIFS as counting rows that meet multiple criteria. Testing whether the count exceeds zero makes each customer contribute only one flag, even when the export contains several qualifying events.

On Invoices, put customer ID in A, invoice date in B and net revenue after credits in C. Compute G2, then F2:

=SUMIFS(Invoices!$C$2:$C$9999,Invoices!$A$2:$A$9999,A2,Invoices!$B$2:$B$9999,">="&C2+61,Invoices!$B$2:$B$9999,"<="&C2+90)
=--(G2>0)

SUMIFS adds values meeting the specified criteria. Fill the formulas down every customer row and extend the ranges if the exports are larger. Covirage is not affiliated with Google.

Work the source comparison

The illustrative customer flags and acquisition-cost register produce this table. Net revenue covers only days 61–90; cost is the agreed acquisition spending assigned to the June cohort, including any shared costs under a documented allocation.

Use the acquired population as the denominator throughout:

Activation rate = Activated customers / Acquired customers

Retention rate = Retained customers / Acquired customers

Cost per retained customer = Acquisition cost / Retained customers

Round percentages to one decimal place and dollars to cents after calculating. The total retention rate is 96 / 300 = 32.0%; the total cost per retained customer is 15,000 / 96 = $156.25. Divide totals rather than averaging source rates. This later cost measure complements customer acquisition cost, which uses newly acquired customers instead.

Prove the result before changing spending

The source counts must reconcile to the customer register: 120 + 80 + 100 = 300. The retention flags must sum to 36 + 40 + 20 = 96. Trace the $42,400 to eligible invoice and credit rows, and the $15,000 to the cost register. Check both amounts for unassigned sources and duplicated joins.

Partner has fewer acquired customers than Search but more retained customers and lower cost per retained customer. Investigate customer mix and onboarding practices before reallocating funds. The comparison does not establish that changing source would reproduce the difference. Have the acquisition owner and customer-success owner review the same exceptions; training account managers covers the handoff into ongoing customer work.

Where it goes wrong

• Unequal cohort ages. Recent customers have not had the same chance to return. Hold them outside a completed-window comparison.
• Changing source assignments. Later interactions must not overwrite the acquisition source used to form the original cohort.
• Counting events as customers. Repeated actions increase activity volume, not the number of people or companies retained.
• Treating revenue as profit. Later invoiced revenue excludes neither service costs nor all future obligations. It is not lifetime value or acquisition payback.

Do this on your own file in Covirage

Bring your customer register, dated activity export, invoice rows and acquisition-cost file to Covirage. Covirage maps the columns, its tools compute and check the cohort comparisons, and the external AI model explains the results with every figure traced to its rows. Create a free Covirage account and bring a sample of your own data; creating the account takes about a minute.

See Covirage for sales insights for the business context.