Blog · Finance metrics and formulas · Financial services
Fintech acquisition costs depend on product type, onboarding expense and what counts as an activated customer. Compare a fully worked acquisition cohort, reconcile the average and learn how to assess an outside benchmark.
Average CAC for fintech is useful only when acquisition spending and the definition of a new customer match the comparison. Divide the agreed cost pool by unique newly activated customers, then separate product types and channels. A cheap registration and a profitable customer relationship are different outcomes.
Activated-customer CAC = Acquisition spending assigned to a cohort / Unique new customers activated within its observation window
The general customer acquisition cost guide explains paid, blended and fully loaded CAC. The fintech question adds another boundary: how far through onboarding does a person or business have to get before being counted?
A consumer payment product might require a first completed payment. A business payment product might require a first settled merchant transaction. A lender might count a first funded loan. Write down the milestone, rather than letting an export's default account count define the measure.
For bank accounts, identification is a separate onboarding consideration. FinCEN's interagency customer identification guidance describes banks' risk-based identity verification procedures. That guidance concerns banks; it does not make every fintech product's onboarding process identical. An application, verification and first use should remain separate fields in the acquisition analysis.
The spending file needs expense date, campaign or acquisition cohort, product, channel, cost category and amount in USD. Include the relevant acquisition staff and agency allocations, incentives and onboarding checks under a documented policy. Separate ongoing customer service and retention spending.
Andreessen Horowitz identifies omitted referral fees, credits and discounts as a common CAC measurement problem. If an incentive already reduces revenue, reconcile its treatment so the same economic cost is not also charged again when calculating the cohort's later contribution.
The customer file needs a stable customer identifier, first registration date, prior-customer flag, acquisition channel, first product, verification date and activation date. Match spending to the acquisition cohort rather than simply dividing this month's expenses by this month's activations.
Choose a complete observation window. In the example below, every January 2026 registration gets 30 calendar days to activate. Reporting waits until the last January registration has that full window. Later activations can be reported separately without silently changing the original result.
A fictional US payments company acquires consumer and small-business customers. Its first-product groups are mutually exclusive: each new customer appears once, and existing customers adding a product are excluded. These figures illustrate the method; they are not an industry survey or a market average.
| Acquisition cost category | Consumer product (USD) | Small-business product (USD) | Total (USD) |
|---|---|---|---|
| Media | 48,000 | 18,000 | 66,000 |
| Referral incentives | 12,000 | 6,000 | 18,000 |
| Acquisition staff and agencies | 24,000 | 12,000 | 36,000 |
| Initial verification and onboarding | 6,000 | 6,000 | 12,000 |
| Total assigned acquisition spending | 90,000 | 42,000 | 132,000 |
The onboarding row includes costs incurred for applicants who never activate. Removing those costs would hide part of what the company spent to produce the successful activations.
The customer export supplies the next table's counts. Each stage is a subset of the preceding stage, and both products use the complete observation window described above.
| First-product group | Registered customers | Verified customers | Activated customers | Activation / registration |
|---|---|---|---|---|
| Consumer | 1,200 | 900 | 600 | 50.0% |
| Small business | 240 | 200 | 150 | 62.5% |
| Total unique customers | 1,440 | 1,100 | 750 | 52.1% |
Consumer activation is 600 / 1,200 = 50.0%; small-business activation is 150 / 240 = 62.5%. The total is 750 / 1,440 = 52.1%, rounded to one decimal place. Dividing the spending above by each stage gives very different acquisition figures:
| First-product group | Spend / registered customer (USD) | Spend / verified customer (USD) | Spend / activated customer (USD) |
|---|---|---|---|
| Consumer | 90,000 / 1,200 = 75.00 | 90,000 / 900 = 100.00 | 90,000 / 600 = 150.00 |
| Small business | 42,000 / 240 = 175.00 | 42,000 / 200 = 210.00 | 42,000 / 150 = 280.00 |
| Blended | 132,000 / 1,440 = 91.67 | 132,000 / 1,100 = 120.00 | 132,000 / 750 = 176.00 |
Dollar ratios are rounded to cents after calculation. The same spending supports a registered-customer figure of $91.67 and an activated-customer CAC of $176.00. Neither number changed because advertising became more efficient; the denominator changed.
Reconcile the cost categories to the acquisition expense schedule and the customer counts to distinct identifiers. Then multiply each activated-customer CAC by its customer count:
600 × 150 + 150 × 280 = 90,000 + 42,000 = 132,000
Blended CAC = 132,000 / 750 = 176
The simple product average would be (150 + 280) / 2 = $215.00. That gives the smaller business cohort the same influence as the larger consumer cohort. The guide to weighting ratios explains why a roll-up must use its underlying denominator.
Keep product additions separate too. A customer's second account can improve the relationship's economics, but it does not create a second newly acquired customer. Use products per customer for that expansion question.
Record the publication date, measurement period, geography, customer type, product and activation milestone beside every quoted average. Then check whether it includes staff, referral rewards, unsuccessful applications and onboarding expense. A dated company disclosure describes that company under its definitions; it does not establish an industry-wide CAC.
Also ask whether the statistic is total spending divided by total customers, an unweighted mean of company CACs or a median company result. Those answer different questions. If the source does not disclose enough to reconcile its boundaries with yours, retain it as context rather than a budget target.
The worked cohort supports an operational decision. Holding consumer spending at $90,000, raising activations from 600 to 750 would lower that product's CAC from $150.00 to $120.00: 90,000 / 750. The reduction is (150 - 120) / 150 = 20.0%. This is a sensitivity calculation, not a prediction that removing an onboarding step will produce those customers.
Growth owns channel spending and attribution. Operations owns the onboarding milestones. Finance owns the cost policy, reconciliation and final ratio. Judge the acquired customers against their subsequent contribution; the LTV to CAC ratio guide covers that separate comparison.
Bring your acquisition spending file and customer export with registration and activation dates to Covirage. Covirage maps the columns, and its tools compute and check the spending, customer counts and cohort ratios; 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.
See Covirage for financial services teams for the supporting industry context.
A good CAC is one that the acquired cohort's contribution can recover within the company's chosen payback period. Set that limit using processing costs, losses, servicing expense and observed customer activity. A competitor's acquisition figure cannot establish your spending ceiling.
Include acquisition-related verification costs when reporting an acquisition measure that covers onboarding, and label that boundary. Keep recurring monitoring and ongoing servicing separate. If another benchmark excludes onboarding, reconcile the difference before comparing the two figures.
Only if the account and customer definitions coincide. One customer may open several accounts or add another product later. Deduplicate the customer identifier, exclude existing relationships and show product additions separately from newly acquired customers.
Use the comparison to ask questions about acquisition methods, rather than to set a direct target. A funded borrower and an actively transacting payment customer pass different milestones, earn different revenue and carry different costs after acquisition.