Blog · Finance metrics and formulas
Average CAC for ecommerce needs a consistent cost definition and comparable customer cohorts. Compare acquisition spend with gross profit after returns and observed repeat purchases, using a worked channel example.
Average CAC for ecommerce is acquisition spending per new buyer, but a useful comparison also asks what those buyers earn after returns and repeat orders. A lower CAC can still leave a larger unrecovered acquisition bill. Compare equally mature cohorts with the same spending definition before treating any average as a target.
Average CAC = Included acquisition spending / Distinct new buyers acquired
Observed gross profit per acquired buyer = (Cohort net revenue - Cohort cost of goods sold) / Original acquired buyers
In the second formula, subtract cost of goods sold from net revenue before dividing. This is an observed gross-profit comparison, before fulfillment and other operating costs that your accounting policy puts below gross profit. It is not a forecast of lifetime profit.
A benchmark needs a label beyond "ecommerce." Paid-channel CAC may include media alone. A fully loaded measure can also include acquisition agency fees, creative work and an allocation of acquisition staff costs. Blended CAC includes all acquired buyers, including those without a paid-channel attribution, against the defined company acquisition budget.
Those are different measures. The general CAC calculation explains the common denominator; the useful question here is whether two reported averages describe the same costs and buyers.
The SBA's marketing and sales guide calls for a complete marketing cost breakdown and consistent measurement of effectiveness. Apply that discipline before comparing your figure with another retailer's headline number. An external benchmark without its cost definition is a prompt to investigate, not a spending limit.
Use a marketing expense export and an order export joined through your existing customer and channel assignments. Keep the following fields visible:
Look back beyond the current reporting period before flagging a buyer as new. A returning customer with a fresh email address should not automatically become a new acquisition. Keep unresolved identity matches in an exception list until the business owner resolves them.
For detailed order exclusions and channel inputs, use the separate ecommerce customer acquisition cost calculation. This comparison begins after those inputs have been reconciled.
A US home-goods retailer compares two acquisition channels. Every acquired buyer has a complete 180-day observation window. The following costs are attributed to acquisition, with no retention campaign expenses included. Shared acquisition staff costs are allocated using recorded time.
| Channel | Media (USD) | Agency and creative (USD) | Acquisition staff (USD) | Total acquisition cost (USD) | New buyers | CAC (USD) |
|---|---|---|---|---|---|---|
| Paid search | 12,000 | 3,000 | 3,000 | 18,000 | 300 | 60.00 |
| Paid social | 14,000 | 2,000 | 2,000 | 18,000 | 400 | 45.00 |
| Total | 26,000 | 5,000 | 5,000 | 36,000 | 700 | 51.43 |
Search CAC is 18,000 / 300 = $60.00; social CAC is 18,000 / 400 = $45.00. The blended result is 36,000 / 700 = $51.43, rounded to cents. Averaging the two channel CACs gives $52.50 and fails the total-spend test because the channels acquired different numbers of buyers.
Now compare the first completed orders. Sales below are after discounts and exclude sales tax. Returns reduce revenue; COGS already reflects the cost reversal for recoverable returned inventory.
| Channel | First-order sales (USD) | Refunds (USD) | Net revenue (USD) | Net COGS (USD) | First-order gross profit (USD) |
|---|---|---|---|---|---|
| Paid search | 36,000 | 3,600 | 32,400 | 18,000 | 14,400 |
| Paid social | 40,000 | 8,000 | 32,000 | 18,000 | 14,000 |
| Total | 76,000 | 11,600 | 64,400 | 36,000 | 28,400 |
This follows the arithmetic set out in IRS Publication 334's gross profit section: returns and allowances reduce gross receipts, then COGS is subtracted to obtain gross profit. The publication supports the calculation structure; your finance team's ledger policy determines the actual cost assignments.
The same buyers then place additional orders inside the observation window:
| Channel | Repeat sales (USD) | Repeat refunds (USD) | Repeat net revenue (USD) | Repeat net COGS (USD) | Repeat gross profit (USD) |
|---|---|---|---|---|---|
| Paid search | 8,400 | 840 | 7,560 | 4,200 | 3,360 |
| Paid social | 6,000 | 1,200 | 4,800 | 2,700 | 2,100 |
| Total | 14,400 | 2,040 | 12,360 | 6,900 | 5,460 |
Combine first-order and repeat gross profit, keeping the original acquired-buyer denominator:
| Channel | Gross profit through day 180 (USD) | Original buyers | Gross profit per buyer (USD) | CAC (USD) | Gross profit less acquisition cost (USD) |
|---|---|---|---|---|---|
| Paid search | 17,760 | 300 | 59.20 | 60.00 | -240 |
| Paid social | 16,100 | 400 | 40.25 | 45.00 | -1,900 |
| Total | 33,860 | 700 | 48.37 | 51.43 | -2,140 |
Search earns 14,400 + 3,360 = $17,760 of gross profit against $18,000 spent. Social earns 14,000 + 2,100 = $16,100 against the same spend. Social acquired buyers more cheaply, but its observed gross profit leaves a larger gap. Neither channel has recovered acquisition spending at the gross-profit level by day 180.
The proof is the reconciliation: 28,400 + 5,460 = $33,860 of cohort gross profit; 33,860 - 36,000 = -$2,140. The total must also equal the sum of the two channel gaps. Reconcile the costs to the marketing ledger and the order amounts to the cohort's sales and refund records before interpreting the difference.
For a longer view, the ecommerce LTV to CAC ratio extends the comparison to observed contribution after variable service costs. Do not call the gross-profit surplus here net profit.
Marketing owns the channel assignment and proposed budget change. Finance owns included costs, return adjustments and the reconciliation. The ecommerce or operations lead explains product mix and service costs. Agree on the observation window together before reading the channel ranking.
The result supports a controlled budget test, a promotion review or a product-mix investigation. It does not show that every additional buyer will cost the historical average. Separate new spending tests from the blended average of customers already acquired.
An industry CAC comparison can provide context when its definitions are visible. Your own comparable cohorts are the stronger operating reference when product margins and repeat behavior differ.
Bring a sample of your own acquisition expense and order exports, including customer IDs, cohort dates, refunds and costs, to Covirage. Covirage maps the columns, and its tools compute and check CAC and cohort figures; the external AI model explains the result with every figure traced to its rows. Create a free Covirage account, which takes about a minute, then set up your business and bring your files.
See customer profitability for the wider comparison of what each customer group earns and costs.
Start with your own mature acquisition cohorts and comparable channel periods. For an outside survey, check its publication date, retailer category, included costs and definition of a new customer. A survey that omits labor or counts orders cannot be compared directly with fully loaded CAC per buyer.
Keep a genuine acquired buyer in the original cohort even if a completed purchase is later refunded. Apply the refund and any related cost adjustments to that cohort's economics. Exclude fraud, test orders and cancellations under a written rule, and use it consistently across periods.
A lower acquisition cost can come with lower baskets, heavier discounts or fewer repeat purchases. Compare gross profit per acquired buyer over the same observation period before moving spending. Then consider fulfillment, payment and return costs to see whether the apparent advantage survives beyond gross profit.