Customer acquisition cost does not appear anywhere in a set of financial statements. It is a management metric assembled from pieces that live in several different places in the books: advertising expense, promotional discounts recorded against revenue, coupon and referral costs buried inside marketplace settlements, agency retainers sitting in professional fees, and sometimes a share of payroll. Where those pieces sit determines whether a seller’s CAC figure is trustworthy or decorative.
The useful version of this question is not what CAC means. It is which ledger accounts feed it, and what a seller has to change to produce it reliably every month.
The definition, and the version that matters
CAC is total acquisition spend for a period divided by new customers acquired in that period. Two variants get used, and confusing them is the most common reporting error.
Blended CAC divides all acquisition spend by all new customers, including the ones who arrived organically. It answers what it costs the business overall to add a customer.
Paid CAC divides paid spend by customers attributable to paid channels. It answers whether the advertising is working.
Blended CAC is easier to compute and harder to game, and it is the right number for a board slide. Paid CAC drives media decisions. A seller reporting one while making decisions that require the other is a common pattern, and it surfaces as advertising that keeps scaling past the point where it makes money.
Where the inputs actually live
The numerator is the hard part, and the pieces are scattered.
Advertising expense
Off platform ad spend arrives as its own invoices and is straightforward. On platform advertising is the problem: it is charged inside marketplace settlements, mixed with referral fees, fulfillment, and storage. A seller whose books carry a single “Amazon fees” account has their entire on platform ad spend hidden inside a line they read as a cost of sales, which means CAC cannot be computed at all and fee load looks worse than it is.
Amazon’s published seller pricing separates the order level charges, referral fees at a percentage of total price or a minimum amount, from optional services charged to the account. The books should follow that separation, with advertising in operating expenses and referral and fulfillment in channel cost of sales.
Promotions and coupons
A discount given to win a first order is acquisition spend in substance, and in the books it lands as a contra-revenue item rather than an expense. That is correct accounting and it means the cost does not appear in any expense account, so a CAC calculation built only from expense lines understates it.
Coupon redemption fees, lightning deal fees, and promotional clipping costs charged by marketplaces sit inside settlements as their own transaction types. Mapped to their own accounts, they are available for the CAC numerator. Flattened into a fees bucket, they are not.
People and tools
Agency retainers, a contractor managing campaigns, creative production, and the software subscriptions that exist for marketing are all acquisition costs by any reasonable definition, and they sit in professional fees, contractor expense, and software. Whether to include them is a policy decision. The requirement is to decide once, document it, and keep the definition stable, because a CAC trend computed on a shifting definition is not a trend.
The denominator problem
New customers is a harder figure than it sounds for a marketplace seller. On most marketplaces the seller does not receive identifying customer information, which means the same buyer purchasing twice may be uncountable as a repeat customer.
Practical approaches: use orders as a proxy and call the metric cost per order rather than CAC, which is honest and useful; use the channel’s own new to brand reporting where it exists; or compute true CAC only for the direct storefront where customer identity is known, and cost per order for the marketplaces. What does not work is presenting cost per order as CAC, because it understates acquisition cost by whatever the repeat rate is.
A worked month
One seller, one month, three channels.
- On platform advertising, from settlement detail: $14,200
- Off platform paid social and search: $9,800
- Promotional discounts, contra-revenue: $3,400
- Coupon and deal fees from settlements: $1,150
- Agency retainer: $2,500
- Creative production: $900
Total acquisition spend: $31,950. New customers on the direct store: 410. New to brand orders across marketplaces: 1,240. Blended new customer count: 1,650.
Blended CAC: $19.36. Against a contribution margin of $12.95 per unit and an average order of 1.3 units, first order contribution is roughly $16.84, so acquisition is running above first order contribution. That is not necessarily wrong for a business with genuine repeat purchase, and it is a fact the owner needs to know deliberately rather than discover during a cash squeeze.
Without promotional discounts and settlement based coupon fees, the same calculation gives $16.06, which looks comfortably profitable on first order. The $3.30 difference is the entire conclusion, and it comes from two inputs that sit in places most sellers do not include.
Making the books produce it
Four structural requirements, all of them chart of accounts work rather than analysis:
- Advertising separated from marketplace fees, by channel, mapped from settlement transaction types.
- Promotional discounts as a contra-revenue account of their own, not netted into sales.
- Coupon, deal, and referral program fees in their own accounts.
- A documented policy on whether people and tools are included, held constant across periods.
At multichannel volume this depends on settlement detail surviving into the ledger, which is the job ecommerce accounting software does, ConnectBooks among the products in that category, syncing marketplace settlement detail into QuickBooks Online, QuickBooks Desktop Enterprise, or Xero with transaction and SKU level granularity intact. A tool that summarizes settlements to a net figure keeps the books tidy and leaves advertising and promotion costs merged into places CAC cannot reach.
Reading the number
CAC on its own says very little. It becomes useful against contribution margin, where the question is how many orders it takes to recover acquisition cost, and against the payback period, where the question is how long that takes in cash terms.
A CAC of $19 is healthy for a business with $40 of contribution per order and a third of customers returning within ninety days. The same figure is a problem for a single purchase product with $15 of contribution. Benchmarks borrowed from other sellers are close to useless for the same reason.
For the broader financial planning this feeds into, the Small Business Administration’s guidance on managing business finances covers the cash timing side, which is where acquisition spend and inventory purchasing collide. Both are paid upfront, and both get recovered from the same customers later.