Unit Economics: What One Customer Actually Costs You, and When You Get It Back

Every owner who tells me their marketing is working has done the arithmetic first. Almost none of them have done it in a way I would sign off on.

The usual version is advertising spend divided by new customers. That number is always flattering, because the two largest costs of winning a customer in a small Canadian business are the owner’s own time and the work that gets done before anyone has agreed to pay. Neither shows up in an ad account.

What actually belongs in the numerator

Acquisition cost is every dollar spent turning a stranger into a paying customer, in the period the spending happened. Pull it out of the general ledger, not out of the ad platform. Most of it is already sitting in accounts you post to every month.

That last row is the one people argue with. The argument is that the owner’s time is free because there is no payroll entry for it. It is not free. It is the scarcest input in the business, and pretending otherwise is how an owner concludes that a channel producing $400 customers at eleven hours of unpaid quoting each is a good channel.

The denominator is simpler and gets fewer arguments: customers who actually signed and paid in that period. Not leads. Not proposals out. Not “verbals”.

Lifetime value calculated on revenue is not lifetime value

The second half of the calculation goes wrong more quietly.

Lifetime value has to be built on gross margin, not revenue. A customer who spends $50,000 a year with you at a 12% margin contributes $6,000. Comparing an acquisition cost against the $50,000 is not analysis, it is theatre. If you do not have reliable margin by line of business, that is the first problem to fix, and it is a separate exercise: see margin by service line.

Then there is the retention assumption. Most owners multiply annual margin by a lifespan they picked because it felt right. If you have been trading for four years you cannot honestly assert a seven-year customer life. Use the period you can actually observe, and say so. Measuring it defensibly at small scale has its own arithmetic, set out in retention maths for a business with sixty clients.

Payback is a cash question

Here is the part that decides whether the business survives the growth it is chasing.

Acquisition cost is paid now. Margin arrives over months. The gap between the two is funded by something: your operating account, a line of credit, or supplier terms. Payback period is the number of months of gross margin needed to return the acquisition cost, and it is a cash measure, so use margin after the cash cost of delivery, before overhead.

The financing evidence says this gap is real and it is being borrowed against. Innovation, Science and Economic Development Canada reports that in 2025, 39% of small businesses requested external financing, and among those requesting debt financing, 45% intended to use it for working or operating capital. The same report shows 75% of borrowers had to pledge collateral, up from 66% the prior year. Growth funded on secured credit is growth where a payback assumption being wrong has consequences beyond a spreadsheet.

There is no Canadian benchmark, and anyone quoting one is quoting a vendor

I want to be direct about this, because it is where most articles on the subject lose their footing. No Canadian government body publishes a benchmark for acquisition cost or payback period. Statistics Canada does not. Innovation, Science and Economic Development Canada does not. If you have read that “good payback is twelve months”, you have read a software vendor’s blog post about software companies.

Set your own threshold instead, from three things you can actually observe:

Your cash position. How many months of acquisition spending can you carry before the line of credit is doing the work? That is your ceiling, and it does not care what any benchmark says.

Your realistic horizon. ISED’s survival data shows 68.0% of Canadian businesses with 1 to 99 employees are still operating five years after they start, and 48.2% at ten years. A payback assumption stretching past the horizon you can finance is not a plan.

Your cost trajectory. In the second quarter of 2026, 64.3% of Canadian businesses expected cost-related obstacles over the next three months, and 48.8% named inflation specifically. A payback calculated on today’s delivery cost and stretched over three years is quietly assuming those costs hold.

Three formulas, run by channel and not in aggregate

Acquisition cost per customer
  = (ad spend + sales salaries + commissions
     + sales software + owner's time at a rate)
    ÷ customers signed in the period

Contribution per customer per month
  = revenue x gross margin %  (margin, never revenue)

Payback (months)
  = acquisition cost ÷ contribution per month

Run it by channel, not in aggregate. An aggregate number tells you nothing you can act on, because it averages the referral that cost you a coffee with the trade show that cost you eleven thousand dollars and produced two customers.

The decision this actually drives

Payback shorter than your cash cycle?
├── Yes → spend more here, deliberately
│         Check delivery capacity first
└── No
    ├── Is margin the problem? → repricing, not more leads
    │     See pricing your services
    └── Is acquisition cost the problem?
        ├── Channel-specific → cut the channel, keep the rest
        └── Across every channel → the offer is the problem

Most owners who run this honestly discover the problem is margin, not marketing. That is a pricing conversation, and it is covered in pricing your services. The rest discover one channel is subsidising three, which is a much easier fix and one you cannot see without splitting the calculation.

The version I actually want to see

Two numbers per channel, updated quarterly: what a customer from that channel costs, and how many months of margin it takes to get it back. Both traceable to ledger accounts, both including the owner’s hours, both built on margin rather than revenue.

That fits on one page. It sits naturally beside the weekly operating numbers and it makes the break-even model forward-looking rather than historical, because you finally know what buying the next unit of volume costs.

If you want the calculation built from your own general ledger rather than from your ad account, send me a year of statements and a chart of accounts. The split between selling cost and delivery cost is usually the part that has never been drawn, and it is the part that changes the answer.