What a customer costs: how to count and not fool yourself
Almost every business owner knows what they pay for advertising. Far fewer know what one customer costs them — and almost nobody counts what that customer brings over their whole relationship with the business. Without these three numbers, every budget decision is made by feel.
The arithmetic is simple: divide everything you spent on a channel in a month by the number of clients, not requests. The difference is decisive — if one in ten requests buys, a client costs ten times what a request does. Include not just the budget but the contractor's fee and your staff's time handling enquiries.
In this article7 sections
Three numbers you cannot decide without
It all comes down to three quantities. They are simple, but they are constantly confused with one another — and that confusion produces wrong budget decisions.
- Cost per enquiryWhat you pay for one enquiry. Divide the month's marketing spend by the number of enquiries.
- Cost per customerWhat you pay for one person who actually paid. Divide the same spend by the number who paid.
- Customer valueWhat they bring over the whole time they work with you, not from a single purchase.
The gap between the first and the second is your funnel. The gap between the second and the third is your profit. Look only at the first and you can spend years pleased with cheap enquiries while not understanding where the money goes.
A worked example
A dental practice spent $1,000 in a month: $700 on ad budget and $300 on the contractor's work. It received 50 enquiries. Of those, 20 booked and 14 turned up.
| Metric | Calculation | Result |
|---|---|---|
| Cost per enquiry | $1,000 ÷ 50 | $20 |
| Cost per booking | $1,000 ÷ 20 | $50 |
| Cost per patient | $1,000 ÷ 14 | $71 |
Notice the gap: an enquiry costs $20, a patient costs $71. That is not an advertising failure, it is loss inside the funnel. Of 50 enquiries, 14 reached the chair — meaning 72% of the money went on people who never became customers.
When an owner says «advertising is expensive», roughly half the time the problem is not advertising. It is what happens between the enquiry and the payment: nobody called back, called back a day later, could not answer the price question, could not book a convenient time.
A useful exercise: work out which produces more money — cutting the cost per click by 20%, or raising the show-up rate from 28% to 40%. Almost always the second, and almost always it is cheaper.
Why one purchase is not enough
That same patient came in for a $60 cleaning. The acquisition cost was $71. It looks like a loss and like the advertising should be switched off.
But the patient came back six months later, then brought their spouse, then had $400 of treatment. Over two years they brought, say, $900. Against a $71 acquisition cost that is a return of more than twelvefold.
This is exactly why the «does it pay off» decision cannot be made on the first transaction. It has to be measured over the horizon a customer actually lives in your business. For food delivery that is a month, for dentistry it is years, for equipment sales it is one large contract and then servicing.
How to tell quickly whether a channel pays off
There is a simple rule people use when there is no time for detailed analytics: customer value should be noticeably higher than the cost of acquiring them. A threefold margin is a common benchmark — it covers cost of delivery, salaries and leaves a profit.
| Value to cost ratio | What it means | What to do |
|---|---|---|
| 5–10x and above | The channel is underloaded | Increase budget while the ratio holds |
| around 3x | A healthy state | Maintain it and improve the funnel |
| 1.5–2x | On the edge | Any price rise pushes you into loss — narrow to what is profitable |
| below 1x | A loss | Stop and investigate the cause |
Where the calculations usually go wrong
- Forgetting your own labour. Only the ad budget goes into costs, while the salary of the person handling enquiries and the owner's time do not.
- Counting every enquiry. Spam, wrong numbers and «are you hiring?» questions land in the statistics and conveniently lower the apparent cost.
- Crediting the last channel. Someone found you in search, read reviews on maps and submitted the form after seeing an ad — and advertising takes all the credit although three channels worked.
- Measuring over too short a period. In niches with long decisions — property, equipment, dentistry — a customer takes weeks from first touch to payment, and a month of data lies.
- Looking at the average across channels. An average of $20 per enquiry can be $8 from maps and $45 from advertising. Decisions made on an average are almost always wrong.
The minimum setup for honest counting
- Record every enquiryCalls, forms, messenger messages — all in one place rather than three different notebooks.
- Tag the sourceEach enquiry must carry a label saying where the person came from. Without it, per-channel calculation is impossible in principle.
- Record the outcomeTurned up, bought, for how much. Without this step you only have cost per enquiry, and that says little about money.
- Count repeat purchasesEven roughly: how many times a typical customer comes back in a year. This turns acquisition cost from an expense into an investment.
- Review once a monthNot daily. Daily swings mean nothing and only push you into nervous decisions.
What to do when the numbers are bad
The order matters: cheap and fast first, expensive and slow later.
- Check the response time to an enquiry. The difference between five minutes and a day is often several times over.
- Listen to ten calls. That is usually enough to find the biggest hole.
- Switch off the most expensive and least useful parts of the advertising.
- Check that the site is not losing people at the form and at the phone number.
- And only then lower bids, change ad copy and rebuild pages.
Frequent questions
What cost per enquiry is normal?
There is no normal in isolation from your average order value. $20 per enquiry is expensive for food delivery and very cheap for selling equipment. Compare against your own customer value, not against someone else's figures.
What if a customer costs more than they bring?
Look at the funnel before the advertising. There is usually more to be gained there than from optimising bids. If the funnel is fine, narrow advertising down to the highest-value queries.
Do we need a CRM to track all this?
A spreadsheet is enough at the start if enquiries are few. A CRM becomes essential once there are more than a few dozen a month: they can no longer be held in your head and data starts getting lost.
How do we count when sales happen offline?
Ask for the source at the point of contact and record it. Not precise, but better than nothing. Call tracking numbers make it precise by identifying the source automatically.
How should referrals be counted?
On a separate line. They cost almost nothing to acquire, so mixing them with paid channels understates the real cost of a customer and creates a false impression that advertising is more efficient than it is.
What if our order values vary hugely?
Count by groups rather than by an average: small orders separately, large ones separately. An average order value with a tenfold spread describes nothing, and decisions based on it come out wrong.
The materials answer general questions. We will look at your specific case — free and without obligation.