What a new customer costs and how much you can pay

The ad dashboard shows what each new customer cost you this month, and the number just sits there on the screen, without saying whether it is good or bad. You compare it with last month, you compare it with what feels reasonable, and you close the laptop with the same doubt you opened it with.

That doubt has an answer, and it depends on your business, not on the platform. By the end of this article you will know the most you can pay for a new customer before losing money, how to turn that figure into what you can pay for each person who gets in touch, and which of the five steps to start with tomorrow.

The steps go from the obvious to what almost nobody looks at. The first ones give you the figure; the last ones tell you when that figure misleads you and when an expensive customer is a better deal than a cheap one.

1. Work out what a sale leaves you, not what you charge

The first thing you need to know how much you can spend on getting a customer is how much they leave you. Not what you charge, but what is left after taking away what it costs to do the work or deliver the product: materials, the hours you pay for, shipping or the card fee.

That number is the margin on a sale, and it is the ceiling for everything else. If a sale leaves you little, no cheap ad will make it profitable; if it leaves you a lot, you can afford to pay quite a bit to bring in whoever buys it.

This is usually where the first surprise appears. Many owners know exactly what they charge for each service and much less clearly what they keep, especially when part of the cost is their own hours, which never get written down. If you do the work yourself, put a reasonable price on your hour and take that away too: otherwise the margin will look bigger than it is.

What counts as the cost of a sale

Only count what would not exist if that sale did not happen. You pay the rent whether you sell or not, so it does not go here; the materials you use on that particular job do.

If you offer several services, start with the one you sell most or the one you want the ads to bring in. One is enough to do the whole sum, and repeating it for the others afterwards takes minutes. Today's step: write down in one line what an average sale of your main service leaves you.

2. Count how many times a customer comes back

A customer does not leave one sale: they leave every sale they make while they stay with you. Someone who comes once and never returns is worth much less than someone who comes back every season, even if both pay the same on day one.

This is called customer lifetime value, and it sounds more complicated than it is. You do not need to calculate it precisely: it is enough to know how many times a typical customer buys in a year or two, and multiply that by the margin from the previous step.

If you do not have the figure

Look up the customers who arrived a year or two ago, in your diary, your invoicing software or your notebook, and count how many times each one has come back. You do not need more than that for a reasonable idea, and the result is almost always a surprise, one way or the other.

If the number comes out low, do not just take it as bad news: it is also a clue. A customer who does not come back is often a customer nobody has reminded that you exist, and a timely message on WhatsApp or by email usually raises that number without spending another euro on ads.

If your business is a one-off purchase, like a renovation or a wedding, the number is one and the rest of the article works the same. In that case, what a customer brings you through recommendations is the part that adds up, and it is worth noting separately. Today's step: write down how many times a customer of yours buys, on average, over two years.

3. Decide how much of that margin you can spend on bringing the customer in

With the two previous steps you already know how much a customer leaves you over their time with you. That is the amount you could, in theory, spend on bringing them in without losing money, but spending all of it would mean working for free for the ad platform.

The sensible thing is to decide in advance what share of that margin goes on bringing in new customers, and to keep the rest for the premises, salaries and your own time. A prudent starting point is a third, which you then adjust depending on how tight your fixed costs are.

An example with numbers

Take a bicycle repair shop, with some example numbers. Each repair leaves an average margin of 30 euros and a typical customer comes back about four times in two years, so each customer leaves 120 euros. If the shop puts a third into bringing in customers, this is its ceiling:

If the shop pays less than 40 euros for each new customer, the advertising pays off. If it pays more, every new customer is costing it money, even if the dashboard celebrates it as a month of growth.

How much you can pay up front

The table tells you whether a customer is profitable over two years, but not whether you can wait two years for that. In the shop, a new customer can cost 40 euros and leave only 30 on the first visit: the rest comes on later visits.

That is why it is worth asking yourself how much money you can have tied up in customers who have not come back yet without struggling to pay the month's costs. That figure sets how far you can raise the budget, even if the table says you could spend more. Today's step: fill in the table with your two numbers and choose your share.

4. Turn that ceiling into what you can pay per contact

Ads hardly ever charge you per customer: they charge per click, per message or per form filled in. To know whether a campaign is going well halfway through the month, before anyone has bought, you need to turn the ceiling per customer into a ceiling per contact.

The sum depends on how many of the people who get in touch end up buying. If one in four people who write to the shop buys, the sum has three steps:

1. Take the most a new customer can cost: in the shop, 40 euros.
2. Divide it by the number of people who need to get in touch for one to buy: in the shop, four.
3. The result is the most you can pay per contact: in the shop, 10 euros.

If you do not know how many people ask before buying

For a month, note down everyone who writes, calls or fills in a form, and at the end of the month count how many bought. You do not need any software: a sheet with two columns is enough. One month already gives you a ratio to start with, and every month you keep noting it makes it more reliable.

This is the number you can really keep an eye on every week in the platform's dashboard. If the cost per contact goes above it, you have an early warning, long before it shows in your till.

And if the cost per contact is fine but customers do not come, the problem is not the ads but what happens next: how long you take to answer, what the page says or what you offer. Today's step: work out your ceiling per contact and write it down next to the dashboard.

5. An expensive customer can work out cheaper than a cheap one

The least intuitive part of all is that the cost of bringing in a customer, on its own, tells you nothing. Two channels or two campaigns can bring you customers at very different prices, and the expensive one can be the better deal if its customers come back more.

If, in the shop, customers who come through a discount ad cost little but never return, and those who come through an ad about the seasonal service cost twice as much but come back every year, the second group leaves more money. Anyone who only looks at the cost per customer would switch off the good ad and keep the bad one.

The catch is that the difference takes time to show. In the first month, the discount ad looks like the clear winner, because it brings more people for less money; the other one's advantage only appears when its customers come back. That is why, before switching off a campaign that brings expensive customers, it is worth waiting long enough to see whether they return.

What matters about a new customer is what they leave after what it cost to bring them in.

How to compare two campaigns

When you have two campaigns or two channels, fill in the table from step three for each one, with how often their own customers come back rather than an average of everyone. It takes a little longer, and it is the only way not to reward the channel that brings people who only come for the deal.

All you need is to note, next to each new customer, where they came from, and check a few months later how often they have come back. Today's step: split your customers from the last quarter by where they came from.

Where to start working out what it costs to get a customer

If you only do one thing this week, make it step two: count how many times a customer comes back. It is the figure fewest businesses have and the one that changes the sum the most, because it decides whether you can pay a little or quite a lot for each new customer.

Then the rest, in order: the margin on a sale, the table with your share, the ceiling per contact and, once you have a few months of data, the comparison between channels. You do not need to do it all at once; each step already leaves you with a figure you did not have before.

With that number and the margin on a sale, you have the figure the ad dashboard does not give you: the most you can pay. And if you would rather work it out with someone, with your numbers in front of you, that is what the diagnostic session is for: half an hour, free of charge, to see how much you can pay for a customer and whether your ads are below it.