Two customers on the same books. The first spends about £8,000 a year, argues about every invoice, pays at ninety days when the terms are thirty, and takes up perhaps a day a month of somebody's time in phone calls. The second spends £3,200, pays on the fifteenth without being asked, and has never once queried a price.
Ask a business owner which one they would drop and almost all of them say the small one, because it is small. Then ask which one they would rather have ten of.
Revenue is not the test, margin per hour is
The reason bad customers survive is that businesses measure them by turnover, and turnover is the only number that makes them look good.
The number that matters is what you keep per hour of your capacity. Take the gross margin on the work, subtract the time that is not the work, and divide by the hours the whole relationship consumes. The hours that are not the work are the ones nobody records: the site visits that came to nothing, the three revisions of a quote, the chasing, the meetings, the complaints, the extra care taken because you know this one will inspect it.
Do that honestly for your ten largest customers and you will usually find that the ranking changes, and that one or two of the big ones are close to zero, and that one is negative.
It is worth doing once a year with actual figures rather than by instinct, because instinct rewards the loud. You will find that some large customers are excellent, which is useful to know as well, and that concentration is a separate risk from profitability. A customer who is 40% of your turnover and profitable is a good customer and a bad position.
The one who always pays late is charging you for the privilege
A customer who pays at ninety days on thirty day terms is borrowing from you, at your cost of money, without asking and without a limit.
Work out what that is worth. If you have £15,000 outstanding with them at any time and your overdraft costs 11%, they are taking about £1,650 a year out of you before you count the hours spent chasing it. On £8,000 of spend, that is most of the margin.
The distinction to draw is between a customer who has had a bad quarter and a customer whose business model includes paying you late. The first is a relationship worth carrying, and the right response is to talk to them, agree a plan and hold them to it. The second will not change, because it is working for them.
For the second sort there are only three moves: get paid up front or on delivery, price the delay in, or stop. What does not work is a stern email, because you have sent seven and they are aware of your tolerance to a fortnight's precision.
You have rights here and they are worth knowing. Under the late payment legislation you can charge statutory interest and a fixed sum for the cost of recovering a commercial debt. Most small firms never invoke it for fear of the relationship. Putting the clause in your terms and mentioning it once, early and pleasantly, changes behaviour more often than people expect.
Scope creep is a discount you never agreed to
The other way a customer becomes unprofitable is a hundred small additions, none of which was worth arguing about on its own.
While you are here, could you just. It is only a small change. I assumed that was included. Each one costs twenty minutes and none of them is worth a difficult conversation, and by the end of the year you have given away a fortnight.
The fix is process rather than confrontation. Write down what is in scope in a way a stranger could read. When something falls outside it, say so at the time, in a friendly and matter-of-fact way, with a price. "Happy to do that, it's £X and adds half a day, shall I go ahead?" That sentence costs nothing to say and is the entire defence.
The customers who object to being told the price of something are telling you something useful about themselves, and it is cheaper to learn it in month two than in year three.
What the capacity is actually costing you
The argument for keeping bad work is always the same: some money is better than no money, and the alternative to this job is an empty day.
Sometimes that is true. In a genuinely slack period, marginal work that covers its variable costs is worth taking, and businesses that turn down work on principle while sitting idle are being precious.
But it stops being true the moment you are near capacity, and most owner-managers reach capacity long before they notice, because the constraint is them. Once you are full, every hour spent on the difficult customer is an hour not available for a good one, and the cost of the bad customer is not their low margin, it is the margin you did not earn instead.
It also costs you in ways that do not appear in a spreadsheet. Difficult work is what makes good staff leave. A job you resent is a job you do more slowly. And a business that is full of work it dislikes stops looking for work it would like, because there is no room to take it if it turned up.
Saying no in a way that leaves the door open
Turning work down does not require a confrontation and should never require a lie about being busy, because they will see your van.
For a one-off you do not want, decline quickly, be pleasant, and if you can, name somebody else who would do it well. Speed is a kindness and recommending a competitor makes you the person they ring next time with something better.
For an existing customer you want to stop serving, the honest version is usually a price. Put them on terms that reflect what they actually cost you, payable in advance, at a rate that makes the work worth doing. Some of them accept and become good customers overnight, which is always a surprise. The rest leave, without a row, and you have your Tuesdays back.
Where this shows up in the game
Contracts in Business Baron carry payment terms, and a customer who pays at ninety days is modelled as exactly that, an unsecured loan you did not agree to. Capacity is finite and taking one contract means declining another, so the decision is always a choice rather than an addition. Turning work down damages your relationship with that customer and nothing else, which is usually a trade worth making and rarely feels like it at the time.