Knowing which customers actually make you money

Your biggest customer is probably your worst one. They buy the most, so they got the discount. They negotiate every renewal, change their mind mid-job, call you on Saturday, and pay in seventy-five days. You keep them because losing them would leave a hole in your sales, and sales is the number you look at.

Meanwhile there is a quiet customer who orders steadily, pays the day the invoice arrives, never asks for anything unusual, and has not heard from you in a year. You spend no time on them at all, which is exactly why they are profitable and exactly why you might lose them without noticing.

Almost every small business ranks its customers by revenue, because revenue is the only number that is easy to see. Revenue tells you how busy a customer makes you. It says nothing about what they leave behind.

By the end of this guide you will have ranked your customers by profit rather than sales, know your profit per hour of attention for each of them, and know which ones to grow, which to reprice, which to change the terms on, and which to let go, with the words to use for each.

(Written for any country. Money appears as plain numbers, read them in your own currency. The method works for a business with ten customers or ten thousand, though large-volume businesses should read the note on segments below.)

The four costs that never reach the invoice

Gross margin is where most owners stop: what you sold it for, minus what it cost you. But between gross margin and money in your pocket sit four things that differ enormously from one customer to the next.

Your time, outside the job itself. Meetings, phone calls, revisions, chasing, hand-holding, the visit to sort out a complaint. Two customers buying the identical thing can consume wildly different amounts of you, and the difference is never billed.

Discounts and free extras. The percentage off, but also the free delivery, the rush job absorbed at no charge, the extra unit thrown in, the sample, the "I won't charge you for that."

Rework, returns and redos. Some customers reject, return, change their mind or need it done twice. That is your materials and your hours, spent again for the same money.

The cost of being paid late. Money owed to you is money you cannot use. At roughly 2-3% a month, a customer who pays sixty days late on a large balance has taken a real amount out of the job, and if you are borrowing to cover the gap, it is not hypothetical at all.

Add those four to the picture and the ranking changes, usually dramatically. In most businesses that do this properly, the top customer by revenue is not in the top three by profit, and at least one customer turns out to be worth nothing at all.

To do this you need, for each of your top customers: what they bought last year, what discount they got, roughly how many hours of yours they consumed, what you spent delivering to them, how many days they take to pay, and what came back. Rough numbers are fine. Pick your ten biggest customers by sales and gather that now. It is a notebook exercise, not an accounting project.

The calculation

One line per customer, over a full year:

Customer profit =
    sales × gross margin %
  − discounts and free extras given
  − (hours of your time × your hourly rate)
  − delivery and travel costs
  − rework and returns
  − cost of the money for however late they pay

Two real-shaped customers, side by side.

Customer A, the big one.

Annual sales                                      480,000
Gross margin 35%                                  168,000
Your time: 6 hrs a month × 12 × 800               −57,600
Delivery: 24 trips × 900                          −21,600
Rework and returns                                −12,000
Paid 45 days later than terms, cost of money       −2,700
                                                 ---------
PROFIT                                             74,100   (15% of sales)

Customer B, the quiet one.

Annual sales                                       96,000
Gross margin 45%                                   43,200
Your time: 1 hr a month × 12 × 800                 −9,600
Delivery: collects their own goods                       0
Rework and returns                                       0
Pays on time                                             0
                                                 ---------
PROFIT                                             33,600   (35% of sales)

Customer A produces five times the sales and only about twice the profit. That is worth knowing. But the number that should actually change your week is this one:

Profit per hour of your attention
Customer A: 74,100 ÷ 84 hours  =    880 per hour
Customer B: 33,600 ÷ 12 hours  =  2,800 per hour

Every hour you spend on the big customer earns you less than a third of what an hour on the quiet one earns. And your hours are the thing you have least of.

Do this for your ten biggest. Then, importantly, do it for two or three small ones you assume are marginal. That is where the surprises usually are.

If you have hundreds of small customers rather than ten large ones, do not do this per customer. Do it per segment: walk-in versus account, delivery versus collection, one product line versus another, one job type versus another. The arithmetic and the conclusions are identical.

Reading your list

Put every customer into one of four boxes, using profit and how much of your time they take.

The bottom-right box is where your problems live, and it is nearly always occupied by someone you consider important.

Two things to notice as you sort.

A customer can be unprofitable without being difficult. Often it is just terms that were set years ago and never revisited, or a delivery arrangement that made sense when they were around the corner.

Time demand is a habit, not a personality. Most high-maintenance customers became that way because it was allowed. That means it can usually be changed without losing them.

Fix it before you fire it

Firing customers is satisfying to read about and rarely the first move. Five fixes, in the order to try them.

Raise the price. The most direct one, and often accepted with less argument than you expect from a customer who knows they are demanding.

We have reviewed our pricing for the year. From the 1st, your
rate moves to [amount]. It reflects the level of service you get
from us, including [the specific extras]. Everything else stays
as it is.

Change the terms. Deposit up front, payment before delivery, shorter credit period, a card or mobile payment on collection. This alone fixes a large share of low-profit accounts, because the finance cost and the chasing time both disappear.

Price the free extras separately. Delivery, urgent jobs, extra revisions, out-of-hours calls, installation. Not to punish anyone, to make visible what has been free. Half your customers will start collecting, and that is the point.

From next month, delivery is [amount] per drop, or free on orders
over [amount]. Collection stays free as always.

Set a minimum order or a callout fee. Small frequent orders are a common profit killer, because the handling cost is the same whatever the size.

Change the service level, not the relationship. Route their routine questions to a set time or a written channel. Give them a scheduled call rather than an open line. Most people adapt without complaint when the alternative is offered pleasantly and consistently.

Give each fix ninety days and re-run the numbers. Many accounts move from the bottom-right box to the top-left with nothing more than a delivery charge and a deposit.

Letting one go, properly

Sometimes the answer is that this customer costs more than they bring, and will not accept any change. That is a legitimate business decision, and it should be handled well. A customer you release politely still sends you referrals; one you drop badly tells everyone.

Do it in this order:

• Never in anger, never in the middle of a job. Finish what you have committed to.
• Give notice and a reason that is about you, not them. "We are changing how we work and can no longer support [arrangement]."
• Offer an alternative. Recommend someone who genuinely suits them better. It is generous, it is often true, and it removes the sting.
• Give a transition period. Thirty days, or until the current order is complete.

Thank you for your business over the last [period]. From [date] we
are changing our terms and will not be able to continue supplying
on the current arrangement. I would be happy to keep working with
you at [new terms], and if that does not suit, I can recommend
[alternative] who would look after you well.

Notice that this is not a dismissal. It is an offer they may accept, and a fair number will. Either outcome is fine. That is what makes the conversation easy to have.

Do more with the good ones

The other half of this exercise is the half people skip. Take your top box (profitable, low maintenance) and do three things.

Sell them more of what you do. They already trust you. Most of them do not know your full range. Tell them, individually and specifically: "you buy X from us. Did you know we also do Y? It would save you the second trip."

Ask them for referrals, by name. Your best customers know people like themselves. Ask each of them once a year, specifically, and offer to look after whoever they send.

Make yourself harder to leave, in a good way. A standing order, a service schedule, holding their spares, remembering their sizes and preferences, a scheduled slot. Convenience is what keeps quiet, profitable customers quiet and profitable.

And thank them. They pay on time, they do not complain, and they have almost certainly heard less from you this year than the customer who shouts. A short personal message at the end of the year costs nothing and gets remembered.

One caution as you grow the good ones: watch what share of your sales any single customer becomes. Above roughly a quarter to a third, they own you. You cannot enforce terms, you cannot raise prices, and the month they leave becomes an emergency. Grow them, and grow others alongside them deliberately.

Keep it alive with one line a month

The exercise is only worth doing once unless you turn it into a habit. One page, one line per major customer, updated monthly:

The two columns that do the most work are hours and days to pay, because they move before profit does. A customer whose hours are climbing and whose payments are slowing is telling you what next year looks like, while you can still do something about it.

Then review the whole list twice a year, when you review your prices. The same afternoon, the same notebook.

What goes wrong

Ranking by sales. The default, and the source of the whole problem. Sales tells you who keeps you busy.

Forgetting your own hours. The single biggest omission. If your time is not in the calculation, every customer looks profitable.

Treating everyone identically. Equal treatment sounds fair and means your best customers subsidise your worst.

Fearing the loss of a big number. Losing revenue that came with no profit costs you nothing and returns your time. Run the number before you panic.

Firing first. Most bad accounts are fixable with a price, a term, or a charge for something that used to be free.

Fixing quietly and hoping they do not notice. Announce changes clearly, in writing, with a date. Silent changes get discovered at the worst moment and read as sharp practice.

Doing it once. Customers drift. A profitable account in January can be a problem by September, and only the monthly line will tell you.

Ignoring the quiet ones entirely. They are the most profitable people you deal with and the easiest to lose, because nobody ever contacts them.

Letting one customer become a third of your business. Then none of the above is available to you, because you cannot afford the answer.

Your first week

Do the profit-per-hour column before anything else, even roughly. It is the number that most often makes an owner change how they spend next week, because it usually shows that the account consuming half their attention is earning them less per hour than the one they have not spoken to since last year.