← Back to the front page
Marketing & CustomersAnalysis

The customer you should turn away

A full diary can conceal weak economics. Before declining a customer, distinguish a poor fit from a price, promise or process that needs fixing.

Mercator editorial desk5 min read
A small green shop on a balance beam is outweighed by an enormous oxblood shopping bag, on warm cream paper.
Original AI-assisted illustration for Mercator.

Imagine a customer who is delighted with your work. They recommend you, praise your flexibility and keep ordering. They also need a revised quotation on Sunday, a delivery outside the agreed window and several small changes after approval. Each request seems reasonable. Together, they have bought a different service from the one you priced.

Turning that customer away may look like poor selling. Yet accepting every request can make a business less reliable for everyone else. The useful question is whether you can serve this customer well on terms that also work for the business.

The invoice is only half the story

Consider a hypothetical owner-managed service business. Two customers each buy a €900 package, excluding VAT. Each job incurs €280 in materials, delivery and other variable costs. The first takes eight hours of the owner’s time. The second takes fourteen, including calls, revisions and rescheduling.

Both leave €620 before the owner’s pay, fixed overheads and tax. But they leave €77.50 and about €44.29 respectively for each hour of the owner’s time. These are illustrative contribution figures, not profit or recommended hourly rates.

If the owner is the bottleneck and has alternative work available, that difference matters. A €620 contribution is less attractive when earning it prevents another worthwhile job. Management accountants formalise this by comparing contribution against the resource in short supply, as ACCA explains in its discussion of limiting factors.

The resource need not be labour. It might be workshop space, a vehicle or a specialist’s approval. Measure the constraint you actually face. An immaculate hourly calculation will not solve a shortage of storage.

Why keeping them feels easier

There is an awkward asymmetry here. A customer who leaves is visible: an empty slot, a smaller invoice total, perhaps an uncomfortable conversation. The cost of keeping them can disappear into evenings, interruptions and promises made to other customers.

Past effort complicates the judgement. You have spent months winning the account; surely it deserves another chance. In their 1985 research on sunk costs, Hal Arkes and Catherine Blumer examined how previous investments of money, effort or time can encourage continued commitment. Their work does not diagnose any particular customer relationship. It offers a useful question: if this enquiry arrived today, with everything you now know, would you quote on the same terms?

That is not a reason to ignore loyalty. Established trust, easier future orders and credible referrals can have prospective value. Distinguish those future benefits from the wish to justify what you have already spent.

First, examine the promise

Before labelling someone a bad customer, inspect what your business invited them to expect. “Bespoke”, “flexible” and “nothing is too much trouble” are attractive words. They are also open invitations unless the service explains their limits.

A customer who asks for repeated revisions may have bought from a page promising unlimited care. A customer who phones three times may be unable to find an order update. In those cases, refusal avoids a problem you helped create.

Record the extra work for a few recent jobs: what was requested, why it happened, how long it took and whether the original agreement covered it. Separate errors you must put right from changes the customer is adding. Count reasonable explanations as part of service, rather than treating unfamiliarity as misconduct.

The exercise may reveal a missing product. Some customers need certainty urgently and will pay for a reserved slot. Others can accept a slower service at a lower price. That is a proposition to test, not a licence to invent surcharges after accepting an order.

Offer a workable version

For a future order, make the choice concrete. The standard price might include one approval round and an agreed delivery window. Extra revisions might need a fresh quotation. An urgent job might require a separately priced slot. Describe the offer in the quotation, before work starts, so the customer can decide whether it suits them.

In our example, suppose the fourteen-hour job remains fourteen hours and its other variable costs remain €280. Matching the first job’s contribution per owner-hour would require a price of €1,365: fourteen times €77.50, plus €280. That is €465 above the original price.

The calculation does not prove anyone will pay it. It shows the size of the gap. If the customer values the extra service at €50 while it consumes €465 worth of capacity at your chosen comparison rate, polite persuasion will not remove the mismatch.

You could reduce the work instead. A single consolidated approval, fewer delivery options or a narrower specification might make the original price viable. Test a defined arrangement on one order. Judge the result against the agreed scope, time and payment terms, rather than against how relieved you feel after the conversation.

When keeping the customer makes sense

The argument changes when capacity is spare. Rejecting a job that contributes €620 does not improve cash flow if the alternative is no job. Owner time still has value, but it is not automatically worth the rate you could earn in a fully booked week.

Learning also matters. A first order can require setup that later orders avoid. A customer may open a worthwhile market or provide unusually useful feedback. Keep such exceptions deliberate: specify what benefit you expect, what it costs and when you will review it. “Potential” is difficult to disprove without a date.

Equally, a spreadsheet cannot make repeated abuse or impossible promises acceptable. Apply clear standards consistently, honour existing commitments and avoid accepting new work you cannot deliver properly.

A refusal can be straightforward: explain which requirement you cannot meet, offer the version you can deliver and, where appropriate, suggest another supplier. There is no need to prosecute the customer’s character.

The aim is to make a promise your business can keep. Sometimes that means a better process. Sometimes a different price. And sometimes it means leaving an order for a business better equipped to fulfil it.

Mercator is a publication by Sofiaga. Sources are linked within this article. Questions or corrections are welcome through our contact page.