Writing · The buyer's screen

One big customer is a screen, not a discount.

Above roughly 15% to 25% of revenue in one customer, whole categories of buyer stop looking. Above 30%, most stop entirely. That is a gate, and gates work differently from prices.

Say your biggest account is 38% of your revenue. Somebody will tell you that knocks a percentage off what your business is worth. Ten percent, maybe fifteen, depending on who is talking.

I will not tell you that, for a simple reason. There is no honest published number for it. Nobody has ever run a proper study that produces a discount percentage for customer concentration in small service businesses. Every figure you hear is a practitioner's estimate dressed up as data. I do not quote those, and you should be careful with anybody who does.

What is real is much simpler, and worse.

What actually happens

Concentration is a screen. It is a yes-or-no gate that a buyer runs before he ever gets to price.

Above roughly 15% to 25% of revenue in a single customer, whole categories of buyer stop looking at the listing. Above 30%, most stop entirely. They do not make a lower offer. They move to the next business on the page, and you never learn they were there.

A screen does not shave a number. It empties the room.

Why a buyer behaves that way

Because he is borrowing money, and the loan does not care about your customer list.

Picture the man buying your company. He is putting in ten percent and borrowing the rest. He has a payment due every month for ten years. Now he looks at your numbers and sees that 38% of the revenue comes from one company, and that company has been loyal to you for 14 years — not to the business, and certainly not to him.

If that customer leaves in his first year, his revenue drops by more than a third and his loan payment does not drop at all. He is out of business in one quarter. So he does not negotiate. He passes.

And his bank gets there before he does. The loan officer runs the same arithmetic and marks the file. Around 78% of buyers at this size expect bank money, so the bank's opinion is not a side issue. It is the deal.

How to measure it properly, this afternoon

Run a sales-by-customer report for the last 12 months, sorted biggest first. Take the top one and the top three as a share of total revenue. That is your number. It takes ten minutes.

Then check for the three traps, because they are where the real answer hides.

1. Related accounts counted as separate customersThree entities, three account numbers, one man signing all three purchase orders. That is one customer. Group by who actually decides, not by how your software files it.
2. Counting jobs instead of dollarsTwo hundred small accounts and one big one can look beautifully spread out if you count invoices. Count money.
3. Ignoring gross profitRun it a second time on gross profit, not revenue, if your books will support it. Sometimes the big customer is the low-margin one — which means he is 38% of your revenue and 19% of your profit. That is genuinely better news, and it is also the case that a buyer will believe, because he can see it in the same report.

Reading your own number

• Under 15% — clean. Nobody screens you out on this.
• 15% to 25% — the room starts to empty. Some buyers stop here.
• Over 30% — most buyers are gone before you meet them.
• Over 40% — that is not a customer. That is a partner who can fire you, and you should think about it that way in every other decision too.

How to fix it, and how long it takes

There are two ways and only two. Grow everything else, or move some of that work somewhere else.

Here is the arithmetic on a $3M shop with one account at $1.15M, which is 38%.

Grow the rest. To get that account down to 25% while keeping every dollar of it, you need total revenue of about $4.6M. That is $1.6M of new work from other customers. At real growth rates for a service business, that is two to three years.

Or shrink the account. To get to 25% while staying at $3M total, that account has to come down to $750,000. You would be walking away from about $400,000 of work you already have.

Almost everybody should choose the first one. It is slower, and it does not cost you the account or the relationship. But notice that neither version is a 90-day project. That is why this is the first thing I put on a plan and the last thing that finishes.

What not to do

Do not fire the customer. Owners get angry enough to think about it. Run the number first. Dropping that account turns 38% concentration into a 38% hole, and filling a hole that size takes the same years you would have spent growing past it — only you start from further back.

Do not sign a long contract and call it solved. Owners love this one, and buyers see through it immediately. A three-year agreement that a customer can cancel on 30 days' notice is a letter, not a contract. Worse, most supply agreements end when the company changes hands — so the one document you were relying on expires at closing.

What does help, in order:

• An agreement in writing that survives a change of control. Ask your attorney for that specific clause, because it is the one that matters and it is usually missing.
• More than one relationship inside their building. If only you know their buyer, that is not a customer relationship. It is your relationship, and it does not transfer.
• Never being the supplier they could replace inside a month. Something they would have to plan around losing.
• Written history. Volume by month for five years, showing they stayed through a downturn.

The other concentrations nobody counts

The same screen applies to four things owners never think of as concentration.

• One supplier. If one distributor is your parts, your credit line and your pricing, that is a single point of failure with a phone number.
• One key employee. The man who is the only one who can do the hard work. Same box on the credit memo as an owner who cannot leave.
• One referral source. If most of your work comes from one builder or one property manager, your customer list has one name on it wearing several hats.
• One license. If the master license the company operates under is in one man's name, and that man is you, the company cannot legally run without you. That is the hardest version of this problem and the least discussed.

Start counting now

This is the slowest thing to fix on the whole list, which makes it the first thing to start. And it is worth fixing even if you never sell anything, for the plainest reason there is: a customer who is 38% of your revenue has a great deal of say over how you run your own company, and he knows it.