Customer Profitability Analysis: Your Biggest Customer Might Not Be Your Best Customer

Every business has its favorite customers. There’s the one who places enormous orders, the one whose logo looks impressive in a presentation, and the one Sales talks about so often you begin to wonder whether they’re on the payroll.

Naturally, those are the customers everyone wants to keep happy. And I understand why.

If someone is sending your business a million dollars a year, you’re probably not going to respond by asking them to take a number and wait in line.

But what does it actually cost to keep that customer happy?

A million dollars in sales sounds wonderful until you discover the customer needs special pricing, custom deliveries, extra support, and what appears to be a personal emergency response team. I’m not saying they’re a bad customer.

I’m saying we should probably do the math before naming a conference room after them.

Let’s meet two customers

Imagine a company with two major customers. Customer A purchases $1 million a year. Customer B purchases $700,000. If we’re ranking customers by revenue, Customer A wins. There’s not much debate.

But Customer A places frequent small orders, negotiates discounts, requests expedited shipments, and occasionally treats customer service like a personal concierge. Customer B orders consistently, uses standard products, pays on time, and generally doesn’t create much excitement.

Which customer is more profitable?

We don’t know yet.

That’s the interesting part. Revenue tells us how much business we’re doing with someone. It doesn’t necessarily tell us how much money we’re making from the relationship.

First, let’s make sure the revenue number is real

I don’t mean anyone is inventing sales. I mean the amount on the original invoice isn’t always what the business ultimately earns. There may be discounts, rebates, returns, credits, or contractual adjustments.

Suppose Customer A receives $1.1 million in invoices but also $100,000 in discounts and credits. The revenue we’re analyzing is $1 million. Net customer revenue = gross billed revenue − discounts − rebates − returns and other revenue adjustments.

Depending on how the company records revenue, those adjustments may already be reflected in the accounting system. I’m not suggesting we subtract them twice. I want numbers that reconcile to the financial statements, because if the customer report and general ledger disagree, I can almost guarantee which question will derail the meeting. It won’t be the interesting one.

Now let’s talk about gross profit

Gross profit is what’s left after subtracting the direct cost of products or services sold. Customer A generates $1 million in net revenue and costs $650,000 to supply. That’s $350,000 of gross profit, or 35%. Customer B generates $700,000 and costs $455,000 to supply. That’s $245,000 of gross profit. Also 35%.

So far, Customer A generates more gross profit. Case closed?

Not quite.

We haven’t accounted for everything the business does to keep those customers supplied and happy. I have yet to encounter a business where fulfilling the order is the only thing that happens between receiving a purchase order and getting paid.

The cost of keeping a customer happy

Customer A requests expedited shipping. Someone arranges it, and perhaps the company absorbs the freight. Frequent small orders mean more picking, packing, processing, and delivery activity. Add custom reports, extra support calls, and returns. Individually, none may look expensive. Together, they can change the economics.

Finance calls this cost to serve: the cost of fulfilling and supporting a customer relationship beyond the product or service costs already counted in gross profit.

Think about planning a birthday party. The cake, food, and decorations are the easy part. Then come the extra store trip, last-minute supplies, and the mysterious additional expenses because someone has decided the party absolutely requires a balloon arch. The original budget didn’t necessarily include everything the event would require.

And before anyone worries we need to track every paper clip, we don’t. I’d start with costs large enough to affect a decision.

Let’s put the numbers together

Annual customer economics	Customer A	Customer B
Net revenue	$1,000,000	$700,000
Product or service cost	$650,000	$455,000
Gross profit	$350,000	$245,000
Gross margin	35%	35%
Additional cost to serve	$210,000	$55,000
Customer contribution	$140,000	$190,000
Contribution margin	14%	27.1%

Illustrative figures, not actual company results.

Well, look at that.

Customer A brings in $300,000 more revenue. Customer B contributes $50,000 more toward the company’s remaining costs and profit. The smaller customer is suddenly looking pretty attractive.

Numbers have a way of interrupting a perfectly good narrative.

What customer contribution means

Customer contribution = net revenue − product or service cost − additional cost to serve. For Customer A, $1,000,000 − $650,000 − $210,000 = $140,000. For Customer B, $700,000 − $455,000 − $55,000 = $190,000.

That remaining contribution helps cover other business expenses and ultimately profit. It isn’t necessarily the customer’s final net profit. We haven’t allocated every corporate expense, and that’s intentional.

Allocated cost isn’t the same as avoidable cost

Suppose we allocate another $170,000 of corporate overhead to Customer A. The report now shows a $30,000 loss. Someone may suggest dropping the customer immediately. But what happens to that $170,000 if Customer A disappears? Does the rent go down? Does the accounting software become cheaper? Probably not by much.

Eliminating the account could cost the business $140,000 in contribution while leaving most overhead intact. Congratulations.

We’ve made the report look cleaner and the business poorer.

Allocated costs are expenses assigned using a method, perhaps revenue share. Avoidable costs are expenses the business could genuinely eliminate by changing or ending the relationship. If I’m deciding whether to keep a customer, I need to know which revenue and costs actually change.

Capacity matters too. If Customer A occupies production capacity that could serve a more profitable account, that’s different from having idle capacity. Same customer.

Different decision.

How I’d calculate cost to serve without losing my mind

This is where someone suggests an elaborate activity-based costing model. Sometimes that’s appropriate. Before creating a spreadsheet that requires its own employee handbook, I’d ask which activities consume meaningful resources, which customers drive them, and whether we can measure the difference.

For distribution, I’d examine shipments, order lines, returns, and freight. For services, support hours and customer-specific work. For manufacturing, setup time, production changes, and special packaging.

If warehouse handling costs $100,000 annually and the operation processes 10,000 comparable order lines, average handling cost is $10 per line. A customer with 1,000 lines would receive roughly $10,000 of handling cost. That assumes comparable effort; I’d refine it if some lines were materially more complex.

I’d also check whether the cost was already included in product costs. Counting it twice doesn’t make the model sophisticated. It just makes the customer look worse.

Some customers are expensive in ways gross margin doesn’t show

Payment terms are a good example. One customer pays in 30 days, another in 90. The second effectively asks the business to finance an extra 60 days of sales.

At $1.2 million in annual sales, assuming even sales throughout the year, that’s $1,200,000 ÷ 365 × 60 = approximately $197,260 of additional receivables. At an illustrative 10% annual financing cost, carrying that incremental balance costs about $19,700 a year.

That isn’t automatically an accounting expense booked against the customer. It’s an economic estimate, dependent on financing arrangements, cash reserves, and actual collections. But it belongs in the conversation. A profitable sale and timely cash collection aren’t the same event.

I explore this in Working Capital Forecasting: Why the P&L Can Be Right and Cash Still Be Wrong. Anyone who has paid for a school field trip and then waited for everyone to reimburse them understands the general concept. The expense happens now. The money comes back whenever the universe cooperates.

Should we get rid of the least profitable customers?

Not so fast.

The purpose of customer profitability analysis isn’t to produce a list of customers to fire. It’s to understand which relationships could work better.

Maybe Customer A needs a minimum order quantity, an expedited-shipping fee, or updated pricing. Maybe the contract hasn’t kept up with labor and freight costs. Or maybe the account provides steady demand, supports purchasing volume, or opens a market that matters.

Those benefits can be real. I’d just want management to explain them rather than attach a vague label saying strategic account and call the matter settled. What’s the benefit? How will we recognize it? When will we reassess? We don’t need to be cynical. We need to know what we’re agreeing to.

Here’s how I’d build the analysis in practice

Start with customers that matter

I wouldn’t analyze every account on day one. I’d start with the largest customers, those with unusual service requirements, and relationships where revenue is growing but margin isn’t following.

Reconcile revenue and gross profit

Match sales, credits, discounts, and rebates to accounting records. Subtract consistent direct product or service costs and examine product mix. Two customers with identical revenue can have very different gross profit because they buy different things.

Trace meaningful service costs

I’d work with Sales, Operations, Customer Service, and anyone who understands what fulfilling the relationship involves. This is one of those occasions when Finance benefits from leaving its office. The spreadsheet may tell me a customer has 150 shipments. The warehouse manager may tell me why those shipments are a problem.

Calculate contribution and test assumptions

Subtract identified cost to serve from gross profit. Divide contribution by net revenue for the contribution margin. Then test estimates. If profitability depends entirely on assuming support hours are unusually low, I’d want that uncertainty visible.

Consider cash, capacity, and the actual decision

Review payment behavior, special inventory commitments, and scarce capacity. Then ask what management can change: pricing, service levels, order minimums, payment terms, or the contract itself. Sometimes the analysis confirms the relationship works well. That’s useful too.

What I’d actually show management

I wouldn’t present a 40-column spreadsheet unless someone specifically requested one. And even then, I’d have questions.

For each major customer, I’d show net revenue, gross profit, cost to serve, contribution, contribution margin, and material working-capital considerations. Then I’d highlight the exceptions: who contributes most, whose service costs are rising, and which contracts need attention.

I’d also show trends. One year of low contribution might reflect a temporary implementation cost. Three years of declining contribution could mean a contract needs work. That’s the same philosophy behind my article on what CFOs actually need in a monthly management reporting pack. Reporting should make a decision clearer. Otherwise it’s just an organized collection of numbers.

The customer I’d want to keep

After all this, I might still conclude Customer A is worth keeping. Maybe service costs can fall, pricing can improve, or the relationship supports valuable capacity. Or perhaps Customer B really is the better relationship. Either answer is fine if the economics support it.

What I wouldn’t want is to call Customer A our best customer simply because it sends the biggest invoices. Revenue matters.

Relationships matter.

So does what’s left after delivering everything we promised.

And if we’re going to name a conference room after a customer, I’d at least like to know whether they’re helping pay for the building.

If your business is growing but the profit isn’t following, I can help examine the economics behind your customers, pricing, and costs. My forecasting and planning work starts with understanding what the numbers mean before deciding what to do with them.