Under standard cost accounting, a product or customer is only profitable once you subtract its variable costs, its share of overheads, its returns and discounts, and its acquisition cost from the revenue it brings in. Revenue ranks by size. Profitability ranks by what is actually left once every real cost is accounted for.
In my experience, most founders can tell you their top three customers by revenue without checking a spreadsheet, but far fewer can tell you their top three by profit, and the two lists are rarely the same. That gap is worth closing, because it's the difference between growing the parts of the business that are actually paying for the rest of it, and growing the parts that only look like they are.
What customer and product profitability analysis actually mean
Customer profitability analysis and product profitability analysis are the same idea applied to two different units of measurement. One asks: once we service this customer properly, what's left? The other asks the same question of a product or SKU rather than an account.
Corporate Finance Institute defines customer profitability analysis as looking at the activities and expenses involved in servicing a particular customer over a given period, rather than simply the revenue that customer generates (https://corporatefinanceinstitute.com/resources/accounting/customer-profitability-analysis/). That framing matters: it's a costing exercise, not a sales-ranking exercise. The same logic runs the other way for products. A best-seller and a most-profitable product are two different rankings, and a business that only tracks the first one is flying on partial instruments.
Why revenue misleads
Revenue is the top line: what came in the door. Profitability is what's left once everything that had to be spent to earn it is taken out. A customer who orders often, in bulk, at a heavily negotiated price, and who returns a meaningful share of what they buy, can post more revenue than a smaller account that pays full price, rarely returns anything, and needs almost no hand-holding, while making the business less money on paper.
In my experience running commercial teams, the highest-revenue account in a business is not reliably the most profitable one, and the gap between the two rarely shows up until someone actually runs the numbers per account rather than per period. The same pattern holds at product level: the product that sells the most units is not automatically the one contributing the most to the bottom line, particularly once you load in the cost of holding stock, processing returns, and supporting the customers who buy it.
The calculation, step by step
Start with contribution margin. Wikipedia's summary of the standard accounting definition puts contribution margin as the selling price per unit minus the variable cost per unit (https://en.wikipedia.org/wiki/Contribution_margin). That figure isolates what a single sale actually contributes before any shared costs are considered, which is the right starting point but not the finish line: it tells you what one unit earns in isolation, not what the customer or product costs the business once everything around that sale is included.
Allocate a fair share of the overheads that unit doesn't carry on its own. Customer service time, account management, warehousing, platform and payment fees, all sit somewhere in the business whether or not they're attached to a specific sale. My rule is to allocate these on actual usage, not on a flat headcount-based guess: a customer who calls the support line weekly is not costing the business the same as one who never does, even if both bought the same product at the same price.
Account for returns and discount rates. In my experience, a return doesn't just reverse the sale; it usually carries a restocking, processing or refund-handling cost on top of the lost revenue. A discount reduces the revenue side of the equation directly. Both need to be applied at the level of the specific customer or product, not averaged across the whole book, or the analysis quietly hides exactly the accounts and lines it was built to expose.
Amortise customer acquisition cost over the relationship, not the first order. Corporate Finance Institute notes that customer acquisition cost is commonly used alongside customer lifetime value to judge the return generated by a new customer (https://corporatefinanceinstitute.com/resources/valuation/customer-acquisition-cost-cac/); that source pairs the two metrics but doesn't itself walk through spreading the cost across a repeat-purchase profile, so this next part is our own method rather than a citation. Take an invented, illustrative case: a customer who cost £40 to acquire and buys once at a £15 margin looks unprofitable on that first order alone. Spread the same £40 across four orders at £15 margin each over the year instead, and the acquisition cost is more than covered by the second order onwards. Charging the full acquisition cost against the first order, rather than spreading it across the purchases that actually followed, is one of the more common ways this analysis gets read backwards.
An illustrative example (not a real business)
To make the mechanics concrete, picture a brand selling two products. Every figure here is invented for illustration, built from round numbers, and describes no client of ours, past or present.
Product A sells for £50 with a variable cost of £30, a contribution margin of £20 per unit, and sells 1,000 units a month, for £50,000 in monthly revenue. It carries a 15% return rate (150 units, each costing £8 in restocking and refund-handling) and a heavier support load, at an average of £3 in support cost per unit sold, because it's the product new customers ask the most questions about before and after buying. Total monthly cost once returns and support are added: £30,000 in variable cost, £1,200 in return-handling, £3,000 in support, leaving £15,800 in monthly profit.
Product B sells for £30 with a variable cost of £12, a contribution margin of £18 per unit, and sells 400 units a month, for £12,000 in monthly revenue. It carries a 2% return rate (8 units, at the same £8 handling cost) and almost no support load, at £0.50 per unit, because the customers who buy it already know exactly what they're getting. Total monthly cost: £4,800 in variable cost, £64 in return-handling, £200 in support, leaving £6,936 in monthly profit.
On revenue alone, Product A wins by more than four times over. Once returns and support are allocated properly, Product A is still the more profitable line here in absolute terms, £15,800 against £6,936, but its profit margin has narrowed from a 40% headline contribution margin to just under 32% of revenue, against Product B's near-58%. In a business where Product A's return rate or support cost ran even higher, or its volume advantage were smaller, that ranking could flip entirely. Nothing about these figures is a claim about any real product mix; they're here to show why the calculation has to be run properly, not skipped in favour of the revenue report everyone already has open.
A closing thought
None of this is complicated arithmetic. In my experience it's mostly a discipline problem: businesses track revenue by default because their sales systems already do it for them, and profitability by customer or product usually has to be built on purpose. If working through your own numbers this way raises more questions than it answers, that's a conversation worth having rather than a spreadsheet to keep wrestling with alone. Our Strategic Counsel work exists for exactly this kind of diagnostic question, and we've written elsewhere about the hidden revenue that sits inside a luxury wellness or aesthetics business, where the same distinction between what comes in and what's actually kept applies just as directly.
FAQ
Is customer profitability analysis the same as customer lifetime value? In general use, no. Lifetime value typically estimates the total revenue a customer is expected to generate over the relationship. Customer profitability analysis typically asks what's left after the real cost of servicing that customer, in a given period, is taken out of what they've actually spent. They're related, but one is usually treated as a revenue forecast and the other as a profit calculation.
What's the difference between gross margin and contribution margin? In standard accounting practice, gross margin is usually calculated at the level of the whole business or a product category, after cost of goods sold, while contribution margin is calculated per unit, per product, or per customer. In my experience, contribution margin is the figure this kind of analysis is best built on, because it's granular enough to compare one product or account against another.
How often should a business rerun this analysis? My rule is at least once a quarter for the accounts and products that matter most, because a customer's ordering pattern, return rate and support needs can shift well before it shows up in the headline numbers.
Do very small businesses need to bother with this? In my experience, yes, and often sooner than founders expect: a business with a handful of customers or a short product range has less room to absorb one unprofitable relationship, not more, because there's nothing else in the mix to dilute it.
Kirsty Newman is Founder of The Boutique Consultancy. She spent almost two decades at some of the largest corporate powerhouses in beauty and luxury, including YSL Beauty, Giorgio Armani Beauty, Lancôme and Rimmel (L'Oréal Group and Coty). She went on to work with a number of startups as they scaled, before striking out on her own. Across this time, she has run seven-figure P&Ls and teams of 16, with direct engagement in venture capital funding and expansion.