Whale Curve: Definition, How to Build One, Live Simulator
Rank every account from best to worst and plot the running total of profit. A minority climbs far above 100%. Then a long tail gives it all back. The Whale Curve is the one picture that makes hidden margin impossible to ignore - and the starting point for getting it back.
Watch your profit change shape as the method gets honest.
Every company has a whale curve: a few relationships create the profit, a long tail quietly destroys it. What that curve looks like depends on three choices. Change them below and watch who really makes you money.
Illustrative model, synthetic dataThe levers companies pull when the curve gets honest.
Seeing the shaded area is step one. Recovering it is a short list of well-known moves. We help you pick and quantify the right ones for your book.
There are only about eight of them, and none is exotic. A loss-making account is usually underpriced for the way it buys rather than underpriced overall, so the first move is to price the behaviour that drives the cost: small orders, rush shipments, heavy support, frequent returns. The second is to stop subsidising the smallest transactions with a minimum order value or a small-order charge, which changes buying patterns faster than any conversation about price. Then come the terms and service levels, where delivery frequency, payment terms and support intensity are re-set to what the account actually earns; and the channel, where a small high-touch account moves to self-serve, to a distributor, or onto a consolidated delivery run. On the supply side the same list continues: rationalise the product references that never carry their cost to serve, attack the cost to serve itself by consolidating deliveries and automating the admin the tail quietly consumes, move commercial time off the value-destroying accounts and onto the head of the curve and its lookalikes, and right-size the capacity the tail was burning so the saving reaches the profit and loss account instead of reappearing as unused capacity. Exit is on the list, and it is last, because almost everything above it recovers more margin at less cost.
Re-price the loss-makers
Price to the real cost-to-serve, not to revenue size. Charge for the behaviour that drives cost: small orders, rush, heavy support.
Set minimum order values
Stop subsidising tiny, high-touch orders with thresholds and small-order fees.
Re-negotiate terms and service levels
Align delivery frequency, payment terms and support to the account's economics.
Shift channel, serve to cost
Move small accounts to lower-cost channels: self-serve, distributor, consolidated delivery.
Rationalise the product tail
Cut or reprice the SKUs that never carry their cost-to-serve.
Cut the cost-to-serve itself
Consolidate deliveries, batch orders, automate the admin the tail quietly consumes.
Reallocate sales effort
Move commercial time from value-destroying accounts to the profitable head and its lookalikes.
Manage unused capacity
Right-size the resources whose practical capacity the tail was quietly burning.
This curve is illustrative. Yours is sitting in your ERP.
A Free Profit Check pulls your whale curve from your own ERP or accounting export. One export, no project, and you will know which quarter of your customers to talk about.
One line, three truths about your customer base.
Standard reporting gives you a single profit number. The Whale Curve unpacks that number into a shape - and the shape tells you who built the profit, who is borrowing against it, and who is quietly destroying it.
Kaplan and Narayanan put the pattern at roughly 20% of customers generating 150 to 300% of total profit, the middle breaking even, and the bottom fifth giving much of it back.
That figure comes from Robert S. Kaplan and V. G. Narayanan, "Measuring and Managing Customer Profitability", Journal of Cost Management, 2001, listed in the sources at the foot of this page. It is worth naming the source, because two different versions of this pattern now circulate and they are not interchangeable. The one above is measured over customers. A second set is quoted over products, where a minority of lines carries the profit and a long tail of slow-moving references destroys it. The product tail behaves the same way, but its percentages come from different studies on different portfolios, so a number taken from one and applied to the other is not evidence of anything. When you see a figure quoted without saying which of the two it counts, treat it as unattributed.
Your P&L adds everything up. That is exactly the problem.
A profit-and-loss statement aggregates. It shows total margin but not which accounts created it and which destroyed it. The costs that sink the tail - picking, shipping, returns, rush orders, account admin - are real, but they are averaged across everyone. ABC and TDABC stop the averaging: every activity is traced to the account that drives it.
How to build a whale curve, step by step.
A whale curve needs two columns and one sort. You need each customer's net contribution, which is revenue minus cost of goods minus cost to serve, and you need the profit the whole book actually reported. Everything else is a running total. The work is not in the chart, it is in the cost-to-serve figure that feeds it: until operational cost is attributed to the account that consumed it, every customer carries the same average and the curve comes out almost flat. Once it is attributed, the same customer base draws a line that climbs well above the profit you reported and then falls back to it. Here are the five steps, in the order you do them.
- Compute net contribution per customer. For every account, take revenue minus cost of goods minus cost to serve. Cost to serve is the sum across activities of the minutes that account consumed multiplied by the cost per minute of each activity. The full calculation is written out here.
- Sort the customers from highest to lowest net contribution. Rank the whole book on that one column, best account first, worst account last. The ranking is the horizontal axis and nothing else determines the order.
- Build the running total. Add each account to the one above it, so every row carries the cumulative net contribution of itself and everything ranked before it, down to the last account.
- Express the running total as a percentage of total reported profit. Divide each cumulative figure by the profit the whole book actually reported. That percentage is the vertical axis. The horizontal axis is your customers in rank order, most profitable first.
- Read the crest and the tail. The crest is the profit you would keep serving only the accounts that pay their way. The decline from the crest back to 100% is what the tail gives away.
Because the last account always brings the running total back to the profit you reported, the curve ends at exactly 100% however steep it got on the way. That is the whole diagnostic: the height of the crest is the profit the head created, and the drop from the crest to 100% is the profit the tail took back.
| Account | Net contribution | Running total | % of reported profit |
|---|---|---|---|
| 1 | 48,000 | 48,000 | 48% |
| 2 | 31,000 | 79,000 | 79% |
| 3 | 22,000 | 101,000 | 101% |
| 4 | 15,000 | 116,000 | 116% |
| 5 | 9,000 | 125,000 | 125% |
| 6 | 4,000 | 129,000 | 129% |
| 7 · the crest | 1,000 | 130,000 | 130% |
| 8 | −6,000 | 124,000 | 124% |
| 9 | −11,000 | 113,000 | 113% |
| 10 · the tail ends | −13,000 | 100,000 | 100% |
Plot the last column against the account rank and you have the curve. Seven accounts out of ten build 130% of the reported profit; the last three hand 30% of it back. Nothing here depends on the size of the business, only on the spread of net contribution across the book, which is why the shape survives from ten accounts to ten thousand. Real books are wider and messier, and the crest usually sits somewhere between a fifth and a third of the way along.
Every account gets one of four verbs.
Once true profit is on the table, the portfolio sorts itself. The Whale Curve is not a verdict on customers - it is an instruction set for what to do next with each one. For a segment-by-segment reading of a real curve, see a deep-dive reading of the whale curve.
Your profit engine. Protect it, deepen it, and win more accounts that look like it. The fastest growth you have is the share you do not yet hold here.
Profitable but leaky. Consolidate orders, simplify the mix and cut cost-to-serve. Small, unglamorous moves that compound across the base.
Underwater on price, not on principle. Re-price, set minimums and charge for the behaviour that drives cost. Most of the tail recovers here.
The last resort, once every other move has failed. A small, deliberate offboarding - never a reflex.
From your data to your curve, in 5-25 days.
A curve is only as trustworthy as the costing behind it. A volume-based or gross-margin view draws a misleading line; a TDABC model draws one you can act on, because every cost is traced to what actually consumes it.
Map the cost structure
We gather financial, operational and transaction data and group it into the activity pools that drive cost - picking, delivery, service, returns, credit.
Trace with time equations
TDABC allocates indirect cost by the real resource each account consumes, not by arbitrary volume keys. The over-serviced and over-discounted finally surface.
Rank, draw, decide
You receive the full ranking, your Whale Curve across every dimension, and a segmented portfolio with the recommended move for each tier.
That is the version we build for you. If you would rather build it yourself, the online TDABC course works through the same three steps on your own data, over three to four weeks.
The Whale Curve, answered.
It is the visual output of a cumulative profitability analysis. Accounts - or products, channels, regions - are ranked from most to least profitable and their cumulative contribution to profit is plotted. The line rises to a peak well above 100% and then declines as loss-making accounts erode the total back toward 100%.
Because a minority of customers typically generate 150 to 300% of total profit. The crest is the profit you would keep if you only served the accounts that pay their way; the decline back to 100% is the margin the tail gives away.
Yes. The same cumulative method applies to any dimension - products and SKUs, sales channels, regions or segments. Each lens usually reveals its own tail of value destruction, which is why we build the curve several ways, not one.
With Activity-Based Costing (ABC) and Time-Driven Activity-Based Costing (TDABC). We trace the true cost of each activity - logistics, service, returns, credit - to the account that consumes it, then rank by net contribution. The build typically takes 5-25 working days depending on data availability.
Segment every account into Grow, Optimise, Reprice or Exit, then model the profit impact of each move before acting - exactly what the simulator above demonstrates. The goal is rarely to fire customers; it is to fix the pricing, complexity or cost-to-serve that put them underwater. That starts with knowing what each customer actually costs to serve; here is how that number is calculated.
The horizontal axis is your customers in rank order, most profitable first and least profitable last, ranked on net contribution. The vertical axis is the running total of that net contribution, usually expressed as a percentage of the profit the whole book reported. Because the last account always brings the running total back to the reported figure, the line ends at exactly 100% however high it climbed on the way. The same two axes work for products, channels or regions; only what you rank changes.
One row per customer, with revenue, cost of goods and cost to serve. Add a net contribution column, sort the sheet descending on it, add a running-total column, then divide that running total by the sum of the whole column to get the percentage. Plot the percentage as a line chart. A spreadsheet does this well once. It stops working when the model has to be rerun every month, cope with thousands of accounts and be explained line by line to people who will act on it, which is the point where the cost-to-serve figures behind it need a model that maintains itself.
They share the ranking idea and differ in what they plot. A Pareto chart accumulates a quantity that is always positive, usually revenue or volume, so its cumulative line can only rise and it approaches 100% from below. A whale curve accumulates profit, and profit per account can be negative, so the line rises past 100%, peaks at the crest, and then falls back to 100% as the loss-making accounts are added. That descent is the whole point and a Pareto chart cannot show it.
Where the whale curve comes from.
- Book
Jonathan L. G. Byrnes, Islands of Profit in a Sea of Red Ink. Portfolio / Penguin, 2010.
Popularised the whale-curve view: a minority of customers earns well above total profit while a long tail gives it back.
- Paper
Robert S. Kaplan and V. G. Narayanan, "Measuring and Managing Customer Profitability". Journal of Cost Management, 2001.
The academic treatment of cumulative customer-profitability curves.
- Book
Robert S. Kaplan and Robin Cooper, Cost & Effect. Harvard Business School Press, 1998.
The foundational text on Activity-Based Costing.
- Book
Robert S. Kaplan and Steven R. Anderson, Time-Driven Activity-Based Costing. Harvard Business Review Press, 2007.
TDABC, the costing method that makes the curve trustworthy.
External evidence · as at 9 August 2026
What companies published, in their own words
4 disclosures, 4 markets, published between 4 June 2026 and 29 July 2026.
A whale curve is an argument about the tail: how much range a business carries, and what the rest of it pays for that range. Here are companies making their own version of that argument in public. Each line is quoted exactly as it was released, with its date and a link to the source.
Geox
IT · 29 Jul 2026
“Cost savings of EUR 19.1 million lifted adjusted EBIT margin to 2.1 percent from 0.2 percent while net sales fell 11.4 percent.”
Source: geox.biz
Signify N.V.
NL · 24 Jul 2026
“adjusted EBITA margin of 6.1% against a confirmed full year range of 7.5 to 8.5% and booked a further EUR 31 million of restructuring costs.”
Source: globenewswire.com
Nobia AB
SE · 17 Jul 2026
“reported second-quarter net sales of SEK 1,498m, an organic decrease of 1 per cent, with adjusted operating profit increased to SEK 128m and adjusted gross margin up to 39.8 per cent, and pointed to additional run-rate savings expected from its cost programme from the third quarter.”
Source: news.cision.com
Treasury Wine Estates
AU · 4 Jun 2026
“set out at its Investor Day a portfolio rationalisation from 76 brands to under 30 and from 1690 individual product lines to 738, delivering $100m p.a. in cost reduction to be fully realised by F29.”
Source: announcements.asx.com.au
Refreshed weekly from a scan of 203 companies. Every line is quoted from the source it links to, and no figure has been added to it. The companies named here are not clients of Cost and Profitability Consulting.
Find out where your hidden margin lives.
The Profit Check takes 10 minutes and no data upload. You get a personalised read on your profitability - and a sense of how steep your own whale really is.
Proof
A distributor in New Zealand. €1.335M of cost-to-serve made visible, then halved, and 830 loss-making customers brought down to 295.
Read the case study →Who you would be talking to
Miguel Guimarães, Founding Partner
Cost and profitability practitioner for 25+ years. Presented the Damco cost-to-serve case at Managing for Profit (Amsterdam RAI, December 2009), on the same programme as Robert S. Kaplan.
Call +351 910 313 731
The whale curve ranks every customer from most to least profitable on net profit (gross margin less the real cost to serve) and plots the running total. At CaP the line climbs to a peak well above the firm's actual profit, then bends back down as the loss making tail is added. Reading left to right shows exactly where the money is made and where it is given back.
| Rank | Customer | Net profit (EUR) | Cumulative profit (EUR) |
|---|---|---|---|
| 1 | National Retail A | 259,874 | 259,874 |
| 2 | Regional Chain B | 171,457 | 431,331 |
| 3 | Wholesaler C | 145,552 | 576,883 |
| 4 | Contract H | 114,530 | 691,413 |
| 5 | Foodservice D | 52,895 | 744,308 |
| 6 | Convenience E | 4,083 | 748,391 |
| 7 | Online pure-play G | 3,488 | 751,879 |
| 8 | Independent F | 1,259 | 753,137 |
| 9 | Small accounts I | -10,222 | 742,915 |
| 10 | New account J | -5,987 | 736,929 |
| Total | 736,929 | 736,929 |
The essentialsa few accounts carry the firm, and the tail destroys 16,208.
Workshops
Bring the method into the room.
One working profitability model, built from real data, that you take home at the end.
Reserve a seat
Reviewed by
Miguel Guimarães
Founding Partner, Cost and Profitability Consulting
More than 150 Time-Driven ABC engagements across 11 sectors since 2010, working within the Kaplan and Anderson framework.
About the author →Published Updated