Lesson 7 · Do & interpret
You have never seen this number about your own company. Most CEOs discover that a fifth of their customers carry all of the profit — and that the rest of the book is quietly consuming it.
11 minutes · Your own customer export · Artifact: your concentration curve
Profit does not. In lesson 6 you saw why that gap exists. Now you are going to measure it in your own business.
This takes about twenty minutes and one export from your accounting system: revenue by customer for the last twelve months. Not gross margin, not an allocation — just what each customer bought. One row per customer.
A fair question: if you only give the tool revenue, how can it tell you anything about profit? The answer is the whole reason the tail exists.
A large part of what it costs to serve a customer arrives with the account itself, not with its size. Order handling, quotes, credit, service calls, freight minimums, the changeover, the phone call about the changeover. A customer buying $40,000 a year consumes a similar amount of that as one buying $4M. Spread evenly across everything — which is what a P&L does — it disappears. Charged to the accounts that actually cause it, the small ones stop covering their own keep.
So three figures off your P&L, plus your own customer list, produce a real curve: your blended gross margin, your overhead as a percent of revenue, and how much of that overhead is driven by the number of customers rather than their size. The shape comes entirely from your book. Only the level comes from the assumptions, and the tool reconciles its total against what you actually earned so you can tell whether you got them right.
Twelve months, one row per customer. A CSV or a paste straight out of Excel. Do not clean it up or group anything — the messiness is the point, and a tidied list hides exactly what you are looking for.
It sorts your customers best-first, charges each one its share of cost to serve, and draws the cumulative curve. Strip the names first if you would rather not have them on screen — the arithmetic does not need them.
Gross margin, overhead, and the share of overhead that comes with the account rather than its size. Then check the reconciliation line: if the operating profit it computes is not close to what you really earned, fix the inputs before you believe the curve.
Where the curve crosses 100%, how high it peaks, and what the tail gives back in dollars.
The tool
Free, no account, and genuinely private — there is no upload and no server. Your customer list is read by this page and nothing else; you can disconnect from the internet and it still works. Closing the tab erases it.
Find the point where your curve first reaches 100% of profit. The percentage of customers it took to get there is your reading.
Then find the highest point the curve reaches — your peak. Anything above 100% is profit your best customers made and the rest of the book gave back. A peak of 120% means the tail destroyed a sixth of what the core earned. Write the peak down as a percentage and as dollars; lesson 9 uses both.
Second read
Pick one item you sell a lot of. Find what your best customer pays for it and what your worst customer pays for it. That spread is your pricing corridor.
If it is wider than about ten points and nobody can explain why on volume, that is not a pricing problem. It is a governance problem, and it is usually the fastest money in the business.
| Customers to reach 100% | Reading | What it means for you |
|---|---|---|
| 10% or fewer | Extreme | You are running a small profitable company that is carrying a large unprofitable one. That is real opportunity, but it is also concentration risk any buyer or lender will price. |
| 10% – 20% | Classic | The most common shape. Your tail is consuming money you are already earning, and the fix is well understood: pricing, cost to serve, and a small number of managed exits. |
| 20% – 30% | Diluted | Lots of mid-sized accounts, no dominant tier. Exits will not move much here. Your money is in pricing discipline and what it costs you to serve. |
| Above 30% | Flat | Either genuinely unconcentrated, which is rare, or your data is grouped too coarsely to show the truth. Re-run it at invoice level before you conclude anything. |
Third read
Name the account you are proudest of. Find it on your curve.
A meaningful share of the time it sits on the downward slope — in the part of the book that destroys profit rather than makes it. That single check tends to land harder than any argument I could make here.
Mid-sized, three regions, earning 8.3% when the sector norm was closer to 12%. The commercial team wanted forty new accounts. Nobody had checked what the existing ones were paying.
Thirteen percent of customers produced 109% of profit. The widest part of the pricing corridor was not big accounts versus small ones — it was customers acquired before and after a price-book revision that had never been applied backwards. About $9M of revenue was still priced on the old book.
They added $6.1M of earnings over seven quarters on revenue that grew 1.5%. No new customers were required. One of the three regions went backwards while they did it, which is covered in lesson 9.
A composite drawn from real engagements and 30+ years of operating experience; company details are disguised and figures are representative, not audited results of a single client.
Write these down. Lesson 9 uses them to size the prize sitting in your own customer book.
Next — lesson 8
