How do you measure concentration so a buyer cannot rewrite it?
Measure sales and gross profit, not just invoices. A 28% revenue customer at 8% gross margin is a different risk than a 28% customer at 42% gross margin. Measure trailing twelve and trailing thirty-six. A new 40% account is a spike. A decade-long 40% account is a dependency. Measure parts and programs, not only legal entities. Three “customers” that are one OEM family are one customer.
CNC machine shops see this weekly: one medical OEM, three part numbers, one quality engineer who can move the book. Job shop versus proprietary product changes switching costs. The percentage can look identical.
Census CBP 2023’s 284,452 establishments and 68.4% small-plant share (Source: Census CBP, 2023) explain why so many files have a whale. Small shops win a program and never diversify. That history is allowed. It is still priced.
| View | What to show | Why it matters |
|---|---|---|
| T12 sales % | Last twelve months by account | Near-term dependency |
| T36 sales % | Last thirty-six months by account | Whether the whale is new |
| T12 gross profit % | Margin contribution | Whether the whale is the profit |
| Program / part | Top SKUs or jobs | Switching-cost story |
| Contract terms | Term, termination, consent | Earn-out design |
| Payment behavior | AR aging on the whale | Peg and collectability |
Quality of earnings will rebuild these views from invoices. If your teaser says 18% and the GL says 34%, the rest of the recast is now suspect.
How do different buyers price the same whale?
How each buyer type prices a manufacturer is the comparison page. The short version:
An SBA operator inside the $5 million 7(a) box still needs cash flow after a replacement wage (Source: SBA). A whale that can leave next quarter breaks the model. They ask for a note or an earn-out you may not want.
A searcher who liked Stanford’s 2024–25 median print — about $16 million purchase, ~6.2x EBITDA, ~25% margin (Source: Stanford GSB) — still has investors who punish single-account risk. They may want a smaller check and more paper.
A strategic who is the whale, or who sells to the whale, may pay for the relationship and still demand a holdback if they are not the account. A PE add-on may pay a cleaner multiple on the diversified remainder and earn-out the concentrated slice.
IBBA Q1 2026’s 83% of >$5 million deals with three or more offers (Source: IBBA) is an argument to run a process even when one customer dominates. One strategic who “already knows the account” loves an exclusive.
BizBuySell’s 2021–2025 manufacturing 2.75x median SDE and 2.04x–3.59x interquartile (Source: BizBuySell / Iconic) is the Main Street band where many concentrated job shops live. Concentration is a common reason files sit in the lower half.
Which structures actually transfer the risk?
Discount the multiple. Simple, painful, done.
Earn-out on named-account gross profit, with a volume floor, a cap, and a measurement period that matches the contract. Seller financing and earn-outs is the seller-side page. “EBITDA after buyer allocations” is how you work for free.
Holdback or escrow tied to renewal or to a defined period without termination. Consent condition: close after the customer blesses the assignment, with a drop-dead date.
Working capital still matters. A whale who pays in 75 days inflates AR. The peg and true-up should not treat that AR as if it were a diversified book.
What should never go in the teaser?
Do not print the customer logo. Do not print “sole source to [famous OEM].” Do not print a percentage that is hope rather than T12. Do not call cancel-for-convenience backlog “recurring.”
Do print a nameless band: “largest account 28–32% of T12 sales; no other account above 12%.” Do print contract shape: “annual PO, 90-day termination.” Do print the quality certificate that is the switching cost, if you actually hold it.
If federal set-aside work is the whale, say that set-aside status may not transfer. If ITAR is the switching cost, say the registration exists without turning the teaser into a street address.
Can you diversify in time, or is that a different sale?
A year of winning smaller accounts can change the T12 if the whale stays flat. Two quarters of “we are calling on new logos” without bookings will not. Buyers annualize the last six months. They do not pay for your pipeline slide.
Illustrative only: a fabricator at $1.2 million revenue, near the BizBuySell manufacturing median revenue neighborhood of about $1.13 million (Source: BizBuySell / Iconic), with one account at 37% and SDE of $290k. A 2.3x conversation is about $667k of enterprise value before peg and debt. An earn-out of $80k on that account’s gross profit over 18 months might bridge a buyer who otherwise walked. Those dollars are illustrations of arithmetic, not a quote.
Firing the whale to “clean the chart” can drop SDE below a livable wage and still leave you with a smaller shop. Diversify with purchase orders, not with a stunt.
What does QoE do to a concentration story?
It tests whether the whale’s margin is real after expedite, after quality escapes, after tooling amortization. It tests whether revenue was pulled forward. It tests whether a price increase sticks. If you stuffed December to look diversified, the report will say so.
Keep the add-back binder honest so the concentration conversation is the only fight. Two fights at once is how exclusives die.
Sell my manufacturing business is the pillar.
How do you prepare a whale introduction without starting a dual-source?
Agree the list, the script, and who dials. Usually the seller introduces, the buyer listens, and nobody mentions price or paper. The script says continuity, quality system, and a named post-close contact. It does not say “we are shopping the company to everyone.”
If the contract requires consent, counsel drafts the request. Informal phone chats first are how purchasing managers open a second RFQ. If the whale is also a competitor of the buyer, clean-team rules apply.
Do the introduction after an LOI unless the contract forces it earlier. Early introductions without a buyer are rumors with a logo.
What supplier concentration sits in the same file and gets ignored?
A single alloy house, a single resin supplier, or a single plating vendor can be as priced as a customer whale. Show it. Dual-source facts help. A sole-source heat-treat that takes 14 weeks is a diligence question. Hiding it next to a hidden customer is how rooms feel dishonest.
Illustrative only: 31% customer, 22% of GP, annual PO with 60-day termination, no other account above 11%, dual-source raw. Earn-out sized to a slice of that GP over 18 months. Process keeps two parties. The file that hid 31% as “about 20%” keeps none.
Should you accept a holdback that is really an earn-out with worse rights?
Read the release conditions. If release depends on the buyer’s “satisfaction” or on EBITDA after allocations, it is a bad earn-out. If release depends on a named account remaining within a band you can audit, it may be a holdback you can live with. Information rights either way.
How do you keep the whale from becoming the only story in the CIM?
Show the rest of the book with the same honesty. A 31% account next to a diversified remainder is a structured deal. A 31% account next to silence about the other 69% is a teaser that forgot to sell the company. Buyers will still price the whale. They should also see whether the remainder is real.
If the remainder is also concentrated in one industry, say the industry percentage. Automotive-heavy or aero-heavy mix is not the same as one legal entity at 31%, but it is still a mix fact.
Illustrative only: 31% whale, next account 9%, industry mix split across three end markets. Earn-out on the whale only. Remainder supports a cash number. That is a company with a whale, not a whale with a building.
How should you treat a whale that is also your landlord or your scrap dealer?
Related-party whales are two stories. Normalize the rent or the scrap price. Disclose the relationship. A 28% customer who also owns the building is a consent, a lease, and a concentration problem in one body. Price all three.
How should you treat a whale that pays slowly but never leaves?
That is concentration plus a peg and collectability story. Age the AR. Do not call 75-day pay “sticky loyalty” without showing the cash cycle. Lenders will not.
A last whale habit: keep a one-page contract extract — term, termination, consent, payment days — in the room before names. Buyers price clauses. They should not wait for the full unredacted PDF to learn the account can leave in 30 days.
One more whale note: if two legal entities are one buying group, combine them on the schedule. Split names are how 18% plus 16% hides a 34% story. Buyers will combine them. You should go first.
If the whale’s quality engineer already walks the floor monthly, plan the announcement around that visit so a buyer tour and a customer audit are not the same Thursday.
Educational information only. Contract consents and antitrust questions belong to counsel. No structure guarantees the whale stays.