The seven things that end a deal on the first call
Seven binary checks that kill an acquisition outright. Run them before valuation, because none of them can be priced around.
Seven things can end an acquisition outright, and all seven can be surfaced in a single conversation before anyone has opened a spreadsheet. They are not scored, they are not weighed against the good parts of the business, and they are not negotiated down with price. They are binary. Either the answer is clean or the deal is over.
Why the first list has to be binary
Every scoring system has the same weakness. Give a business enough points for recurring revenue, margin and customer quality and a high score will eventually carry a fatal flaw across the line, because the flaw is worth minus fifteen and the rest of the business is worth plus ninety. That arithmetic is wrong. Some problems do not subtract from a business, they end it, and no amount of quality elsewhere compensates.
So the screen runs in two stages. The first is a set of binary gates that either pass or stop the process, run before any valuation work. The second is the scoring, which only ever runs on what survived. Anything else and you end up doing three weeks of diligence on a company that was never buyable, which is expensive, and worse, you get attached to it while you do.
The seven below are drawn from the German trades and industrial services market, which is where I look. The specific statutes are local. The shape is not, and if you buy small companies anywhere you will recognise most of them, because they all come from the same place: something essential to the business is attached to a person, a permit, or a piece of land rather than to the company you are actually buying.
The seven
The owner is the only qualified person
In a German licensed trade, the company can only stay on the Handwerksrolle if it has a technical manager who holds the master qualification, and that person has to hold a genuinely dominant position in the technical side of the business. If the seller is the only one, the company does not just lose its best technician on completion day. It loses the legal basis on which it trades.
One customer is more than a quarter of revenue
Above roughly a quarter, that customer is not a customer, they are a counterparty with veto power over your returns. Look for the hidden version too: five separate accounts that turn out to be five sites of one group, or one procurement decision-maker who happens to control several of them.
The books cannot be reconstructed
Not "the books are messy", which is normal and fixable. Unreconstructable: revenue that only exists as cash, a set of filed accounts that cannot be tied back to bank statements, or an owner who explains that the real numbers are better than the declared ones. That last one is an offer to buy a fraud and a valuation you cannot defend.
Contamination on land the company owns
Under the German soil protection act the owner of a site is liable for cleaning it up as the party responsible for its condition, whether or not they caused the problem. The constitutional court capped that liability, in the normal case, at the market value of the land as if it were clean. That is a cap, not a comfort: the worst realistic case is that the property is worth nothing and you pay its full clean value again to remediate it.
The staff transfer rule breaks your structure
Buy a German business as a set of assets rather than shares and the employment relationships transfer to you automatically anyway. Terminating someone because of the transfer is void, the seller stays jointly liable for a year on obligations already due, and each employee can object in writing within one month of being informed. Plan an asset deal around leaving part of the workforce behind and you have planned something that does not exist.
A licence is attached to the leaver, not the company
The refrigeration rules are the clearest example. Under the current EU F-gas regulation the company itself needs a certificate on top of the individual certificates its technicians hold, and refrigerant can only be sold to a business that qualifies. Lose the only certified person and you have a company that cannot legally buy the substance its entire service book depends on.
The business is bankable and the seller wants all cash now
This one is not a defect in the business, which is why it gets missed. A clean, profitable, well-documented company whose owner wants maximum cash at completion is a competitive auction. Somebody with a lower cost of capital and no interest in operating it will pay more than you should. That is a fine outcome for the seller and a bad use of three months.
Read them together and a pattern shows up. Five of the seven are about something the company depends on that the company does not actually own: a qualification, a customer relationship, a certificate, a piece of contaminated ground, a workforce that has statutory rights independent of your plans. The other two are about whether the numbers are real and whether you are the right buyer at all.
Almost everything that kills a small acquisition is a dependency the balance sheet does not show, because it was never the company's to begin with.
Curable, incurable, and the difference that matters
The seven are not equally final, and treating them as if they were will make you pass on good businesses. Two of them can be cured before completion, which moves the target to a watch list rather than off the list. The rest cannot be cured by anything you are able to do as a buyer.
- Curable
- The condition can be fixed before you sign, by the seller or by you, and you can verify the fix. A second qualified person can be hired and registered. A certificate can be obtained by an existing employee. The deal waits for the fix, it does not proceed alongside it.
- Incurable
- Nothing you can do changes it inside the deal timetable. Contaminated ground stays contaminated. Concentration cannot be diversified in ninety days. Books that were never kept cannot be recreated. A seller who wants all cash does not want your structure.
The distinction sounds obvious and is routinely fudged, usually in one direction. A buyer who likes a business will decide that a curable problem is being cured, sign on that basis, and discover on completion day that the second qualified person took another job in April. So the rule is that a curable gate keeps the target on watch and nothing more. It becomes a live deal on the day the cure is documented, not on the day it is promised.
There is one more asymmetry worth naming. Passing on a business that would have been fine costs you a business. Buying one carrying an incurable gate costs you the purchase price, the years you spend on it, and the option value of everything you did not look at while you were busy. Those are not comparable errors, and a screen should be built knowing which one it is designed to avoid.
| The gate | Curable before completion | What it costs if ignored |
|---|---|---|
| Owner is the only qualified person | Yes, hire and register a second | The legal basis to trade |
| One customer over a quarter of revenue | No, not in ninety days | A counterparty with veto power over returns |
| Books cannot be reconstructed | No | A valuation you cannot defend, or a fraud |
| Contamination on owned land | No | Up to the clean value of the site, twice over |
| Staff transfer breaks the structure | No, it is statute | A plan that does not legally exist |
| Licence sits with the leaver | Yes, certify an existing employee | A company that cannot buy what it sells |
| Bankable seller wants all cash | No, it is about them | Three months in an auction you lose |
Surfacing all seven in one conversation
None of this needs a data room. Each gate has a question that fits into a normal first conversation with an owner, phrased so that the honest answer and the evasive answer sound different. The order matters slightly: the qualification and the customer questions come early because they end the most deals, and the cash question comes last because it is about the seller rather than the business.
- Who else in the business holds the qualification, and how long have they been here
- If you were off for three months, who signs off the technical work
- What share of revenue came from your largest customer last year, and the year before
- Are any of your top accounts part of the same group, or buying through the same person
- Who does the bookkeeping, and are the filed accounts the whole picture
- Do you own the premises or rent them, and what was on the site before you
- How many people are on the books, and are any of them on something other than a standard contract
- Which certificates sit with the company, and which sit with individuals
- What does a good outcome look like for you, and how much of it needs to be cash on the day
The last question does more work than the other eight combined, and it is the one most buyers skip because it feels premature. It is not premature. It tells you whether you are in a negotiation or an auction, and it tells you before you have spent anything. An owner whose answer involves a handover, a legacy, and staff who are looked after is a different transaction from one whose answer is a number.
What the gates deliberately do not do
A screen this blunt will be wrong sometimes, and here is how. It will occasionally stop a business that could have been made to work by a buyer with a specialist advantage, someone who already employs three qualified people and can supply one on day one, or who has remediated contaminated sites before and prices that risk properly. If that is genuinely you, then the gate is not your gate.
What the gates are not is a substitute for judgement on the upside. Clearing all seven says nothing about whether a business is any good. It only says the deal will not die of something structural. The actual question, whether this is a genuinely good company that happens to be run badly, is a separate exercise and it runs afterwards. I have written about that screen in what makes a business worth buying, and about what diligence looks like once a target survives both in due diligence for operators. If you own a business and want to know how these read from the other side of the table, the same list is what I work through when someone approaches me to sell.
One closing note on discipline, since this is where lists like this usually fail. The value of a binary gate is entirely in the fact that it is never overridden. The first time you carry a failed gate forward because the rest of the business is exceptional, you have converted the whole thing into a scorecard with strong opinions, and you will do it again on the next deal with slightly less hesitation. Write them down before you have a live target in front of you, because writing them down afterwards is just describing the deal you already want to do.
The short version
- Seven conditions end a deal outright: sole qualified person, customer concentration above a quarter, unreconstructable books, contaminated owned land, a staff transfer rule that breaks the structure, a licence tied to the person leaving, and a bankable seller who wants all cash now.
- They run before valuation, and they are binary. A gate that can be outvoted by a good score is not a gate.
- Two are curable. A cure puts the target on watch until it is documented, never on the deal path while it is promised.
- Five of the seven are the same problem: the business depends on something it does not own.
- All nine surfacing questions fit inside one normal conversation, and the one about what the seller actually wants is the most useful of them.
Questions I get on this
What level of customer concentration is too high when buying a business?
Can you avoid taking on employees in an asset deal in Germany?
Who is liable for contaminated land, the polluter or the new owner?
Sources: § 7 Handwerksordnung on the registered technical manager; § 613a BGB on transfer of employment on a business transfer; BVerfG, 1 BvR 242/91 (16 February 2000) on the limits of owner liability for contaminated sites under the Bundes-Bodenschutzgesetz; Regulation (EU) 2024/573 on F-gas company and personnel certification. None of this is legal advice.