The cases in this part of the textbook are individual stories. This lesson steps back to the collection they came from: the founder's interview corpus of 60+ failed and struggling buyers. It reports what that corpus says in aggregate, exactly how much weight each number can bear, and why so few of these stories ever reach a dataset you could look up.
Read the numbers here as what they are. They are one practitioner's tally from his own conversations. They are not industry statistics, and you should never repeat them as if they were.
Principle
A number is only as good as its frame. Every figure in this lesson carries its source in the same sentence, and when you use it, yours should too. "Of the failures in one founder's interview corpus" is a true sentence. "Most acquisitions fail because of seller fraud" is not a sentence anyone can support.
The second principle is the reason the corpus exists at all: dead searchers do not fill out surveys. The failures that end worst are the ones least likely to be counted anywhere. A buyer who settled with a non-disparagement clause cannot tell the story. A buyer in bankruptcy is not answering a researcher's questionnaire. So the published picture of how buyers fare is drawn mostly by the people who are still standing. Treat any rosy aggregate with that in mind.
Case
What the corpus is
Since 2024 the founder of Searcher School has held conversations with 60+ buyers whose acquisitions failed or are struggling. Some became the published case record in Buyer Beware; most did not. A subset was recorded and transcribed. The stories were later distilled into a set of anonymized pattern cards, each mapped to the red flag library (RF-01 to RF-12) and to the seven-link failure chain taught in lesson 2.5.1.
That is a convenience sample. Buyers who failed found the founder, or he found them, because he had failed himself and said so publicly. It is not random, it is not weighted, and it was not designed as a study. Its value is depth, not representativeness: hours with each buyer, in their own words, about what they missed.
What it says
Two aggregate findings may be stated, each only in the frame below.
| Finding | The only correct frame |
|---|---|
| About 65 percent of the collected failure stories involved material seller fraud or misrepresentation | The founder's own tally of the failures he has studied. Not an external study, not an industry rate. |
| Roughly 8 out of 10 of the failures studied ended in some mix of litigation, bankruptcy, and a settlement carrying non-disclosure or non-disparagement terms | Anecdotal, the founder's estimate from the failures he has studied. Not a measured rate. |
Hold the first finding next to the published record. Of the stories in Part II, Tom Brennan's tire fraud, Paul Kessler's inventory, Victor Lang's phantom machines, Eric Sandoval's missing expenses and Dan Cole's painted backlog all turned on a seller who misrepresented the business. Alan Brewer's sellers and Devin Price's sellers told the truth, and those buyers lost anyway, to a walk-out and to a cycle. Both kinds of failure are real. The corpus says the first kind is common enough that proof, not trust, has to be your default posture. Lesson 2.5.2 turns that into practice.
Hold the second finding next to what you can read. The reason you rarely hear these stories is in the finding itself: litigation is slow and expensive, bankruptcy is private pain, and settlements often buy silence. The lessons get buried with the companies, and the next buyer walks in believing failure is rare.
Other themes recur across the corpus without a reliable count, and this textbook states them only as patterns: buyers describing deal fatigue as a reason they signed despite doubts; access to employees restricted until days before closing; stress tests sized to a round number instead of to the specific way the business could break; and the personal guarantee turning a company's failure into a household's.
Why the formal data misses it
Look at what the available data can and cannot see.
- Fund studies see funds. The long-running studies of traditional search funds track investor-backed searchers, not self-funded buyers using an SBA loan, and they depend on who reports. A 2025 Yale analysis of data from 12 serial search fund investors, covering 768 completed deals, found returns well below the figures in earlier studies, and its authors described the results as directional at best because a dozen investors could introduce sampling bias (M1). Never apply fund aggregates to a self-funded deal, in either direction.
- Academic failure studies are small. A 2014 study of failed external takeovers of small and medium firms worked from five cases, interviewing sellers and buyers, and traced the failures to the buyers' cognitive biases (B12). That is valuable and honest about its scale. It is also a reminder of how thin the formal record of buyer failure is.
- Loan data sees loans, not stories. SBA data records whether a loan defaulted, not why, and not what the guarantor lost. This lesson quotes no single default rate, because a rate means little without its year, its loan population (all 7(a) loans or only change-of-ownership loans) and whether it is annual or cumulative. Timing also distorts the picture: a 2025 vendor analysis of SBA loan data concluded that pandemic-era stimulus suppressed defaults until a catch-up in 2023 and 2024 (Y14). A loan that is still current can sit on top of a buyer who is draining his savings to keep it current, as Devin Price's case shows.
None of those sources is wrong. Each answers a narrower question than the one you are asking, which is: what tends to go wrong for someone like me, and how would I see it coming?
Steps
Step 1: Use the corpus to set posture, not to forecast
The corpus cannot tell you the probability that your deal fails. It can tell you which questions earlier buyers wish they had asked. Use it for that: as a list of failure modes to check, ordered by how often they came up.
Step 2: Quote it correctly or not at all
If you cite these findings to a partner, a lender, or a spouse, keep the frame attached:
- Say "the founder's interview corpus of 60+ failed and struggling buyers," never "studies show."
- Say "about 65 percent of the failures he studied," never "65 percent of acquisitions."
- Say "anecdotally, roughly 8 out of 10," never "80 percent."
The same discipline applies to every number you put in front of anyone during your search. A figure without a source is folklore.
Step 3: Correct for survivorship in everything you read
When you read a success story, a podcast interview, or a returns study, ask who could not be in the sample. Settled buyers, bankrupt buyers and buyers still grinding through a slow decline rarely appear. Adjust your expectations downward for that silence, not upward for the enthusiasm.
Step 4: Start your own base-rate log
From your first deal book, keep a record of every business you review past the screen:
| Field | Example |
|---|---|
| Deal and date | Deal 14, March |
| Stage reached | CIM, call, LOI, diligence |
| Misrepresentation found? | Yes: add-backs not supported by invoices |
| Red flags found | RF-04, RF-11 |
| Why you stopped or continued | Seller could not document $90,000 of add-backs |
After twenty or thirty deals you will have your own base rate, from your own market, in your own notebook. It will be small and it will be biased in its own ways, but it will be yours, and it will tell you more about your next deal than any aggregate.
Step 5: Talk to the people who did not make it
Before you close, find two or three buyers in your industry or your loan size whose deals went badly, and ask them what they would never skip again. They are harder to find than success stories. That is the point.
Deliverable
A one-page Base-Rate Card that you keep with your deal files:
- The two corpus findings, each written with its exact frame.
- One sentence on what each finding changes in your process (for example: every seller number is verified against a bank statement or tax return before it enters your model).
- Your own base-rate log, updated after every deal that passes the screen.
- The names of the failed or struggling buyers you have spoken with, and the one check each told you never to skip.
This lesson is education, not investment advice. The interview-corpus findings are one practitioner's tally from his own conversations, not industry statistics, and they do not predict the outcome of any particular acquisition.