Diligence for the smallest buyers

Baby boomers own 41% of U.S. businesses, and roughly 6 million small businesses will change hands by 2035 as those owners retire. The buyers are already lined up. The deals are not closing: the close rate on boring-business deals dropped from 86% to 48% in the last three years, per the trend line in the September 25 IdeaBrowser edition, which named diligence automation as one of the startup ideas the wave calls for.

Two IdeaBrowser entries three weeks apart, and one podcast episode about a very large sale, show why. Every check that protects a buyer, quality of earnings, valuation, red-flag detection, exists. Each one was built and priced for transactions far bigger than a Main Street deal. October's sources show the same checks being rebuilt at small-deal prices, aimed at the buyers with the least protection in the market.

The check costs 7% of the deal

A small business is priced off the seller's profit number, so a padded number means the buyer overpays on day one and services debt on earnings that do not exist. The standard protection is a quality-of-earnings report: an independent check of which profits are real. The October 24 IdeaBrowser entry prices that check at $10K to $25K, against a typical small-business sale price of $350K. The same report on a $3M deal costs under 1% of the price. On a $350K deal it can cost 7%, before legal fees, valuation, and the buyer's own time.

The entry's worked example shows what the check catches. A seller claims $150,000 of profit. Remove a $20,000 marketing refresh with no receipt and subtract a $40,000 cost to replace the owner's own job, and the number a lender may count is $90,000. The receipts collected in the entry repeat the pattern at larger sizes: a business priced on $850K of earnings where the check supports $800K, a seller's pack claiming $800K where the check supports $640K, and cases where $500K of EBITDA vanishes once a replacement manager's salary is subtracted. Over 21% of broken letters of intent are caused by financial discrepancies, discovered after both sides have spent money on lawyers and loan applications. A self-funded searcher looks at more than 50 deals to close one, and 95% of searchers never close at all.

The new SBA rule stops at $3M

On October 1, 2026, the SBA replaced its lending rulebook, and the change draws the protection line exactly where the problem is not. For 7(a) acquisition loans at a purchase price of $3M or more, the lender must now order an independent quality-of-earnings report and use its earnings figure when sizing the loan. The debt-service floor rose to 1.25x for first-time acquisitions, and projections no longer count toward it. A lower checked number now directly shrinks the loan a buyer can get.

Below $3M, a lender may ask for the check but is not required to. That is where most of the volume sits: the SBA 7(a) acquisition segment closed 7,003 deals worth $8.29B in calendar 2025 through September, an average ticket of $1.18M, up 34.58% year over year. The mandate covers the top of that segment. The average deal, and everything under it, still runs on the seller's number plus whatever the buyer can afford to verify.

The seller side of the same deals is nearly as exposed. The October 3 IdeaBrowser entry, a valuation product for Main Street owners, reports that 35% of owners have no idea what their company would sell for, while nearly half of U.S. small-business owners are over 55 and only 54% have a succession plan. In roofing, its wedge niche, the spread between a 3x and a 12x multiple on the same earnings is the difference between a modest retirement and a funded one, and 56 private-equity roll-up platforms were hunting those companies by the end of 2024. The existing valuation tools named in the entry sell to advisors, not to owners or buyers. Both sides of the smallest deals are negotiating against counterparties with better numbers than theirs.

What diligence looks like at the top of the market

The October 26 episode of My First Million covers the sale of Tock, a reservations platform, to Squarespace for hundreds of millions, and its diligence section is a useful contrast. The buyer's diligence list ran to 896 questions. Bankers and lawyers, four lawyers on one call at roughly $3,200 an hour combined, projected four to six weeks to work through it. The selling team converted the list to shared documents, assigned owners, and finished in three days. The seller also declined to hire an investment bank and offered to do the banker's job for 2% of the deal, against the 3% to 4% quoted, payable only if the deal closed.

At that size, diligence is staffed, adversarial, and fast because both sides can pay for it out of the deal. At $350K, the buyer gets a seller-prepared pack and a choice: spend 7% of the purchase price checking it, or sign and find out later. The 48% close rate is what that choice produces at scale.

The rebuild is already priced

The October 24 entry maps a market rebuilding the check in layers, and the prices show how far the cost has already fallen:

• Finsider's Addback screening layer is free. It launched on October 1, 2026, the day the new SBA rules took effect, running ten standardized checks with source-row lineage.
• Centurica lists a $199 Quick Deal Review; ACE CPAs sells a $499 pre-offer review, credited against a later full report, returning a proceed, renegotiate, or walk verdict in 3 to 5 days.
• Rapid Diligence sells a $750 preliminary vetting report and a $1,097 monthly advisory with unlimited reports.
• A full report still sits far above: the median "QoE Lite" proposal on DueDilio for deals under $5M is $12,000, and full engagements run $7,500 to $35,000.

The entry is blunt about where the remaining gap is. The free and cheap layers produce screens: a list of risky add-backs, candidate adjustments, flags. None of them produces the lender-grade number, the earnings figure a lender will actually size debt on under the new rules, with the add-back scoring, replacement-manager cost, and debt-service math tied to the rulebook's language. Getting that number today means paying $8,900 to $10,000 after signing the letter of intent, which is when a bad answer is most expensive. The proposed product in the entry prices the verdict itself at a $99 deep report per deal and a $99 to $149 monthly searcher plan, with the serious money on a $500 to $2,000 monthly tier sold to brokers and lenders, the parties who lose fees every time a deal dies in diligence.

Where the sources push back

The same entry lists the reasons this may not work. The pre-offer screen slot is already crowded, and one competitor is free, so a new entrant fights on price from the first day. Below $3M there is no mandate forcing anyone to buy a check, and small-deal buyers are, in the entry's words, famously cheap. The input data is unreliable by construction: seller packs contain scrambled tables, and sellers lie. A wrong lender-realistic verdict on one visible deal can end the product's credibility in a community, Searchfunder alone counts roughly 10,000 members, that talks to itself constantly.

The valuation entry adds a timing problem from the other side. A seller-side tool has to hold an owner's attention for the two to five years between a first valuation and an actual sale, and its payoff arrives at a transaction that may never happen. Cheap diligence sold to buyers monetizes immediately, per deal. Cheap valuation sold to sellers monetizes late, if at all. The sources support the buyer-side wedge more strongly than the seller-side one, even though the underlying data gap is the same.

What this means for a solo builder

• Sell the verdict, not the screen. Screens are already free; the unsold product is a defensible earnings number a lender accepts, delivered before the offer.
• Price against the deal, not the seat. A $99 report on a $350K purchase is a rounding error next to a $10K report, and it only has to be right often enough to beat signing blind.
• The buyer may be the wrong customer. Lenders must keep preferred-provider lists under the new SBA rules, and brokers lose fees when deals die late. Both have budgets the individual searcher does not.
• Distribution is a community, not a search term. Exact-product searches return low double-digit monthly volume; the buyers gather in Searchfunder and acquisition forums, and credibility there is the asset.
• Accuracy is the whole brand. One public wrong call in a small community outweighs a hundred quiet correct ones, so the product's confidence limits need to be printed on the report, not buried.

What would change this conclusion

Two developments would redraw this market. If the SBA extends the mandatory quality-of-earnings rule below $3M, demand for cheap verdicts becomes compulsory overnight, and the winners are whoever lenders already trust. If lenders refuse to accept AI-assisted checks at any price, the cheap layers stay a pre-screen and the $12,000 median report keeps the market. Until one of those happens, the smallest buyers remain the least protected party in a 6-million-deal transfer wave, and that gap is the opening.

All articles · Research home