When the Buyer Reads Your Books in a Day

Two years ago, a buyer's diligence team spent its first week in a data room simply finding out what was in there. That week is now closer to a day. Generative AI is embedded in most institutional deal workflows, and the change has less to do with buyers finishing sooner than with what they do with the time they get back.

The tooling is no longer experimental

Adoption has moved past the pilot stage. Of surveyed corporate and private equity organizations, 86 percent have integrated generative AI somewhere in their M&A workflow, and 35 percent of those adopters use it specifically for due diligence. More than 60 percent of private equity firms report using at least one generative AI tool for sourcing, screening, or diligence, and roughly 80 percent of those users say it reduced manual effort.

The pressure is coming from the capital side as well. In a survey of 405 limited partners, 99 percent said they want their fund managers integrating AI into dealmaking, and 66 percent specifically expect it in diligence. For a general partner raising a fund in 2026, having no answer to that question is a problem in the room.

What the tools actually do is narrower than the marketing suggests, and more useful. They extract change-of-control provisions across a contract set, flag agreements that cannot be assigned to a new owner, reconcile financial anomalies between the general ledger and the reported statements, and locate representation-and-warranty language across a data room holding several thousand documents. Work that used to take a junior team a week of reading now takes hours. Early adopters describe spending about one day summarizing a data room where the same output previously took five.

Faster review has produced deeper review, not shorter deals

The reasonable assumption is that faster document handling shortens the diligence period. That is not what has happened in the middle market. Diligence still runs four to six weeks inside a term-sheet-to-signing window of roughly six to twelve weeks, and in practice the whole exercise tends to cluster at eight to ten weeks. Timelines have held roughly steady.

What changed is depth. Time recovered from document review has been redeployed into analysis. The most thorough financial diligence reports in 2026 stop treating workstreams as separate exercises and instead reason across them, mapping a financial finding against as many as nine parallel streams covering technology, legal, commercial, and other areas. A customer concentration issue surfaced in the revenue analysis now gets tested against the contract set, the sales pipeline, and the integration plan in the same report rather than in three disconnected memos.

Quality of earnings work sits at the center of this. It is the single largest workstream in financial diligence, typically around 30 percent of total effort. A quality of earnings analysis stress-tests reported profit to separate earnings a new owner can expect to keep from results that depended on one-time events, related-party arrangements, or accounting choices that will not survive a change in ownership. Adjustments that a seller considers obvious housekeeping, the owner's above-market compensation, a below-market lease from an affiliated entity, a large one-time contract, are precisely what the analysis is built to isolate.

Buyers have also become more willing to act on what they find. Dealmakers surveyed at ACG DealMAX in late April put certainty of execution near the top of their priorities, meaning speed, structure, and financing clarity, and 77 percent identified misaligned price expectations between buyers and sellers as the primary barrier to getting deals done. An analysis that lands three weeks into exclusivity and moves the earnings base by 8 percent is a price conversation, not a footnote.

Why sellers started commissioning the analysis first

The response from prepared sellers has been to run the exercise on themselves before going to market. A sell-side quality of earnings report, commissioned by the seller two to three months ahead of launch, has become standard practice on institutionally advised deals. At least 90 percent of private-equity-backed transactions now include one.

The outcome data supports the spend. Sellers who came to market with a sell-side quality of earnings report cleared roughly 7.4 times total enterprise value to EBITDA, against 7.0 times for those who did not. In a middle market where private equity multiples have been sitting flat at 7.2 to 7.5 times, four tenths of a turn is not a rounding difference. On a business with $8 million of EBITDA it is worth about $3.2 million.

The defensive value is easier to quantify. On transactions valued between $5 million and $15 million, a sell-side report is estimated to save $200,000 to $1 million by removing the buyer's leverage to re-trade price after exclusivity begins. A re-trade works because the buyer discovers something the seller did not disclose, at a point when the seller has no competing bidder left and considerable sunk cost. A finding the seller already documented, quantified, and explained is not a discovery. It is an agreed adjustment, and it was priced into the offer.

Cost is modest against the numbers involved. Sell-side reports run roughly $15,000 to $25,000 for businesses under $3 million of EBITDA, $25,000 to $50,000 in the $3 million to $10 million range, and $50,000 to $75,000 or more above that. Sell-side work generally prices at 40 to 60 percent of a comparable buy-side engagement, because it is not adversarial and the provider has cooperative access to management. Preparation takes four to eight weeks, which is the real reason to start early: the report is only useful if it exists before the first buyer call.

Where the advantage is concentrating

The market is not applying this scrutiny evenly. Multiples are holding firm for businesses with recurring revenue, low customer concentration, and a management team that can operate without the owner in the building. Businesses that are owner-dependent or reliant on a single large customer are drawing longer processes and closer examination. Faster, more capable diligence widens that gap, because the characteristics that make a company hard to underwrite are now identified in week one instead of week four.

Two developments are worth tracking through the rest of the year. First, whether the disclosed-adjustment premium holds as sell-side reports become universal among advised sellers; an advantage that everyone has is a baseline, and the penalty then shifts to sellers who arrive without one. Second, whether buyers begin extending the analytical reach of these tools into forward-looking work, testing a seller's forecast against its own historical accuracy rather than accepting the model as presented. Sellers who have never compared their prior three budgets to actual results should expect that a buyer will.