Inside due diligencewhere deals come apart.

My article on the sale process and my article on deal structure covered the stages of a sale and what is actually being negotiated. Earlier editions of The Advice Firm Blueprint set out the five readiness gaps buyers test in an advice firm and how to fix each one. This article is a different thing. It is the process itself: what a buyer actually sends you, who is asking, in what order, and what happens to your deal when something in the answers does not hold up.

Due diligence is the point where the deal stops being a conversation and becomes an audit. Founders who have prepared find it demanding. Founders who have not find it destabilising.

Four workstreams, not one

Due diligence is usually run as four parallel workstreams, often by four different teams who do not talk to each other as much as you would expect. Understanding which one is asking a question tells you what the question is really about.

• Legal and corporate. Share register and ownership history, articles, shareholder agreements, options, any past restructuring, leases, supplier and client contracts, employment contracts, litigation and disputes.
• Financial. Statutory accounts and management accounts, revenue by client and by type, recurring versus initial split, debtor and creditor positions, deferred income, adjusted profit workings, and reconciliation between what your systems say and what your accounts say.
• Regulatory and compliance. Permissions, the compliance monitoring programme and its findings, complaints and their outcomes, professional indemnity history and exclusions, file review evidence, suitability documentation, any past business review activity, and how the firm has evidenced its approach to Consumer Duty.
• Commercial and operational. Client demographics and concentration, adviser productivity, retention data, platform and provider dependencies, technology stack, staffing and key person exposure, and growth attribution.

The question behind the question

Most diligence questions have a second question underneath them. Answering only the surface question is what makes a founder look unprepared.

A request for revenue by client is not really about revenue by client. It is about concentration, and whether the top relationships sit with the founder. A request for the complaints log is not about the complaints. It is about whether the firm handled them consistently and learned anything. A request for the compliance monitoring findings is not about the findings. It is about whether the firm identified its own issues before anyone else had to.

The pattern buyers are testing across all four workstreams is the same one: does this firm know itself? A business that surfaces its own problems, with evidence of what it did about them, is priced differently from one where the buyer discovers those problems unaided.

The data room

Everything above lands in a data room, and the state of it tells the buyer something before they read a single document. A room that is complete, indexed, consistently named and internally consistent signals a business that is run rather than merely owned. A room built in a panic over three weeks signals the opposite.

Two practical points. First, version control matters more than founders expect; the same figure appearing differently in two documents will generate a question, and enough of those questions changes the tone of the process. Second, the room is usually populated under time pressure while you are still running the firm, which is precisely why the preparation work should have happened before the process started, not during it.

What happens when diligence finds something

It usually finds something. That is not the problem. What matters is the size of the issue, whether it was disclosed, and whether it can be quantified. Buyers have a fairly standard set of responses.

• Price adjustment. The issue is quantifiable and permanent, so it comes off the number.
• Specific indemnity. The issue is identifiable but uncertain in size, so the seller carries it contractually after completion.
• Increased retention or escrow. More of your consideration is held back for longer against the risk.
• Condition to completion. The issue must be resolved before the deal closes, which usually means delay.
• Withdrawal. Reserved for issues that are large, unquantifiable, or that change the buyer's view of the seller.

The last of those is rarer than founder anxiety suggests, but it is not rare after non-disclosure. An issue you flagged yourself gets negotiated. The same issue found by a buyer's adviser gets treated as a question about everything else you told them.

Where deals in this market actually come apart

From what I have seen, the failure points cluster in a small number of places.

• Advice files that cannot evidence the reasoning behind past recommendations, particularly in higher-risk areas. The advice may have been sound. If the file cannot demonstrate it, the buyer prices the uncertainty, not the advice.
• Client relationships that are contractually the firm's but practically the founder's, with nothing in the diligence pack that shows otherwise.
• Management information that does not reconcile to the statutory accounts, which undermines confidence in every other number in the room.
• Regulatory matters that were live, or arguably live, and were not disclosed early.
• Founder capacity. Diligence runs for weeks alongside a business that still needs running, and founders who have no depth beneath them either stall the process or take their eye off trading. Both show up.

Your job during the process

A founder's role in diligence is not to answer everything personally. It is to have appointed someone internally to own the process, to have your advisers positioned as the channel for anything contentious, and to protect enough of your own time to keep trading in line with the forecast the buyer is underwriting.

The firms that come through this well are almost never the ones with no issues. They are the ones who knew what their issues were before the buyer asked.