What you ask for, and the order you ask for it in, determines what you learn. The response to the request is frequently more informative than the documents themselves.
13 min read · Level 3
There is a widely held view that a demanding request list sours a relationship you will need afterwards, and that a light touch buys goodwill worth more than the documents. It is not a foolish position - private transactions run on continuing cooperation, and a buyer who behaves like a prosecutor gets a worse business afterwards even when they get a better price. But the trade is usually misdescribed. What damages a relationship is not the length of the list; it is being asked for the same thing four times because nobody tracked the first three, being pressed hardest on the items that matter least, and having a fortnight of work commissioned and then abandoned without explanation. Sellers almost never resent rigour. They resent disorder, because disorder costs them time and tells them the buyer is not serious. The uncomfortable conclusion is that the diligence process is also an audition, and the thing being auditioned is you - which means the most self-interested reason to run an orderly list has nothing to do with what you learn from it.
A diligence request list is the set of documents and access you ask for, written down and sent. In a private transaction it is usually the first thing that makes the process real, and it is the single most revealing document you will produce, because it commits you to what you think matters.
Order is the part people get wrong. A request list is not a shopping list to be handed over in one block; it works in waves, and each wave is chosen so that the cheap things that could end the conversation happen first. Asking for three years of audited statements and a full contract set in week one buys you a month of reading before you discover something that was visible in a two-page ownership register on day two.
The response is data in its own right. A request can be answered, answered slowly, answered partially, answered with a substitute document, or refused. Each of those is information about the counterparty, the state of the company's records, or the thing being asked about - and unlike the documents themselves, it cannot be prepared in advance.
Because the request list is the only part of the process you fully control. You cannot make a business better than it is, and you cannot make a seller more forthcoming than they intend to be, but you decide what is asked and in what sequence, and that decision sets the ceiling on what the process can possibly discover.
And because a well-ordered list is what protects you from the costs of diligence itself. A process that spends 60,000 before reaching the ownership register has made the walk-away decision far harder than it needed to be, for reasons that have nothing to do with the business. Sequencing is not administration; it is how you keep the option to stop cheaply for as long as possible.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
The same manufacturer, revenue of 4,200,000. Three waves.
Wave one - cheap, fast, and capable of ending the conversation. The ownership register and the share certificates. A list of every charge, lien or security registered over the company or its assets. The last twelve months of bank statements, unredacted. A list of every customer with revenue beside it. Any litigation, threatened or live. Two weeks, no advisers, almost no cost.
Wave two - the picture. Thirty-six months of management accounts, monthly. The receivables and payables ledgers. Every contract representing more than 5% of revenue. The employment terms of anyone the business cannot run without. Insurance, licences and permissions, with their renewal dates and their transfer terms. Tax filings and any correspondence with the tax authority.
Wave three - the expensive confirmations. An independent review of the quality of the earnings. Reference calls with customers and former employees. A technical or site inspection. This is the wave that costs real money, and the whole point of the first two is to avoid paying for it on a transaction that was never going to happen.
Now what the monthly accounts did. Annual revenue of 4,200,000 averages 350,000 a month. The last three months averaged 275,000. Annualised, 275,000 times twelve is 3,300,000 - which is 21.4% below the figure the whole conversation had been conducted around. The annual number was not wrong. It was an average across a year in which the business had been shrinking for a quarter, and an annual figure cannot show that. Only the monthly series can, which is why it is on the list and why it is asked for by month rather than by year.
And now the refusal. The seller provides everything in wave one except the unredacted bank statements, offering instead a summary of receipts prepared by their accountant. The summary may well be accurate. But the substitution has converted a primary source into a management representation, and it has done so silently - a buyer who files the summary under 'bank statements received' has lost the distinction without ever being told it was gone. The correct response is neither to accept it nor to walk out. It is to ask, in writing, what specifically prevents the statements from being provided, and to record the answer. 'Our bank charges for historic statements' is a different world from 'we would prefer not to'.
The first mistake is sending everything at once. It buries the items that could end the process quickly inside a hundred that cannot, and it signals that nothing on the list is more important than anything else - which is exactly the impression you do not want to give.
The second is asking for annual figures where the question is about a trend. Annual revenue is an average, and an average is where a deteriorating quarter goes to hide. Monthly, for thirty-six months, is the format; anything less and the series is too short to tell a decline from a season.
The third is accepting a substitute document without noting the substitution. This is the quiet one. The summary arrives, it is responsive, it is filed, and the fact that a primary source became an interested party's own account is lost permanently.
The fourth is treating a slow response as rudeness rather than as data. A company that cannot produce its own receivables ledger in a fortnight is telling you something true and important about how it is run, and it will still be true after you own it.
The fifth is failing to write down what was never provided. At the end of a long process the file is full, everyone is tired, and the three items that were promised and never arrived are the hardest things in the world to remember. Keep the list with its status column, and read the unanswered rows immediately before deciding.
A list written after you have met the owner is shaped by what you have already been told is available and what you sense they would rather not discuss. Draft it from what the business is, then adjust. The first version is the honest one and is worth keeping even after it changes.
Everything that is cheap to obtain, fast to read and capable of ending the conversation goes first: ownership, charges and liens, bank statements, customer revenue, litigation. The expensive confirmations go last, after the cheap facts have had every chance to stop you.
Monthly for thirty-six months, not annually for three years. The annual figure is an average, and the decline you most need to see is exactly what an average removes. This single formatting choice finds more than any other item on a request list.
Requested, received, partial, substituted, refused, with a date. Update it the day something arrives. The status column is the document you actually read at decision time, and if it is maintained only in memory it will show every row as received.
The reason is the information, not the refusal. A bank that charges for archived statements, a contract with a confidentiality clause, a genuine legal restriction - these are ordinary and answerable. A reason that is never actually given, or that changes when pressed, is a different finding entirely, and it is one you can only make if you asked in a form that leaves a record.
Include a request whose answer you have independently established from a public register or a prior document. The response calibrates everything else in the file - not because you expect to catch anybody out, but because you learn how carefully this counterparty answers when the stakes are low, which is what you then need to assume about the answers you cannot check.
In any private transaction, and any time you can ask a company directly for material - a small business purchase, a private placement, a loan, a joint venture, a substantial stake in anything not publicly filed.
A request list is the wrong instrument where you have no leverage to ask - buying a listed share, most fund investments, anything where the material is a published document set. There the equivalent discipline is reading the filings adversarially, and sending a long list of questions to an investor relations desk is a way of feeling thorough while learning nothing.
In transactions that later disappoint, a regularly recurring detail is that the finding which caused the loss sat inside a document that was requested, acknowledged, promised, and never actually delivered. It is rarely refused - refusal would be noticed. It is deferred: it is being pulled together, it is with the accountant, it will come with the next batch. Because deferral is polite and open-ended it produces no moment at which anybody decides to proceed without the document, and by the time the file closes there are two or three such rows that nobody has looked at as a group.
Before deciding, read only the rows that are not marked received, as a list, with nothing else in front of you. That short list is the shape of what you do not know.
Across small-business purchases, the single formatting choice that most often changes the conclusion is asking for monthly management accounts over three years rather than three annual sets. Annual figures present a business as a sequence of three data points, which is too few to show anything but direction, and each point is an average that has already absorbed whatever happened inside the year. Thirty-six monthly points show seasonality, the month a large customer stopped buying, the step change when a price rise went through, and the quarter in which growth stopped - none of which is visible in an audited annual statement that is entirely accurate.
Wherever you are given an annual figure, ask for the monthly series underneath it. It costs the seller almost nothing and it is the request most likely to change what you believe.
Create a free account to read the rest and apply it to your own investing.