The price moves before the signature, and not after.
If you are inside two weeks of sending a letter of intent, this is the window.
Start a conversation
Before the letter of intent
You have a number in your head.
What is it resting on?
Before you put an LOI out, the Thesis Test challenges the assumptions behind the price you are prepared to pay, about the company and about you as the one buying it, and tells you which are verified, which still need verifying, and which could materially change your number.
$2,000 fixed. One target, one price, seventy-two hours. A written analysis, then a debrief. No seller contact, no data room, no management access.
The most leverage, the least evidence
Everything you have real leverage over gets decided before the letter of intent, at the point in the deal where you know the least.
What sits in front of you post-NDA is thin by design: a confidential information memorandum, two or three years of returns, whatever internal financials the seller was willing to send, and maybe one or two answers to questions you asked. On an attractive listing you are also on a clock, against buyers who have decided not to think about it as long as you have.
So the offer gets built out of the material and the rest gets built out of inference. That is not a mistake; there is no version of this stage where it does not happen. The price has to rest on something, and at this point most of what it rests on is reasoning rather than evidence.
Some of that reasoning is sound. Some of it entered the picture on the first call, went unchallenged for three weeks, and by now feels like a fact.
The Thesis Test tells you which is which, while the number can still move.
A Quality of Earnings takes the seller’s financials as its object and asks whether the earnings are real. That work now starts earlier than it used to. Software can take a P&L apart before the letter of intent, in minutes, and hand you a sharp list of questions about the company. Use it.
It still answers a question about the company. The Thesis Test takes a different object, and one you already own.
Your thesis. Every assumption the offer price depends on, including the ones you have stopped noticing you are making.
Whether each one was verified, can still be verified during diligence, or can never be verified at all.
A thesis rests on two kinds of assumption, and only one of them is in the documents.
Which revenue actually recurs, what the margin trend means, which costs will repeat. The pack can calibrate these, and an earnings screen can sharpen them.
Whether you will run it or hire for it, and at what cost. What your financing depends on staying true. What your first year assumes about the people and customers you inherit. None of it is in the seller’s documents, because none of it is about the seller.
The second kind is where a thesis most often breaks without anyone noticing, and no document review, by a person or by software, can reach it. It lives in your head until someone asks you to say it out loud. Nothing here waits on seller cooperation. The material you already have is used to calibrate the thesis, not to audit the company.
A written analysis inside seventy-two hours, and a forty-five minute debrief once you have read it.
The document opens with an inventory of every assumption the price depends on, each one sorted into three places:
Then the load points: the two or three assumptions that break the price if they turn out to be wrong. Usually there is one. Sometimes two. They are as likely to sit in your plan as in the company. Each is written with where the belief entered the deal and who actually checked it. The value of the whole exercise is knowing which ones those are before you write the number down, not after.
Last, the questions to carry into the LOI period, ordered by what they are worth.
The document states its own scope on the second page: what was examined, what was not, and what would change the read. It is written to be forwarded. Your lender, your attorney and your accountant can each read it without needing it explained, and none of it evaluates their work.
Sometimes it does. The load points have real checks behind them, the sourcing traces back to something other than a conversation in week one, and nothing in the file is carrying more weight than it can hold.
That result is written the same way as any other, because the document is a record rather than a verdict: every assumption, where it came from, who verified it, what remains open, and in what order to close it. What you hold afterward is a dated account of what the price rests on, built to be handed to anyone else on the deal.
A clean read is not advice to proceed. It says that what you believe is checkable, and it names the parts you are taking on faith, so that you take them deliberately rather than by default.
Through the secure portal: the confidential information memorandum, the returns, whatever internal financials you have, and the price you are contemplating. If you have run a first-pass earnings screen, send its output too. The answers you typed into it, such as the replacement salary, the owner’s hours, and whether a cost repeats, are assumptions, and they belong to your thesis. Nothing you do not already hold.
Recorded, with your consent, because the useful part is usually the follow-up question rather than the first answer. This is where the thesis actually gets said out loud. Software can read the pack. The thesis is not in the pack.
The written analysis, through the portal. Pre-LOI moves quickly; a review that arrives after the letter goes out is a receipt, not a decision.
After you have read it, not before. The document does the work; the call is where you push on the load points and decide what to do about them.
The full fee credits against a Quality of Deal engagement on that same target, within ninety days. You do not pay twice for the same thinking.
The engagement
The Thesis Test is $2,000, flat. Not a ratio, not a rate, and not scaled to the size of the deal, because at this stage you may be looking at three targets and expect to lose two of them.
It covers one target and one contemplated price. A second target is a second engagement at the same fee. If the file genuinely needs the full method, that is a Quality of Deal Sprint and the fee is quoted as one.
The credit is per target, not per client, and it is the whole fee. Run a Thesis Test on a company, then take a Quality of Deal engagement on that same company within ninety days, and the two are read as one engagement paid in two parts: the $2,000 comes off, and what you pay across both is the engagement fee. It does not travel to a different deal, because the thinking does not travel either.
The Known vs. Assumed Ledger is a twelve-question instrument you run yourself, against a deal you are actually in. It separates what you have verified from what you have assumed and dressed as known, and takes about twenty minutes.
It will not resolve anything. It will tell you how many assumptions you are carrying and whether the gap is one you can live with. Plenty of people run it and decide they do not need anyone.
Free with a subscription to the Decision Signal Journal.
It arrives with the welcome note, immediately.
More on the category at Quality of Deal.
If you are inside two weeks of sending a letter of intent, this is the window.
Start a conversation