A letter of intent is the most over-read artefact in a pre-seed deck. It is a genuine signal, and it is a bounded one, and almost all the trouble comes from collapsing those two facts into a single claim. The founder who has three signed letters wants them to mean traction; the investor who is offered them wants to know whether they do. Both are right to take the letters seriously and both are one careless step from reading more into them than a letter can hold. This is a field note on where that line sits, and it is the general question rather than the first-person one: not the story of our own three letters and how their overlap decided which features were table stakes, which is told elsewhere, but the question an investor faces with any deck that leads with letters of intent. Given a short stack of non-binding notes of interest, what can you actually conclude?
To keep it concrete we calibrate against a real pre-seed spread from a raster-log-digitisation venture, three letters of intent from four buyer types, an independent explorer, a private-equity asset holder, a national oil company, and an independent operator, sitting under a 6.7 billion dollar serviceable market and a 1,200 dollar per-seat list price. The numbers are real; the reading is the general one.
A note of interest is not a signed order
Start with what a letter of intent is, mechanically, because most of the over-reading comes from forgetting it. A letter of intent is a signed statement that a named party is interested in buying, under terms to be agreed, at some future point. It is signed, which is why it feels weightier than a warm email, and it is non-binding, which is why the weight is smaller than the signature suggests. Nothing in it obliges the buyer to pay, to accept a price, or to sign the contract it gestures at. That is not a defect of the instrument; it is the instrument. A letter of intent exists precisely so a buyer can register real interest without committing to terms that neither side has settled.
This is why the first discipline is to refuse the word traction for it. Ries drew the line that does the work: a vanity metric flatters and an actionable metric informs a decision [2]. A letter of intent read as booked revenue is a vanity metric wearing a signature. Read as evidence that a named buyer, whom you can call and who will take the call, feels the problem badly enough to put their name on paper, it is one of the more actionable signals a pre-seed company can produce, because it is a claim about a real person's real problem rather than a number the founder generated. The letter carries information about the job the buyer needs done; Ulwick's outcome-driven framing is that what a buyer will commit to is the outcome they are trying to reach, and a letter of intent is a commitment to caring about that outcome, not to a price for it [3].
The count is the weak part; the spread is the signal
The instinct is to read a stack of letters by its height. Three feels better than one and worse than ten, and a founder who has three is tempted to present them as three units of the same thing. That is the wrong axis. Three letters is a small number by any statistical standard and no amount of presentation changes that; what a pre-seed investor can actually learn from them is not in the count but in the spread of who signed.
Here the four buyer types matter more than the three letters. An independent explorer, a private-equity asset holder, a national oil company, and an independent operator are not one market sampled four times; they are four buying contexts with different budgets, different procurement, and different reasons to want the same tool. Moore's segmentation argument is that early buyers are emphatically not a homogeneous mass, and that treating them as one is how a company mistakes a beachhead for a market [4]. Applied to letters of intent it inverts the naive reading: ten letters all from the same kind of buyer tell you one segment is interested, a narrower fact than it looks, while three letters from distinct buyer types tell you the problem is not local to one niche. The spread is the signal. The count is close to noise until it grows into the hundreds, which at pre-seed it never has.
The exhibit above is the reading made mechanical. The left column is the evidence ledger, five claims a letter of intent might be asked to carry, each a filled bar where a non-binding note supports it and an empty bar where it does not. Problem-interest and buyer identity fill; price acceptance, a signed contract, and booked revenue stay empty, because a letter of intent does not touch any of the three. The right column sets the honest reading, a fixed demand vector of three letters across four buyer types, against the fantasy, an orange bar of revenue that exists only because someone chose to book the letters as paid seats. Drag the lever and watch dollars appear out of nothing, then notice that even fully imagined they are a rounding error against the 6.7 billion dollar serviceable market. That is the argument in one motion: the letters point somewhere real, and the moment you convert them to a dollar figure you are inventing, not measuring.
What the letters are silent about
The three empty bars each name something a pre-seed reader must go get elsewhere and must not credit to the letters. The first is price. A 1,200 dollar per-seat list price is a founder's proposal, and a letter that does not name it has said nothing about whether the buyer will pay it; a signed interest in the problem is fully compatible with a flat refusal of the price once it is on the table. The second is the contract. A binding order has terms a note of interest leaves open, and every open term, volume, start date, support, exit, is a place the deal can die after the letter is signed. The third is revenue, downstream of the first two: there is no invoice, nothing collectable, nothing that belongs on a projection as booked.
Blank's customer-development discipline is the one to hold here: early evidence is hypotheses to test against real buyers, not facts to bank [1]. A letter of intent is a strong hypothesis, it says this buyer will probably engage, and a weak fact, it guarantees nothing. The failure mode is treating the strong hypothesis as the weak fact and skipping the test. The letters tell an investor where to point the diligence, at whether these named buyers will accept a price and sign terms, which is exactly what a demand vector is: a direction that tells you which experiment to run next, not the answer to it.
Reading the vector, not the number
Put the pieces together and a coherent reading falls out, neither the founder's inflation nor the cynic's dismissal. The letters are real evidence that the problem is felt across more than one niche, information a cold market-size slide cannot provide, and they are silent on price, contract, and revenue, so any number derived from them is invented and the honest financial reading of three pre-seed letters is zero booked dollars. The rule this leaves us with is to read the vector and refuse the number. For a founder that means presenting letters as evidence of cross-segment demand and a queue of named buyers to convert, never as revenue, because an investor who catches the inflation discounts the genuine signal along with the invented one. For an investor it means crediting the direction while pricing in that price, contract, and revenue are all still ahead, so the diligence question is not how much is this worth but will these specific buyers accept terms.
Limitations
This is a field note built around one real pre-seed spread, not a survey of many, so its numbers illustrate the reading rather than establish a base rate. Three letters from four buyer types under a 6.7 billion dollar serviceable market and a 1,200 dollar list price are the actual figures from one raster-log-digitisation venture; a different market or letter count would move the specifics even though the structure, signal in the problem and the spread, silence on price and contract and revenue, would hold. The imagined dollars the instrument prints are an illustrative what-if to make the gap visible, not a forecast; the letters priced nothing. We also treat the letters as sincere, which pre-seed letters usually are, and say nothing about a letter procured to decorate a deck, where even the problem-interest signal weakens. And a letter of intent is one signal among several; nothing here argues it should be read in isolation, only that when it is read it should be read for the vector it is and not the traction it is not.
Where the signal stops and the work starts
The habit worth keeping is to let a letter of intent do its actual job and no more. Its job is to point, to tell a founder which buyers are worth converting and an investor which segments are already leaning in, and it does that job well when the letters come from genuinely different kinds of buyer. Its job is not to stand in for the sale, and a deck that leans on letters as though they were bookings is quietly announcing that the sale has not happened yet. Read as a demand vector, a short stack of pre-seed letters of intent is one of the more honest artefacts a young company can show. Read as revenue, it is the least. The letters are the same either way; the difference is entirely in the reading.
References
[1] Blank, S. The Four Steps to the Epiphany: Successful Strategies for Products that Win. K&S Ranch (2005, 2013 ed.). The customer-development case that early evidence is a set of hypotheses to be tested against real buyers rather than facts to be banked. https://web.stanford.edu/group/e145/cgi-bin/winter/drupal/upload/handouts/Four_Steps.pdf
[2] Ries, E. The Lean Startup. Crown Business (2011). The distinction between vanity metrics that flatter and actionable metrics that inform a decision. https://theleanstartup.com/book
[3] Ulwick, A. W. What Customers Want: Using Outcome-Driven Innovation to Create Breakthrough Products and Services. McGraw-Hill (2005). The outcome-driven view that what a buyer will commit to is the job they need done. https://strategyn.com/outcome-driven-innovation-process/
[4] Moore, G. A. Crossing the Chasm: Marketing and Selling Disruptive Products to Mainstream Customers, 3rd ed. HarperBusiness (2014). The segmentation argument that early buyers are not a homogeneous market, which is why the spread of buyer types behind a set of letters carries more signal than their raw count. https://www.harpercollins.com/products/crossing-the-chasm-3rd-edition-geoffrey-a-moore
