Letter of Intent — The Product Playbook

A letter of intent is a signed, non-binding document where a business customer states, in writing, that they intend to buy if you build what you described. You take a one-page spec to a real buyer, and you ask them to put their name on it.

It's the B2B answer to one question. Will they actually spend? In enterprise, nobody swipes a card on a landing page. The deal is too big, the buying takes too long, and the person who wants it isn't the person who signs the check. So the currency you collect here isn't dollars yet. It's a signature from someone willing to put their name on the intent to spend them later.

That puts this squarely in the desirability question, whether anyone actually wants the thing and will pay for it, at the willing-to-pay end of it. Not "do they like the idea." A survey answers that, and it's worth almost nothing. This answers the harder one. Will a real buyer commit to it before it exists?

How to run it cheaply

You don't build anything. That's the whole point.

You write one page. The problem you're solving, the solution you're proposing, and the rough price. No product behind it, no demo, no slide deck that took a week. One page a target customer can read in two minutes and understand exactly what they'd be buying.

Then you take it to the buyers you already believe want it, and you ask for the signature. The ask is plain: "if we build this as described, we intend to purchase." Non-binding, on purpose. You're not asking them to commit money or risk a lawsuit. You're asking them to put their name on the intent, which is a far higher bar than a nod in a meeting and a far lower one than a contract.

Write the success number down before you start, the same rule every experiment on the Lean Experiments catalog leans on. How many of the buyers you approach have to sign before you'll commit to the build? Pick that number first, so it can't bend later when you're staring at a stack of polite maybes.

Worked example (illustrative)

Picture this. You're building compliance software for mid-size manufacturers, and the build is months of real engineering. Before you write a line of it, you want to know the deals are real.

So you write one page. The problem (the audits they dread, the spreadsheet they maintain by hand). The solution (what your tool would do). The rough price. Then you take that page to eight target customers and ask each of them to sign a letter of intent: if you build this as described, we intend to purchase.

You set the bar first. At least four of the eight sign, or you don't build it. (Both the case and the number are illustrative.)

Here's the honest part. A signature isn't a contract, and some signers will still walk when the invoice shows up. That's real. But run it the other way. A buyer who won't sign a non-binding letter, who won't even put their name on the intent when it costs them nothing, was never going to buy the finished thing. You just found that out before you spent the quarter, instead of after.

That's the trade. A signed page is a weaker yes than a wire transfer and a much stronger yes than enthusiasm in a room. In enterprise, it's often the strongest yes you can get before the thing exists.

When to reach for it, and when not

Reach for a letter of intent when you're selling to businesses, the deals are large, and the buying cycle is long enough that a pre-sell page can't carry the weight.

That last part is the line between this and its closest neighbor. The pre-sell takes real money up front, and money is the truest signal there is, hardest to fake. So when you can take the money, take it. But in enterprise you usually can't. The buyer needs procurement, legal, a budget cycle, signatures that aren't theirs. Asking them to swipe a card today isn't a cheaper test, it's an impossible one. The letter of intent is what you reach for when the real check is six months and four approvals away and you still need a read now.

Don't reach for it when a card swipe is actually on the table. If you're selling to individuals or small businesses who can buy today, a pre-sell beats a signature every time, because a paid invoice is a harder yes than a signed intent. Don't manufacture an LOI where money would move.

Don't reach for it before the cheap signals have come back positive, either. Think of the experiments as a ladder. Cheap, weak-signal tests sit at the bottom, and expensive, strong-signal commitments like this one sit near the top. A letter of intent sits high on that ladder, where the cost of the yes is real. If you don't yet know whether anyone wants the thing at all, run a demand test first, a landing-page test or interviews, and climb to the signature only once the cheap evidence earns it.

And don't confuse it with the A/B test. That one needs a shipped product and a customer already using it, and it tunes how often you get the yes. The letter of intent runs before anything exists, and asks whether you'll get the yes at all. One optimizes a yes you have. The other goes and gets one.

Credit where it's due: the experiment taxonomy here, the letter of intent and the 44-experiment library it belongs to, comes from Testing Business Ideas by David Bland and Alex Osterwalder.