Three Bits Survive the Meeting

Last week I told investors the same story five times in four days. It landed twice. It sank without trace once. The other two times it was politely misunderstood.

On the train home I reread Shannon’s A Mathematical Theory of Communication. It’s a 1948 paper about telegraphs and telephones, and after that week it read like a post-mortem on how a true idea can leave one person’s mouth and turn up in someone else’s notes as something else.

I marked every part that explained where my pitch had leaked. It wasn’t a short list.

Information is selection

Shannon’s opening move is the one everybody remembers and hardly anybody takes seriously: he throws meaning out entirely. The engineering problem, he says, is that one message gets selected from a set the receiver was already considering. Information is how much that selection narrows the possibilities.

My deck did not enjoy that framing.

If a wealth-platform CIO hears “AI-powered financial intelligence” and silently files it under another AI vendor, you’ve transmitted zero bits. The set they were entertaining didn’t narrow. You never showed up inside any of the buckets they actually care about (analyst-capacity problem, personalisation at scale, compliant advice automation), so nothing got selected and nothing got communicated.

The 30-second version of our pitch (“personalised investment advice at scale, without hiring more analysts”) lands inside a problem the buyer already recognises, then narrows it. Give me ten minutes and I’ll keep adding words until the signal drowns.

Entropy is surprise per symbol

Shannon defines the entropy of a source as the average surprise of the next symbol. High entropy means the next thing out of your mouth was hard to predict. Low entropy means it could have come from almost anyone.

“We’re AI-powered, building the future of finance” has near-zero entropy for a fintech investor. Every pitch on their calendar says some version of it. There’s nothing to update on.

“10 signed LOIs representing 500K+ end-users, Liminal-backed, SIX as brokerage partner, SRO coverage across FX, commodities, and digital assets” is high entropy. Very few companies on earth could emit that exact sequence. Every clause moves the posterior.

I now read my own decks with a highlighter, crossing out anything a generic competitor could say word for word. What’s left is usually half the length and twice as useful.

Spend the long codes on what’s new

Morse gave “E” a single dot and “Q” a long sequence because E is common in English. Good coding matches each symbol’s length to how often it turns up.

Pitches work the same way. Concepts the receiver already processes fluently (fiduciary duty, advice at scale, MiFID) get short codes. Lean on them and don’t re-explain. The listener’s brain autocompletes and it costs you nothing.

Concepts the buyer hasn’t indexed yet (causal knowledge graph, temporal asset overlay, Mastra runtime) get the long codes. Spell them out, give the example, earn the bandwidth.

Founders usually get this exactly backwards. They over-explain what the buyer already knows (the why now for AI in finance, the regulatory tailwind, the market size) and rush the part that makes the company different. Swap the lengths.

Redundancy fights noise, but it has a budget

Shannon puts English at roughly 50% redundant. Which is why you can still read a sentence riddlled with tyops. Redundancy buys you error correction.

Sales and fundraising channels are spectacularly noisy. The buyer is distracted, half-remembers the last call, has three competitors loitering in the back of their head, and will probably retell your pitch to a partner, in their own words, tomorrow morning.

So you want redundancy across touchpoints: cold email → deck → meeting → follow-up, each carrying the same core message in a different encoding. A good founder story survives any one of those channels dropping out.

What you don’t want is redundancy within a single message, saying the same thing three ways inside one paragraph. That burns capacity the receiver could have spent on a second point. Plenty of decks are stuffed with it. The traction slide shows the same number as a bar chart, a logo wall and a sentence, when it could have used that space to land the next bit.

The craft is knowing which kind of redundancy you’re in.

Only three bits made the train home

This was the note that hurt.

A 30-minute investor meeting has a capacity of maybe two or three bits that actually survive to the partner meeting: [what category are they in], [one differentiator], [one proof point]. That’s it. Everything else either reinforces those three or leaks out the sides.

Shannon’s coding theorem says that below capacity you can get arbitrarily high fidelity. In principle, you can land your message perfectly. Above capacity, no amount of cleverness rescues you, and talking faster doesn’t get you around the limit.

So I had to pick which three bits to ship. For Motif right now, they’re [B2B2C wealth intelligence], [causal graph, not RAG/sentiment], [traction plus a regulated path].

Every meeting tempts me to add a fourth. The fourth doesn’t survive. Worse, it elbows one of the first three out of the investor’s notes.

Equivocation tells you what to fix next

After a pitch, the receiver is still left with some uncertainty about what you actually are. Shannon calls this equivocation: H_y(x), the conditional entropy of the source given what was received.

That leftover is diagnostic. It tells you exactly where the channel failed.

So the second meeting’s agenda should be whatever the equivocation was after the first, not a re-transmission of what already got through. Obvious, and almost nobody does it. Founders tend to re-run the deck with slightly different adjectives and hope volume will do what clarity didn’t.

It’s also why a good investor-FAQ doc earns its keep so fast. You write it once, and it drives down equivocation on the handful of questions you know come up on every call. Marginal cost: a weekend. Marginal return: every conversation afterwards starts one step closer to yes.

The semantic content is “irrelevant”

Shannon’s most provocative line sits awkwardly on the first page: the semantic aspects of communication are irrelevant to the engineering problem.

That’s an ugly sentence when you’re the one pitching. Meaning is the whole reason you built the thing. You didn’t spend two years on the product so it could be squeezed into three bits over a rosé.

The right reading isn’t nihilistic, though. The meaning can be fully formed and true inside your head and still not arrive. Sales and fundraising are the transmission layer for the good thing, and the good thing doesn’t transmit itself.

On the train I opened the deck and cut the fourth bit. Again. Shannon didn’t tell me what to say. He just gave me a less flattering estimate of how much of it would survive the wire.

Want to chat? X / LinkedIn