Hire the Wrong Expert
What Bletchley Park's codebreakers figured out about cognitive diversity — and why most startups get it backwards
1941 - The British government took out an advertisement in The Daily Telegraph.
Not for engineers. Not for mathematicians. Not for trained intelligence analysts.
They ran a crossword competition.
The ad was a recruitment pipeline in disguise. Anyone who could finish the Telegraph‘s famously difficult crossword in under 12 minutes received a letter in the mail. Come to London. We have something we’d like you to help with.
Those recruits ended up at Bletchley Park, a country house north of London that housed Britain’s codebreaking operation during World War II. Their mission: crack the Enigma cipher. The German military used Enigma to encode every communication: troop movements, supply chains, U-boat coordinates. The machine had more possible settings than atoms in the observable universe. Germany’s engineers had designed it to be mathematically unbreakable.
Bletchley’s answer was not to hire the world’s best cryptographers.
It was to build the world’s strangest team.
The roster reads like someone accidentally merged three academic departments and forgot to assign offices.
Alan Turing, a Cambridge mathematician obsessed with the theory of computing. Gordon Welchman, an algebraic geometer. Dilly Knox, a papyrus scholar who’d spent two decades deciphering ancient Greek texts. A professional chess champion. A museum curator. A bridge player. Linguists, classicists, crossword constructors.
Not a single person had worked in intelligence. That was, mostly, the point.
Researchers who’ve studied Bletchley call it “organizational porosity” — knowledge moving sideways instead of up. No formal org chart. A linguist could walk into a mathematician’s hut and describe a pattern he’d noticed in intercepted messages. The mathematician would see something the linguist couldn’t. The chess player would see something neither of them caught.
The crossword solvers found the first critical cribs: known-plaintext fragments that gave the mathematicians something to anchor on. The reason: they were looking at the cipher as a puzzle, not a mathematical object. They noticed that German operators, bored on overnight shifts, signed off with predictable phrases. The same words appearing at the same position in messages sent at the same time each day.
The mathematicians would never have found that. They were looking for structural weaknesses in the algorithm.
The crossword solvers were looking for human laziness.
They needed each other to see the whole thing.
Most companies do the opposite when problems get hard.
When something won’t budge, the instinct is to find whoever has solved that exact problem before. Sales is struggling, so you hire a VP Sales from a company that figured it out. Product isn’t landing, so you bring in a CPM with a stronger pedigree. Feature velocity is low, so you add senior engineers.
The logic is intuitive. Hard problem, deep expertise. Makes sense.
But domain expertise comes packaged with domain assumptions. The seasoned VP Sales knows how to run a playbook — the one that worked at her last company, in her last market, against competitors she understood. The experienced CPM knows what “good” looks like, which means he also knows what “impossible” looks like. Expertise is pattern recognition. The patterns aren’t always right.
The people at Bletchley who made the first breakthroughs weren’t the ones who knew the most about codes. They were the ones who’d never been told what made codes impossible to break.
Psychologists call this functional fixedness: the inability to see a tool used for anything other than its intended purpose. A hammer is for nails. Experts develop professional versions of this. A cryptographer looks at a cipher and sees a mathematical object. That framing carries assumptions about where the weaknesses must be.
The crossword solver doesn’t have that. She sees the cipher as something humans made, which means humans are using it, which means humans are taking shortcuts. That observation is worth more than a lot of advanced mathematics in the right moment.
This isn’t a case for hiring generalists instead of specialists. Bletchley needed Turing. The Bombe machine he built, which tested possible Enigma settings at mechanical speed, was the infrastructure the entire operation ran on. Without the mathematical rigor, the human observations went nowhere.
But without the crossword solvers’ cribs, Turing’s machine had nothing to test.
The specialists and the generalists were each other’s raw material.
Pixar figured this out around 2000 with the Braintrust.
The Braintrust is a group of directors and writers who watch rough cuts of films in development and give feedback. No hierarchy. No authority. The director can take every note or throw them all out — the Braintrust doesn’t decide anything.
Ed Catmull, in Creativity, Inc., describes the goal as giving filmmakers a moment to see what they’ve made through someone else’s eyes. People invested enough to give real feedback, but not invested in the film itself.
That’s organizational porosity. Walls between projects stay thin. What the animation team learned on Finding Nemo flows into what the story team is building on The Incredibles. Not through a knowledge management system. Through people who move between rooms.
Most early-stage startups do this by accident. The designer, the engineer, and the founder sit in the same customer call. The support rep walks to the product manager’s desk and says, “You should hear what this customer told me.” The walls are thin not by design but by necessity — you don’t have the headcount to build silos yet.
Then the company grows. The walls go up. Disciplines separate.
And the hardest problems stop getting solved.
I’ve watched this pattern run in both directions. At Property Meld, when the product team was stuck on a workflow problem that engineers and designers had been circling for weeks, a customer service rep who’d never attended a product meeting walked in and said: “They just want to know what’s happening to their request.” The whole frame changed in sixty seconds. She wasn’t solving the technical problem. She was pointing at it from the right angle.
At AgCents, a cattle lending startup I work with, the breakthrough in underwriting logic didn’t come from the finance team. It came from a rancher describing forty years of buying decisions that no spreadsheet had ever captured. The finance team had the model. He had the variable they’d been missing.
Different rooms, different angles. Same problem, now visible.
The Bletchley lesson, stated plainly:
For your hardest problem — the one that’s been stuck for months — the answer probably isn’t more domain expertise. It’s a different lens on the same problem.
Find the crossword solver. Not metaphorically. Find someone with no context about what you’ve already tried and no stake in why it hasn’t worked. Someone whose pattern recognition comes from somewhere else entirely. Give them full access to the problem. Don’t explain your framework before they start. Don’t tell them what you’ve already ruled out. Let them describe what they see.
Then sit with your mathematician and figure out what to do with it.
The chaos at Bletchley wasn’t dysfunction. It was the operating system. Different disciplines don’t just bring different answers — they bring different questions. And in 1941, it turned out the right question wasn’t mathematical at all.
It was: why do German operators keep sending the same phrase at the start of every morning weather report?
Nobody trained in cryptography thought to ask that.
The museum curator did.
Todd - Great post. When I'm trying to solve a business problem problem or figure out how to build a system, I look in parallel spaces for insights. Since I can't hire Alan Turing, I research Philosophy, political science, military strategy, etc. You frequently find solutions that can be adapted to your domain.
Founders want the toolbox for progress, not perfection.Founders’most expensive lesson is:hiring mistakes.So,how to avoid such mistakes—MR.TODD GAGNE post—Hire the Wrong Expert—draws from history,real life examples and explains—What Bletchley Park's codebreakers figured out about cognitive diversity — and why most startups get it backwards.
MR.TODD article is an EYE-OPENER—We’ve been taught that the surest way to win is to gather the most talented people. But stacking a team with stars doesn’t guarantee success. In fact, it often undermines it.Take the 1980 U.S. Olympic basketball team. They were just college kids, facing NBA All-Stars in a series of exhibitions. On paper, the pros should have crushed them. Instead, the college players won four out of five games, including one by 31 points. The less talented team consistently defeated the stars.
I always Learn something worthy to adapt from each post of MR.TODD and that has helped me and my associates in avoiding/correcting mistakes before it’s too late.