The difference between the people who build companies and the people hired to run them, and how to read it.
Dean Foerter · Signal Labs · Second edition, October 2026
Investors name the founder the most important factor in a deal, and the research agrees, yet it is the one variable still assessed by gut while everything else is measured to the decimal. The research on founder-led firms and founder reputation says the person at the top is priced. Read on one instrument, 139 of America’s largest leaders divide sharply. The people hired to run companies converge: about three in four read as one type, the institutional Commander. The people who built companies do not. More than half of founders read as one of four builders, the Architect, the Sorcerer, the Trailblazer and the Revolutionary. Five of 94 hired chief executives do: about ten to one. Only 27 percent of founders read as the Commander, and each of the four builders wins, fails and must be governed in a different way. Our own pilot found no link between founder character and returns, so this is not a way to pick winners. It is a way to read the founder correctly, and know which one you are funding.
Part One
You back founder-led businesses, so start with the variable you already know matters most and measure least. Ask an investor what they are really betting on in a founder-led company and they will tell you: the founder. In the largest survey of venture capitalists ever run, ninety-five percent named the management team an important factor, and nearly half called it the single most important one, ahead of the product and the market. They are right, and they know it.
And yet the founder is the one part of the deal still assessed by gut. The financials get a quality-of-earnings review to the decimal; the person the whole thesis rests on gets a ninety-minute meeting, a few reference calls and a dinner. The same firm that would never accept a financial model on a hunch will bet nine figures on a leader on one. Everything else is measured. The founder is a feeling.
Gompers, Gornall, Kaplan & Strebulaev, “How Do Venture Capitalists Make Decisions?”, Journal of Financial Economics, 2020 (survey of 885 investors).
That gap is the whole opportunity. The founder’s character is the company’s operating system: it decides what gets built, who stays, how the business behaves under pressure and what is left after the cheque clears. And the person who builds a company is not the same animal as the person a board later hires to run it.
We measured it. We read 139 of America’s biggest leaders, 45 founders and 94 hired Fortune 100 chief executives, on one instrument that reads how a person shows up in public, not the resume. The result is stark. The people hired to run companies converge: 71 of 94, about three in four, read as a single type, the Commander. Founders do not. Twelve of 45 read as the Commander, 27 percent, and a third lead from the Commander’s family, so founders are operators too. But two thirds of them are something the hired almost never are. Building a company is an act of character. Running one is a job a board selects you into, and boards select for safety.
So a founder arrives as one of a small set of types, and they divide in a way that matters to you. One is the Commander. The other four build.
The Commander runs the machine and holds order at scale. This is the type a board almost always hires, the safe pair of hands an institution selects for. Among founders it shows up as the owner who controls what they built. It is the one type you can also hire.
The other four are the builders. These are the four you actually bet on.
That is about ten to one. Counting the Architect’s and the Sorcerer’s families whole, with the Trailblazer and the Revolutionary, 24 of 45 founders read as a builder on our first read and 27 on the second, against 5 of 94 hired chief executives. Of those 94, 2 read as the Architect, 2 as the Sorcerer, and none as the Trailblazer or the Revolutionary. You can almost never appoint them. That is the whole point. The Commander is who you get when you hire. The builders are who you get when someone builds.
One honest caution. A person can shift between reads: 18 of our 45 founders read as a different character on their second sitting. The group shares did not shift. The Commander, the Architect’s family and the Sorcerer’s family each came back within two people on both runs. So we give shares, and name a founder only where both reads agree. We would rather be the weather forecaster than the black box.
Part Two
This difference is not trivia, and it is not soft. The founder is a priced variable, and the market pays a great deal for the person at the top.
Founder-led firms beat the market. A strategy that simply bought founder-CEO firms earned 8.3 percent a year above benchmark, and 4.4 percent even after controlling for size, industry and CEO characteristics (Fahlenbrach, Journal of Financial and Quantitative Analysis, 2009).
Reputation buys funding, and a better price. A founder’s track record raises both the odds of funding and the valuation it comes at, and reputation lowers a firm’s cost of capital and cushions its bad news (Hsu 2004, 2007; Pfarrer, Pollock and Rindova, 2010).
And it is priced even without substance. The market pays for founder reputation whether or not the results are there. Theranos and WeWork are that fact run in reverse, which is the surest proof that it is priced at all.
What we do not claim. Our own pilot found no link between founder character and returns. No type is a better bet than another. The value is in reading correctly: knowing which founder you are funding, how they win and how they fail, and governing them that way.
Not because the money is careless, it diligences the numbers to two decimal places, but because the deciding variable, the founder at the centre, never had a reading. You cannot govern what you cannot read. The rest of this is that reading.
Part Three
Which changes the question at the table. Not “is this a good company,” every company on your desk looks good on paper. The question is “what kind of founder am I backing, and can I read them well enough to govern it.” The rest of this is how. Read the sector first, then the person against it.
Each industry breeds its own kind of founder, so every sector has a normal, and the normal is what you read a founder against. A type that is unremarkable in one industry is a flare in another. Our samples by sector are small, so this edition gives the direction, not the decimal, and names a founder only where both of our reads agree.
Technology. The builders cluster here. The Architects and the Sorcerers in our sample are almost all technology founders (Chesky, Collison, Ek, Dorsey, Cannon-Brookes; Huang, Musk, Zuckerberg, Spiegel). So a tech founder who reads as a Commander is worth stopping on. Sometimes it is rare operating discipline in a category that needs it (Dell, Ellison, Kurtz, McKinnon); sometimes it is a sign the real visionary has already left and an administrator is holding the seat.
Finance. Finance sells control, and some of its founders read from the Commander’s family. The founders who do not are usually the interesting ones, and in our reads they are inventors: Richard Fairbank, who turned data into a new kind of lender, reads as the Scientist; Thomas Peterffy, who built a brokerage out of code, and Max Levchin both read as the Tinker. All three sit in the Sorcerer’s family.
Consumer and retail. Consumer founders read as makers and fighters more than operators (Steve Ells, the Artisan; Kevin Plank, the Warrior). A consumer brand is one person’s taste made real, so bet on how well the founder and the brand match, and treat the day the founder leaves as a real danger: a hired chief most often reads as the Commander, and a Commander cannot taste. The retail founders who built empires often read as stewards of what they built (Howard Schultz, the Shepherd; Bernie Marcus, the Worker), and the handoff from founder to hired operator is where the character of the company is kept or lost.
Read the sector first, then the person against it. The normal tells you what to expect. The exception is the thing worth the whole meeting.
Each question below is a behavioural probe aimed at the type’s failure mode. The answer matters less than what it reveals; the line beneath says what you are listening for.
Systems thinking made physical. The product is the argument, and the company is itself a designed object. Compounding advantage through iteration and taste.
Polish over shipping, the product beautiful and late. Under-investment in distribution (“it should sell itself”). People treated as components. The company inherits the founder’s introversion.
What did you ship in the last two quarters, and what did you kill?
Listening for a shipping rhythm and kill discipline, versus polishing one beautiful thing forever.
Who sells this when you are not in the room?
Listening for whether distribution exists as a system, or only as the founder’s reluctant side job.
Show me the last time a customer changed your design.
Listening for whether the market can reach the product, or the design is a monologue.
Delivers what the experts said was impossible, and recruits believers the way others recruit staff. The highest ceiling of the four: when it works, you get Nvidia.
The demo outruns the deployment. Reality distortion becomes the operating culture, and the burn scales with belief. Timelines become promises. The gap between claim and shipped widens until it swallows the company.
Walk me from your last three public claims to what shipped.
Listening for the size and direction of the gap between claim and delivery. Every Sorcerer has one; the bet depends on whether it is closing.
Who on your team tells you no, and when did you last take it?
Listening for whether a real counterweight exists, or the reality distortion is now the org chart.
What breaks if the breakthrough is eighteen months late?
Listening for a plan with dates in it, versus faith with a burn rate.
New-market instinct and appetite for risk. Sees the opening before the map exists and moves while others are commissioning studies. Builds by sequence: a trail, then the next trail.
Over-extension. The character needs a frontier, so it manufactures one, sometimes past the point the core can fund it. Operations are somebody else’s problem, and the last trail is abandoned as the next one opens.
Which trail did you finish, and what does it earn today?
Listening for at least one frontier turned into a business, versus a portfolio of exciting starts.
What are the unit economics of the newest venture, in your own words?
Listening for whether they know, not whether the numbers are good. A founder who cannot recite the economics has stopped caring about them.
What would make you stop?
Listening for any answer at all. The failure mode of the type is that nothing does.
A cause bigger than the company, aimed at an industry that has stopped questioning itself. The Revolutionary wins by changing the rules everyone else plays by, and recruits customers and staff to the cause, not just the product. The mission is the marketing. Anita Roddick, whose Body Shop took on how a whole industry made and sold its products, is the pattern; she read as the Revolutionary on both of our reads.
The cause outgrows the company. Every decision becomes a vote on principle, compromise feels like betrayal, and the founder fights the partners, regulators and board who could help the cause scale. When the system finally changes, the company can lose its reason to exist.
What are you for, in one sentence that does not mention what you are against?
Listening for a positive position that survives the old system’s fall. If the sentence needs the villain, so does the company.
Which principle have you traded for progress, and why?
Listening for whether the founder can make a trade, or treats every trade as surrender.
Who has changed your mind in the last year?
Listening for at least one person outside the cause the founder truly listens to. That person is your future board ally.
The models are not wrong. They are half a ledger: they price the business to the decimal and read the founder over a dinner. Signal Labs is a strategy consultancy for the personal brand. We read the founder and the top team, and build the founder’s brand on what we read: who they are at their best, the team on one map and where it will break, and how they show up in every room that decides.
You don’t bet on companies. You bet on people. The only edge left is whether you can read them.
Read yourself →Method. 139 leaders: 45 founders, each read twice, and 94 non-founder Fortune 100 chief executives, each read once, October 10, 2026. Read by our Decoder from public name and company; not our Diagnostic. Shares are from the first founder read, and the group shares held across both runs. A founder counts as one of the four builders if they read in the Architect’s family (Maker) or the Sorcerer’s family (Wizard), counted whole, or as the Trailblazer or the Revolutionary: 24 of 45 founders on the first read, 27 of 45 on the second, against 5 of 94 hired. A founder is named only where both reads agree; hired chief executives are given as counts, never names. Sector samples are small; read the direction, not the decimal. Our own pilot found no link between founder character and returns. Reputation findings are drawn from the peer-reviewed literature cited above. A directional argument about the shape of the field, not a verdict on any individual. This second edition replaces the September 2026 edition. Signal Labs · signallabs.ca · The Behaviour Code.