I've interviewed forty people for one role and hired nobody. Is my bar wrong, or is it really the market?
What the lenses had in common. Sharpest when a pattern names a tension you hadn’t fully named yourself — weakest when it summarises something you already know.
Reflection Check and Premortem both attack the same weakness: you have no evidence your bar measures capability rather than familiarity. Put two strong current staff through the loop, and write a specific predicted failure for the three best people you rejected. If either exercise comes back thin, the market was never the problem.
From The Reflection Check The Pre-Mortem
Braintrust's insight is structural: nobody says the true thing unless it's safe to. If a bad hire damages an interviewer's standing and a missed hire costs them nothing, forty rejections is exactly what the system was built to produce. Making the vacancy's monthly cost visible changes the arithmetic in the room.
From Pixar's Braintrust Semmelweis and Handwashing
Ants in a death spiral are each following the rules correctly. All forty rejections on one page, with stated reasons, will surface the killing stage or the never-approving interviewer in about an hour — and none of it is visible from inside a single defensible decision.
From The Ant Colony
Often the sharpest provocation lives inside a single lens, not in the synthesis above. Skim them; one tends to land harder than the others.
You haven't earned your rebuttal until they agree you understood them.
Read this lensBefore trusting a measurement, check the instrument against a known quantity. Forty rejections is not primarily information about forty candidates — it's information about your process, and you have never calibrated it. The single cheapest test exists and almost nobody runs it: put your current best-performing team members through your own interview loop. If they'd fail, you are not measuring capability. You are measuring something else and calling it a bar.
Run your two strongest team members through the real loop with a colleague who doesn't know them. If either would be rejected, the instrument is broken and every conclusion you've drawn from forty interviews is unsafe. This takes one afternoon and it's the only honest starting point.
The Rogerian / active-listening discipline: repeat the other side's position back in your own words until they confirm 'yes, that's right' — then you may respond.
The principle: Most people in conflict aren't arguing, they're waiting to be understood. The reflection check removes the thing they're actually fighting for — being heard — and the heat usually drops with it.
Right, provable, and rejected anyway.
Read this lensSemmelweis showed that doctors washing their hands cut maternal deaths dramatically. The profession rejected it for decades — not because the data was weak but because the implication was intolerable: that respectable physicians had been killing their patients. The evidence pointed at the people evaluating the evidence. Forty failed interviews point somewhere too, and the reason the 'it's the market' explanation is so comfortable is that the alternative implicates you.
Email five rejected candidates and ask, genuinely, what the process was like and where it felt off. Some won't reply. The ones who do will tell you things no internal review will surface — and the discomfort you feel about sending that email is the Semmelweis reflex, on time and on schedule.
In 1847 Vienna, Ignaz Semmelweis showed that maternal mortality fell tenfold when doctors washed their hands between cadavers and the maternity ward. The data was unimpeachable. The medical establishment rejected it as an insult to gentleman-doctors. Semmelweis died in an asylum. Decades later, germ theory vindicated him.
The principle: Evidence does not move people who would be insulted by the conclusion. The strongest data fails when the implication is 'you have been killing people for years'. The carrier of the truth often pays a higher price than the people refusing to hear it.
Tear the work apart without tearing the room apart.
Read this lensPixar's Braintrust works because of two rules: it has no authority over the director, and it critiques the film rather than the film-maker. That combination is what lets people say the true thing. Most hiring loops fail on exactly this — feedback is gathered from interviewers who know a hire means less work for them, or who fear being blamed for a bad hire, and nobody is structurally free to say 'this candidate is fine and we are being precious'.
Put a number on the vacancy — the delayed project, the overtime, the burnout on the team covering it, per month. Bring it to the next debrief. Hiring loops reject endlessly when the cost of 'no' is invisible and the cost of 'yes' is vivid.
Pixar's Braintrust gives directors brutally candid feedback on unfinished films — but the notes carry no authority; the director decides what to use. Candour is total, control stays with the maker.
The principle: Hard disagreement and a working alliance can coexist if you separate honest input from command. The fight stays about the work because nobody can force the outcome.
Imagine the project has already failed. Now write the obituary.
Read this lensA premortem assumes the failure has already happened and asks the team to explain why. It works because it licences people to voice doubts that sound disloyal in advance. Run it on your best rejected candidate: it is now eighteen months later and hiring them was a disaster — why? If the answer is thin, or reduces to a communication style you found grating, you rejected a good hire for taste. If it's substantial, you'll finally have written down what the bar actually is.
Do this for the three best candidates you turned away. Write the predicted failure for each in two sentences. Vague answers mean you were pattern-matching on similarity, which is both a hiring failure and a legal exposure — and it's the most common cause of a forty-interview drought.
Klein's pre-mortem inverts retrospective analysis: before a decision is committed, the team imagines themselves a year in the future, with the project having ended in disaster. Each person privately writes down why it failed. Because the premise is settled — failure happened — people who hesitated to voice doubts feel safe to articulate them. The technique reliably surfaces 30% more failure modes than open 'what could go wrong' brainstorming.
The principle: Pessimism phrased as fact is heard more cleanly than pessimism phrased as opinion. Move the question from 'might it fail' to 'why did it fail' and the silent worries become visible.
Leave signals, not instructions.
Read this lensAn ant colony finds the shortest path to food with no ant knowing the route. Simple local rules and pheromone reinforcement produce sophisticated collective behaviour. But the same mechanism produces the death spiral: a circular pheromone trail that ants follow until they die of exhaustion, each one behaving correctly. Your loop may be in one. Every individual rejection was defensible; the aggregate is a circle nobody chose and nobody can see from inside it.
Lay all forty rejections out on one page with the stated reason for each. Patterns invisible at the level of a single decision — a stage that kills everyone, one interviewer who never approves, a demographic skew — become obvious in aggregate, and this is the single highest-yield hour available to you.
An ant colony with no manager finds the shortest path to food, builds a nest, defends it, sorts the brood by age. The mechanism is pheromone trails — each ant lays a chemical signal slightly stronger when it finds food, slightly weaker when the path runs cold. The colony's intelligence is in the signals, not in any individual ant.
The principle: Signals scale where instructions don't. A system that broadcasts what is working and what is failing — and lets the individual actors decide what to do with the signal — coordinates without anyone holding the plan in their head.
Lenses ranked relevant but kept off the exploration — the runner-up reading. Each links to the lens itself, not an applied reading of your question.
Buy what the market under-prices, not what the market wants.
Build their argument stronger than they did, then answer that.
Wear one mode of thinking at a time, in turn, deliberately.
Keep your opponents inside the tent.
Everyone agreed to a trip nobody wanted. Silence is a decision too.
Make it safe to stop the line, or the fault ships anyway.
Argue against the problem, not across the table.
The breakthrough belonged to whoever published first.
drag · scroll to zoom · click anything