没有人在买「你这个人」
他们在买「选了你之后，我会不会后悔」

EN Nobody is buying you. In a hiring, funding, partnership or resource conversation, you are a product that cannot be opened and inspected before payment — so the buyer isn't admiring you, they're pricing uncertainty. And the sharpest part of that uncertainty is personal: if this goes wrong, whose signature is on it. Four scenes, one machine. The hiring manager buys not having to re-run the search in six months and fears being visibly wrong. The investor buys a shot at something large plus your not walking away, and fears looking foolish to peers. A prospective co-founder buys two years they don't have to spend themselves, and fears being locked in. A decision maker granting you resources buys their own goal moving forward, and fears spending their credibility on you for nothing. Adjectives — passionate, fast learner, hard working — cost nothing to say, which is exactly why they carry no information: signalling theory says a signal only separates people when it is cheaper for the genuine article than for the pretender. Verifiable work, a named referee, a real exit clause: those cost the faker more than they cost you.

EN There is exactly one test for a personal value proposition: can it be proven wrong? "Responsible, good communicator, product-minded" excludes nobody and can be falsified by nothing, so it carries zero information. "I took three sub-twenty-person e-commerce teams from a two-day reconciliation to two hours, and one of them will take your call" can be checked, and can blow up in my face — which is why it counts. Four parts: who exactly, from what state to what state, where the evidence sits, and who should not hire you. The fourth is the counter-intuitive one: naming what you're bad at makes the other three more credible, because someone who can do anything has told you nothing. The underlying finding from personnel selection is that concrete past behaviour predicts future performance far better than self-described traits. Be honest about the strength of it, though: Schmidt and Hunter's famous 1998 validity table was substantially revised downward by Sackett and colleagues in 2022 after correcting for range-restriction artefacts. The ordering survived — structured, behaviour-based evidence still comes out on top — but treat the individual coefficients as contested, not settled.

EN The cringe isn't modesty and it isn't ethics — it's a bookkeeping error. You've recorded self-promotion as taking something from the other person, when the actual entry is that someone who needs your kind of help currently can't find you, and you both lose. Three findings make the ledger clearer. First, the illusion of transparency: you badly overestimate how visible your nervousness is, and simply being told this reduces speech anxiety and improves how observers rate the talk. Second, the liking gap: after a conversation, people systematically underestimate how much the other person liked them — the version of you that just spoke landed better than the version replaying in your head. Third, and most usefully, disclosing information about yourself activates the dopamine reward circuit — nucleus accumbens and ventral tegmental area — strongly enough that people will give up money for the chance to do it. Which reframes the whole thing: talking about yourself is a reward, so a monologue takes that reward away from them. Ask first, listen for their exact words, then place your evidence next to the problem they just named. The honest boundary is that self-promotion does carry a real social cost, and the research shows it falls unevenly — which is an argument for third-party evidence over adjectives, not for silence.

EN The biggest opportunities in your life mostly don't arrive through a pitch you delivered. They arrive because someone you barely know mentioned you, in one sentence, in a room you weren't in. So the thing worth polishing isn't your self-introduction — it's the sentence they can remember and repeat. Two constraints: short enough to survive being retold, specific enough that it couldn't be retold about someone else. "A friend of mine's pretty good, does internet stuff" fits anybody and therefore hooks onto no actual need. "There's a guy who rescues the ancient reports nobody dares touch" is narrow and visual, so it fires automatically the moment someone complains about exactly that. Then hand it over with a trigger attached — "if you ever hear someone moaning about legacy reporting, throw my name in" — because people don't scan the world for opportunities on your behalf, they retrieve a name when a cue shows up. Granovetter's 1973 survey found most people who found work through contacts heard about it from someone they saw only occasionally; the sample was small and half a century old, but a 2022 Science paper running randomised experiments across roughly twenty million LinkedIn users over five years confirmed the causal version: moderately weak ties do the most for job mobility.