Trust: How It Gets Built, Spent and Lost

"I trust them" is three different sentences

You'd trust a friend with a secret but not with returning a laptop on time. You'd trust a colleague to hit a deadline but not to keep a confidence. You'd trust a doctor's diagnosis while not particularly liking them.

These feel like different amounts of one thing. They aren't — they're different things entirely, and treating trust as a single dial is why people get it so wrong about themselves.

Because here's the part that stings: when trust in you is low, you almost certainly know it. What you usually don't know is which of the three components is the one that's missing — and since the repair for each is completely different, guessing wrong means working hard on the wrong thing for months.

---

The structure: three components, not one

The research converges on roughly the same three. Different names, same shape.

1. Reliability — do you do what you said?

The most mechanical, the most measurable, the most commonly broken.

Not "are you a good person." Just: the gap between what you said you'd do and what happened. Did the message arrive when you said? Did the work come Friday? Were you there at seven?

Small promises count more than large ones, which is deeply unfair and completely true. Nobody tracks whether you delivered the big thing — that was visible and everyone remembers. What gets tracked, silently, is the pattern of small ones: the "I'll send it tonight" that arrived Tuesday, four times.

The fix is unglamorous and almost mechanical: promise less, and narrow it. "I'll get to it this week" is a promise you can fail without noticing. "I'll send it Thursday" is one you can keep.

2. Competence — can you actually do the thing?

Separate from reliability, and often confused with it. Someone can be entirely dependable and still not be who you'd ask.

The nuance worth having: competence is domain-specific and people over-generalise it wildly. Being excellent at one thing buys unearned credibility in unrelated things — which is why confident people in one field are dangerous in another, and why "I don't know, that's outside what I'm good at" is a sentence that raises your credibility rather than lowering it.

Overclaiming is the fast way to destroy this one. Saying you can do something and then not being able to costs more than declining would have — because now the failure is visible, and it retro-taints everything you said you could do.

3. Care — are you acting in my interest, or just yours?

The hardest to demonstrate and the one that decides how the other two get read.

If someone believes you're on their side, a missed deadline is a mistake. If they don't, the same missed deadline is evidence. Care is the lens the other two are seen through.

It's demonstrated in small, cheap, specific ways — and, crucially, when there's nothing in it for you: remembering the thing they mentioned once · flagging a problem that makes you look worse · telling them something they'd rather not hear · not using something they told you in confidence when using it would have helped you.

The strongest signal is a cost you paid that they can see. Care claimed costs nothing and proves nothing. Care that cost you something is the only version that reads.

---

How trust actually moves

Not symmetrically, and that's the practical part.

It builds slowly, in small units. There's no single act that establishes it. It accumulates through repetition of unremarkable things — which is why new relationships feel effortful and long ones feel easy.

It's spent, not just held. Asking a favour, being late, needing patience, taking a risk on someone's behalf — all of these draw down a balance. Which is fine. Trust is for spending; a relationship where nobody ever draws on it isn't trusting, it's distant. The failure is spending without noticing and never repaying.

It falls fast and unevenly. Negative events weigh more than positive ones — the "bad is stronger than good" asymmetry is one of the more robust findings in social psychology. One clear breach can undo a long run of reliability, and no amount of protesting that the ratio is unfair changes the arithmetic.

Which breach matters depends on the component. Break reliability and you're careless. Break competence and you're out of your depth. Break care and it's different in kind — it's not a mistake, it's information about who you are. Care breaches are the ones that don't recover on their own.

---