Trust as a System, Not a Personality Trait

Why 'I just don't trust them yet' isn't a useful starting point

It's common to hear founders describe trust as something that either exists or doesn't, almost like a personality match — you either click with someone and trust them, or you don't. This framing is a problem because it gives you nothing to actually do about it. If trust is just a feeling, the only options are to wait and hope, or to give up on the relationship, neither of which is much use when you're running a small team that needs to function now.

A more workable model treats trust as something built deliberately out of three components: what authority someone has been given, what evidence exists about how they've used it, and how fairly that evidence gets reviewed. Change any one of the three and trust changes with it — which means you have levers to pull rather than a mood to wait out.

Component one: stated authority

Trust can't accumulate around a decision nobody has been explicitly given the authority to make. If a team member is quietly making judgement calls that were never formally handed to them, any mistake feels like a betrayal of trust that was never actually granted — and any success goes unnoticed because it was never framed as a test of anything.

State the authority clearly and in writing where it matters: what they can decide alone, up to what limit, and what needs a check first. This is the same idea covered in decision rights and delegation, and it's the foundation trust is built on top of — without it, there's no clear baseline to measure evidence against.

Component two: evidence, deliberately generated

Trust needs something to be built from, and that means someone has to be given a chance to act before they've fully proven themselves — otherwise nobody ever gets the opportunity to generate the evidence at all. This is uncomfortable for founders who want proof before they extend any authority, but it's the wrong order: proof only exists after authority has been used a few times.

The practical approach is to extend a small, bounded amount of authority first — a modest spending limit, a narrow decision area — and treat the first several instances as evidence-gathering, not as a test with instant pass-or-fail stakes. Keep a simple private note of how it goes: did they use the authority sensibly, did they flag the right things, did they ask when they should have asked. This becomes the factual basis for extending more authority later, rather than a vague impression.

Component three: fair, regular review

Evidence only builds trust if it's actually reviewed, and reviewed fairly. If good decisions go unacknowledged and only mistakes get noticed, the felt experience for your team is that trust can only go down, never up — which discourages exactly the kind of initiative you're trying to encourage.

Build review into a rhythm you already have, like a one-to-one, rather than treating it as a special, occasional event. Name specific instances: 'you decided X last week and that was the right call because Y' matters as much as addressing what went wrong. Over several cycles, this explicit review is what actually shifts a working relationship from cautious to genuinely trusted, faster than time alone would.

• Keep short, factual notes on how authority has been used, not general impressions
• Acknowledge good judgement calls specifically, not just as a passing 'good job'
• Address concerning decisions promptly and specifically, tied to the actual event
• Use a fixed review point (weekly or fortnightly one-to-one) rather than ad hoc conversations
• Adjust the stated authority level explicitly when the evidence supports it, in either direction

One mistake shouldn't erase a pattern

A common way trust breaks down unfairly is when a single mistake, especially a visible or costly one, gets treated as proof that trust was misplaced all along — even against a longer run of good decisions. This is a natural human bias (recent, vivid events dominate memory), but it's corrosive if you let it drive how you treat someone going forward.

When something goes wrong, look at it against the pattern, not in isolation. Was this one decision out of many, most of which were sound? Was it a genuinely new situation the person hadn't faced before, or a repeat of something they should already have learned from? Treating a single miss proportionately, rather than as a verdict on the whole relationship, is often the difference between someone growing into more responsibility and someone quietly disengaging because they feel one mistake wiped out months of good work.

When distrust is real, name it directly

Sometimes the evidence genuinely doesn't support extending more trust, and the honest, if uncomfortable, thing to do is say so directly rather than let it leak out as unexplained hovering or reduced authority with no stated reason. An unnamed loss of trust is far more damaging to a working relationship than a direct, specific conversation about what would need to change.

A workable script: 'I want to be straightforward with you — after the last two client calls went the way they did, I'm going to ask you to loop me in before you commit to a delivery date, until we've had a run of these go well. That's not permanent, and here's what I'd want to see to change it back.' This is honest, specific, and gives a path forward, which is far better for the relationship than silent, unexplained restriction.

Rebuilding trust after it's been damaged

Rebuilding trust after a serious breach — a missed deadline that damaged a client relationship, a decision made outside stated authority, a repeated pattern of avoidable mistakes — needs the same explicit system as building it in the first place, just starting from a lower baseline. Skipping straight back to full authority because enough time has passed, without any fresh evidence, tends to produce another disappointment.

Agree explicitly what a reasonable path back looks like: a narrower authority level, a shorter review cycle, specific things you'll be looking for. Frame it as a plan with an end point in sight, not an indefinite punishment, so the person has something concrete to work towards rather than a vague sense they're permanently under a cloud.