IX Coach

The Trust Triangle

The four distinct assessments that make up trust — and how to repair each one specifically

What is the Trust Triangle and how do you use it?

The Trust Triangle is a model that breaks trust into separate assessments you can diagnose one at a time, rather than treating trust as a single all-or-nothing thing. In Charles Feltman’s version (The Thin Book of Trust) there are four: sincerity (I mean what I say), reliability (I do what I say), competence (I have the ability to do what I say), and care (I have your interests at heart). Trust breaks down when any one of these assessments fails — and repairing trust requires identifying which assessment was damaged, then addressing that specific one rather than trying to rebuild trust in general.

Most people treat trust as a single thing that you either have or don’t. Charles Feltman’s framework disagrees: trust is a judgment made across four separate dimensions, and knowing which dimension failed is the key to targeted repair. A colleague who is sincere and competent but unreliable has a very different problem than one who seems reliable but doesn’t actually care about your interests. The practices here apply the Trust Triangle to building, maintaining, and repairing trust in real relationships.

Practices

How these practices connect

Each practice targets one of four distinct trust assessments Feltman identifies -- sincerity, reliability, competence, and care -- rather than treating trust as one undifferentiated quantity. The underlying mechanism is that trust is repaired and built the same way: by closing the gap between what a person signals (in words or commitments) and what they actually do, dimension by dimension. Global gestures like a generic apology or a vague promise to 'do better' fail because they don't target the specific assessment that broke.

How do I know which trust practice I actually need?

Start by diagnosing which of the four trust dimensions -- sincerity, reliability, competence, or care -- is actually broken, rather than assuming trust is damaged in general. A sincerity failure needs honesty correction, a reliability failure needs behavioral follow-through, a competence failure needs capability building or role clarification, and a care failure needs visible alignment of interests. Picking the wrong repair, like offering to try harder when the real issue is that someone doesn't believe you consider their interests, wastes effort and can deepen the rupture because it doesn't address what was actually assessed as broken.

Why does making smaller promises build more trust than making bigger ones?

Reliability trust is built by the ratio of promises kept to promises made, and that ratio degrades fast when ambition outruns capacity. Making smaller, guaranteeable commitments and consistently exceeding them shifts the evidence base in your favor, whereas large aspirational promises that are routinely unmet teach the other person to treat your word as an estimate rather than a commitment. This connects directly to sincerity: a promise made in optimism still lands as a broken promise to the person relying on it, regardless of your intent.

Is admitting I don't know something going to make people trust me less?

Being transparent about competence limits protects trust rather than damaging it, because the real threat to competence trust is being caught overstating ability, which combines a competence gap with a sincerity failure at the same time. Naming what you can do, what you can't, and what you'd need to learn gives the other person accurate information to calibrate their reliance on you. The evidence caveat here matters: in some environments expressing uncertainty gets misread as weakness, and this benefit is most reliable in relationships that already have some goodwill.

If someone hasn't earned my trust yet, why would I extend it first?

Waiting for someone to fully earn trust before extending any is a self-fulfilling trap, because neither party ever gets the opportunity to demonstrate trustworthiness. Trust is invited by extending it first, in bounded, low-stakes situations, which is the mechanism that makes new relationships and post-rupture repair possible at all. This carries real risk -- extending too much too fast leaves you exposed, and the right amount to extend is a judgment call rather than a formula, so it works best as a deliberate, incremental move rather than a single leap.

Someone keeps not doing what they said -- should I just talk about the missed tasks or bring up trust directly?

Framing it only as a performance issue -- 'you didn't do what you said' -- addresses the symptom but leaves the underlying trust assessment unresolved, because it doesn't name that your reliance on the person has actually changed. An explicit trust conversation, naming directly that this has affected your trust, is the only move that addresses the assessment itself rather than managing around it through avoidance or escalation. This does require a baseline of relational safety; in some cultural contexts or power-imbalanced relationships, that directness carries real risk and the calculation of whether to have the conversation changes accordingly.

Can I just tell someone I care about their interests instead of proving it?

No -- care trust is specifically damaged when someone appears to act only in their own interest, and it is repaired and built through visible acts, not statements, especially ones that cost you something. Performative care-signaling without an actual cost is often detected as such, and when it is, it doesn't just fail to build care trust, it erodes sincerity trust too, compounding the original problem. The strongest signal is visible sacrifice or advocacy that benefits the other person during a moment of competing interests, which is also exactly when care is most needed and least assumed.

Do all these trust practices come from the same research, or is some of this more solid than the rest?

No -- the evidence quality varies across these practices. Diagnosing the specific trust dimension and having explicit trust conversations are described as mechanistic: well-reasoned and directionally supported, though Feltman's four-dimension labels are a practitioner framework rather than a single validated taxonomy. Sincerity, reliability, competence transparency, care, and trust extension are observational: consistently reported patterns, some widely cited in organizational settings, but without the same weight of formal experimental study behind them as trialed interventions.

How do you choose between these?

The choice depends on where you are in a specific situation, not on who you are as a person. If trust already feels broken and you don't know why, start by diagnosing which dimension failed before doing anything else, since the wrong repair can waste the opportunity or deepen the rupture. If you're in a new relationship or rebuilding after a rupture with nothing yet to diagnose, extending trust first in a bounded way is what creates the conditions for any of the other dimensions to be demonstrated. If a specific commitment was missed, reliability repair through smaller promises applies; if the gap is about what you said versus what you meant, it's sincerity; if it's about whether you're capable, competence transparency; if it's about whose interests are being served, visible care. When the problem has been sitting unaddressed and is affecting how you rely on someone, the trust conversation is what surfaces it directly instead of letting it drift under a performance-issue framing.

What these approaches do not settle

These practices don't fully resolve tension with each other: extending trust first is explicitly a bet made before evidence exists, while the other six practices are about building evidence, so the model doesn't specify exactly how much unearned trust to extend before diagnosis becomes possible. Most of the mechanisms here are observational and drawn from organizational or practitioner contexts rather than controlled trials, and Feltman's four-dimension framework itself is one practitioner taxonomy among several used in trust repair research, not a settled consensus. The trust conversation and full transparency about competence both assume a baseline of relational safety or existing goodwill that not every relationship has, and in power-imbalanced or high-risk contexts the direct approaches this page recommends may carry more cost than benefit.

Practice this with IX Coach

Practice this with IX Coach

Frequently asked questions

What is the Trust Triangle and how do you use it?

The Trust Triangle is a model that breaks trust into separate assessments you can diagnose one at a time, rather than treating trust as a single all-or-nothing thing. In Charles Feltman’s version (The Thin Book of Trust) there are four: sincerity (I mean what I say), reliability (I do what I say), competence (I have the ability to do what I say), and care (I have your interests at heart). Trust breaks down when any one of these assessments fails — and repairing trust requires identifying which assessment was damaged, then addressing that specific one rather than trying to rebuild trust in general.

How do you diagnose which trust dimension is broken before trying to repair?

When trust breaks down, the instinct is to apologize for "not being trustworthy" — a global statement that addresses none of the four dimensions precisely. Feltman’s insight is that each dimension fails for different reasons and requires different repair. Sincerity failures require honesty correction; reliability failures require behavioral follow-through; competence failures require capability building or role clarification; care failures require visible alignment of interests. Generic trust repair misses the specific lever.

How do you build sincerity by saying only what you mean?

Sincerity trust is damaged incrementally by small misalignments between what is said and what is meant — social white lies, overpromising to avoid conflict, saying "I understand" when you don’t. Each instance is small; cumulatively they create the sense that the person’s words are not reliable guides to their actual state. The repair and the prevention are the same: close the gap between internal state and verbal output.

How do you build reliability by making smaller, keepable promises?

Reliability trust is built by the ratio of promises kept to promises made — a ratio that degrades fast when ambition outruns capacity. The common failure is making large, aspirational commitments that feel motivating to state but are routinely unmet, while the relationship learns that promises are estimates. Deliberately scaling down to commitments you can guarantee shifts the evidence base: every kept small promise adds to the reliability account.

How do you be transparent about your competence limits?

Competence trust is damaged in two ways: by genuinely lacking ability, or by overstating ability and being caught. The second is far more damaging, because it combines a competence gap with a sincerity failure. Proactively naming the limits of your competence ("I know this domain well; this adjacent area I’d need to research") protects sincerity trust and gives the other person accurate information with which to calibrate their reliance on you.

How do you demonstrate care through visible acts in the other person’s interest?

Care trust is the belief that the other person considers your interests, not only their own. It is damaged when someone appears to act in their own interest at the other’s expense, even if unintentionally. The repair and the building are the same: make visible the ways in which you considered their interests — especially when it cost you something. Visible sacrifice or advocacy that benefits them at some cost to you is the strongest care signal.

How do you extend trust deliberately to invite it?

Feltman also defines distrust: the decision to not rely on another’s sincerity, reliability, competence, or care. Trust cannot be demanded, only invited — and it is built by extending it first in low-stakes situations, providing the other person the opportunity to demonstrate trustworthiness. Demanding full trust before any extension is a self-fulfilling trap: neither party can demonstrate they’re worthy of it.

How do you have explicit trust conversations when trust is at stake?

Trust problems are almost universally handled indirectly: through avoidance, rumor, or escalation rather than explicit conversation. Indirect management allows assumptions to harden and the relationship to drift further. An explicit trust conversation — "I want to talk about something that has affected my trust in you" — is uncomfortable but is the only move that addresses the underlying assessment rather than its symptoms.

How do I know which trust practice I actually need?

Start by diagnosing which of the four trust dimensions -- sincerity, reliability, competence, or care -- is actually broken, rather than assuming trust is damaged in general. A sincerity failure needs honesty correction, a reliability failure needs behavioral follow-through, a competence failure needs capability building or role clarification, and a care failure needs visible alignment of interests. Picking the wrong repair, like offering to try harder when the real issue is that someone doesn't believe you consider their interests, wastes effort and can deepen the rupture because it doesn't address what was actually assessed as broken.

Why does making smaller promises build more trust than making bigger ones?

Reliability trust is built by the ratio of promises kept to promises made, and that ratio degrades fast when ambition outruns capacity. Making smaller, guaranteeable commitments and consistently exceeding them shifts the evidence base in your favor, whereas large aspirational promises that are routinely unmet teach the other person to treat your word as an estimate rather than a commitment. This connects directly to sincerity: a promise made in optimism still lands as a broken promise to the person relying on it, regardless of your intent.

Is admitting I don't know something going to make people trust me less?

Being transparent about competence limits protects trust rather than damaging it, because the real threat to competence trust is being caught overstating ability, which combines a competence gap with a sincerity failure at the same time. Naming what you can do, what you can't, and what you'd need to learn gives the other person accurate information to calibrate their reliance on you. The evidence caveat here matters: in some environments expressing uncertainty gets misread as weakness, and this benefit is most reliable in relationships that already have some goodwill.

If someone hasn't earned my trust yet, why would I extend it first?

Waiting for someone to fully earn trust before extending any is a self-fulfilling trap, because neither party ever gets the opportunity to demonstrate trustworthiness. Trust is invited by extending it first, in bounded, low-stakes situations, which is the mechanism that makes new relationships and post-rupture repair possible at all. This carries real risk -- extending too much too fast leaves you exposed, and the right amount to extend is a judgment call rather than a formula, so it works best as a deliberate, incremental move rather than a single leap.

Someone keeps not doing what they said -- should I just talk about the missed tasks or bring up trust directly?

Framing it only as a performance issue -- 'you didn't do what you said' -- addresses the symptom but leaves the underlying trust assessment unresolved, because it doesn't name that your reliance on the person has actually changed. An explicit trust conversation, naming directly that this has affected your trust, is the only move that addresses the assessment itself rather than managing around it through avoidance or escalation. This does require a baseline of relational safety; in some cultural contexts or power-imbalanced relationships, that directness carries real risk and the calculation of whether to have the conversation changes accordingly.

Can I just tell someone I care about their interests instead of proving it?

No -- care trust is specifically damaged when someone appears to act only in their own interest, and it is repaired and built through visible acts, not statements, especially ones that cost you something. Performative care-signaling without an actual cost is often detected as such, and when it is, it doesn't just fail to build care trust, it erodes sincerity trust too, compounding the original problem. The strongest signal is visible sacrifice or advocacy that benefits the other person during a moment of competing interests, which is also exactly when care is most needed and least assumed.

Do all these trust practices come from the same research, or is some of this more solid than the rest?

No -- the evidence quality varies across these practices. Diagnosing the specific trust dimension and having explicit trust conversations are described as mechanistic: well-reasoned and directionally supported, though Feltman's four-dimension labels are a practitioner framework rather than a single validated taxonomy. Sincerity, reliability, competence transparency, care, and trust extension are observational: consistently reported patterns, some widely cited in organizational settings, but without the same weight of formal experimental study behind them as trialed interventions.

Research

Related concepts

Articles that cover this