You're standing in front of a used car, and you don't know anything about engines. You have no idea whether the timing belt has been replaced, whether the transmission is failing, whether the mileage on the odometer is honest. The question you're supposedly answering — is this a good car — requires expertise you don't have and can't acquire in the twenty minutes you're standing there.
So your brain quietly does something else. It answers a different question instead: do I trust the person selling it to me?1 And it hands you that answer with the same confidence it would have given you if you'd actually inspected the engine, because as far as your conscious experience is concerned, you never noticed the swap happened.
This is attribute substitution — Kahneman's term for what the mind does when asked a hard question it can't efficiently answer: it silently substitutes an easier, related question, answers that one, and reports the answer as though it addressed the original.1 You didn't decide to change the subject. The substitution happens below the threshold where you'd notice, which is exactly what makes it powerful — and exactly what makes it exploitable by anyone who understands the swap is happening.
The hard question, in the car example, is a technical one: is this machine mechanically sound. The easy substitute question is a social one: is this person the kind of person who'd lie to me. Humans have roughly half a million years of evolved, richly-developed circuitry for answering the social question — reading faces, tracking reputations, detecting deception in tone and posture. We have essentially zero evolved circuitry for assessing a combustion engine.1 The substitution isn't laziness. It's your brain correctly routing the decision to the department that's actually equipped to make it — it's just that the department it routes to isn't the one the original question was addressed to.
Once you see the swap, it explains a startling amount of everyday decision-making that otherwise looks irrational. Buying a house. Choosing a contractor. Picking a coffee machine you have neither the time nor the engineering background to research fully. In each case, the stated question is technical or comparative — which option is objectively best — and the actual question being answered is almost entirely social: does this person seem trustworthy, competent, like they'd be ashamed to sell me something bad.2
This produces a specific, testable prediction: substitution is strongest exactly where the category is one you have no personal expertise in, and weakest where you do. Someone who races cars for a hobby is far less likely to substitute the trust-question for the engine-question, because they can actually answer the engine-question themselves. The substitution isn't a fixed trait of the person; it's a function of the gap between the question's difficulty and the asker's specific competence.
The cleanest demonstration of the mechanism working against the seller's short-term interest — which is also the strongest evidence it's a real cognitive phenomenon and not just a convenient story sellers tell about themselves — comes from a camera shop.3
An elderly, comfortably wealthy couple walked in wanting to buy a roughly $900 camera. The salesperson asked a series of questions, established they'd realistically use perhaps four percent of that camera's capability, and down-sold them to something cheaper and better suited. On a narrow, transactional reading, this makes no sense — the salesperson took a direct commission hit for no visible reason.
But the salesperson wasn't optimizing the sale in front of them. They were optimizing the trust-question a future customer would unconsciously be answering, the next time this same couple — or anyone they told the story to — needed camera advice. Professional pride was explicitly part of it: some part of the salesperson's own identity was staked on giving the right recommendation rather than the maximally profitable one, in a category (specialty retail, high perceived expertise) where that professional identity is unusually strong.3 The short-term commission loss purchased something with a longer payoff: the customer's answer to "do I trust this person" got permanently, expensively reinforced — expensively for the shop in the moment, cheaply in terms of what it will be worth the next time that trust-question gets asked.
You're advising a solo consultant who's losing pitches to a larger competitor with a worse track record but a slicker deck. Don't tell them to build a better deck. Ask what question the prospect is actually answering when they choose between the two options — it isn't "which consultant has objectively delivered better outcomes," because the prospect usually can't verify that claim either way in the time available. It's "which one of these two people do I trust to be straight with me when something goes wrong." Redirect the pitch prep away from credentials and toward whatever concretely demonstrates trustworthiness in the room — specific, checkable claims, visible willingness to say "that's not something I'd recommend," anything that gives the substituted question a clean, positive answer.
Later, you're reviewing a junior salesperson's approach and notice they're leading every pitch with technical specifications the customer has no way to independently verify or even fully understand. Point out what's actually happening in the customer's head during that pitch: they're not evaluating the specs, they're evaluating the salesperson, using the specs pitch itself as the input — does this person seem like they know what they're talking about, does their confidence feel earned or performed. Coach the salesperson to notice they're being judged on a different axis than the one they think they're presenting on.
A year later, a client asks why their online reviews — genuinely excellent, verified, detailed — aren't converting sales the way word-of-mouth referrals do. Explain the substitution mechanism directly: a written review, however positive, is a weaker input to the trust-question than a referral from someone the buyer already has an independent trust relationship with, because the review still requires the reader to separately trust that the review itself is genuine — an extra inferential step a personal referral skips entirely.
The strongest evidence is the down-sell case study specifically because it runs against the seller's immediate financial interest — a self-interested account of "buyers trust trustworthy sellers" would be easy to dismiss as motivated reasoning, but a seller's own account of voluntarily reducing their commission for the sake of the trust-relationship is harder to explain away as anything other than the mechanism actually operating.
The tension the source doesn't resolve: attribute substitution is presented as an automatic, largely unconscious process, but the down-sell case study describes a salesperson who seems to understand the trade-off consciously and is deliberately playing the long game. If experienced practitioners can consciously exploit their own customers' unconscious substitution, is the mechanism itself unconscious, or is it unconscious only in the buyer while remaining fully conscious and strategic in the seller? The source doesn't distinguish these.
Open question: does attribute substitution weaken with repeated exposure to the same seller — does trust accumulated over multiple interactions eventually let the buyer answer something closer to the real underlying question directly, or does the substitution simply persist indefinitely, with trust functioning as a permanent proxy rather than a bridge toward genuine expertise?
This directly extends the general behavioral-economics literature on heuristics and biases (Kahneman & Tversky's original attribute-substitution framing, applied here specifically to commercial trust decisions rather than the classic judgment-under-uncertainty examples like the Linda problem). Sutherland's contribution isn't the mechanism itself, which is well-established, but the specific application: showing that an entire industry's worth of personal-branding and reputation-management practice can be read as downstream consequences of one well-documented cognitive shortcut, rather than as a separate, sui generis marketing phenomenon.
Behavioral-Mechanics — Manipulation Technique: Authority Costume. (mandatory handshake — this psychology page names a mechanism with a direct tactical counterpart per the psychology-behavioral-handshakes standard.) The authority-costume technique exists specifically because attribute substitution runs on any plausible signal of trustworthiness or expertise, not only genuine ones — a costume (a white coat, a confident voice) can trigger the exact same substitution this page describes, supplying the buyer's brain with a cheap, fabricated answer to "does this person seem like they know what they're talking about" that gets accepted with the same confidence a genuine answer would carry. The insight the pairing produces: this page explains why the substitution happens (the buyer has no efficient way to answer the real question); that page shows the substitution has no built-in fidelity check — it accepts a costumed signal as readily as an earned one, which is exactly what makes attribute substitution exploitable rather than merely convenient.
(Note: likability-trumps-performance-in-service-brands — a page covering the Royal Mail "brand quake" case, a strong structural match for this mechanism at brand scale — is part of a parallel writer's assignment in this same batch and may not exist yet at time of writing. Once built, consider adding it as a third handshake or swapping it in for the one above; verify it exists before linking.)
Behavioral-Mechanics — Value Comparison vs. Premium Justification. That page's split (customers either compare products or justify a purchase against their own identity) can now be read as two different answers to two different substituted questions. The value-comparison lane substitutes "which option scores best on visible features" — itself already a simplification of the true underlying question of long-term satisfaction. The premium-justification lane substitutes something closer to this page's trust-and-identity question directly. The insight neither page alone produces: premium branding doesn't eliminate attribute substitution, it redirects which substitute question gets asked — from a comparative, feature-based substitute to an identity-based one — and a brand's entire positioning strategy can be understood as choosing which easy question it wants to be judged by, since customers were never going to answer the hard original question either way.
Sharpest implication: almost no consumer decision in a category the buyer doesn't personally have expertise in is actually being made on the terms it appears to be made on — the real decision is being outsourced, silently and automatically, to whichever proxy question the buyer's brain judges itself competent to answer, and understanding what that proxy question actually is matters more than optimizing the answer to the question you assumed was being asked.
Generative questions: