There’s a quiet war fought at dinner tables everywhere. On the one side: people who love coriander, that citrusy herb that lifts a curry or brightens a salsa. On the other side: people who insist it tastes like soap. For years, haters were dismissed as fussy. Then science stepped in a genetic variant near a cluster of smell-receptor genes on chromosome 11 changes how strongly people detect the soapy-tasting compounds in the herb. Coriander haters were never being difficult. It’s biology.
Your clients are the same. And it’s time insurance started cooking, and talking, accordingly.
The one-size-fits-all recipe is broken
Most insurance conversations serve the same base ingredients to everyone: the same coverage tiers, the same jargon, the same “protect what matters most” script. It’s a recipe built for an average palate that almost nobody actually has.
A 32-year-old sole designer running a design studio from her spare room doesn’t experience risk the way a 58-year-old manufacturing director does. A family that just weathered a flood doesn’t hear the word “excess” the way a family that’s never claimed does. Yet both are often handed the same explanation and the same sense of what “enough cover” means.
This isn’t a product failure; most insurers have plenty of variety in the pantry. The failure is in the serving. Two clients can hold identical policies and walk away with completely different relationships to their own protection: one reassured, one resentful; one who renews without a second thought, one who feels quietly overcharged for something they never understood.
Risk perception is genetic, too – just not literally
The coriander gene is a neat fact, but the real lesson is the category it represents: an invisible, involuntary difference in perception that produces wildly different reactions to the same stimulus. Clients carry their own version of a “risk chromosome” shaped by upbringing, past claims, industry exposure, media exposure, and temperament.
Some clients see danger everywhere and want to over-insure against ghosts. Others are constitutionally optimistic and miss exposures a broker can see clearly. Neither is wrong, any more than either coriander camp is wrong about what they taste. But don’t trust clients to tell you which camp they’re in. Ask someone whether they want “just the headline cover” or the full detail, and most will pick the option that requires less effort, not the one that’s actually true.
People are unreliable narrators of their own risk psychology. The fact-find is a starting hypothesis, not a diagnosis. The better test is behavioural: which framing produces fewer panicked calls after a near-miss, fewer lapsed renewals, fewer complaints when a claim doesn’t go as imagined. Watch what clients do, not just what they say they want.
The product isn’t the policy
Here’s the reframe that changes everything: clients rarely buy the thing on the schedule of cover. They buy the feeling of not having to think about it anymore.
The “irrational” premium people pay for reassurance, or a good story, isn’t irrational; it’s solving a different problem than the spreadsheet solves. A client doesn’t lie awake calculating expected claim value. They lie awake imagining the phone call where something’s gone wrong, wondering if they’ll be looked after. The policy is the mechanism. Peace of mind is the purchase.
That means two policies with identical technical value can deliver very different felt value per rand of premium. A R5 million liability limit means little to someone who’s never imagined a R5 million claim; a same-day claims hotline with a named contact might mean everything to someone who’s pictured being left on hold while their business burns. Brokers who optimise purely for technical adequacy can still lose the client’s trust if they never address the actual anxiety being purchased against.
Three ways to taste-test your advice
Diagnose before you serve — then watch; don’t just ask. A chef checks for allergies before plating. Yet how many fact-finds actually probe how a client processes risk, rather than just what risks they face? Questions like “Have you ever had a claim go wrong?” or “What keeps you up at night about the business?” reveal a client’s palate. But treat the answers as a working theory, refined against what actually happens at renewal and at claim — who calls anxiously, who never opens the policy document, who only complains when the story doesn’t match the outcome. That’s the real fact-find, and it never stops.
Reason with story, not just structure. The dish itself — “the policy” — often doesn’t need to change. The seasoning does: the language, the emphasis, the narrative. Underwriting language is built for probabilities and exclusions, but almost nobody processes risk through statistics. They process it through story. A cautious client needs the exact story of what happens on the day something goes wrong, not a percentage. A confident, growth-focused client needs the story reframed around upside and continuity, because leading with worst-case will switch them off. Same policy, same numbers, different story and the story does more persuasive work than the number ever will.
Mind the defaults. They decide more than the pitch does. Behavioural economics has one unglamorous, enormous finding: people overwhelmingly stick with whatever they were first given. The default level of cover, the default excess, the add-on presented as “recommended” — these matter more than anything said afterwards to justify them. If a client’s first exposure is a bare, price-optimised quote, that anchor is hard to shift later, however good the advice that follows. Brokers who want personalisation to stick should spend as much care on the first framing a client sees as on the ongoing relationship. Inertia, not persuasion, usually decides where a client ends up.
A fourth point worth adding: Expect complaints from people you fed correctly. Some clients will still say the policy “tastes like soap” even when it’s properly priced and explained. That’s not necessarily a sign you got it wrong — some people are never going to enjoy complexity, however well it’s prepared. The job isn’t converting everyone into a lover of detail; it’s recognising which dish someone will never enjoy, and finding another way to deliver the same nutrition — a simpler summary, a shorter call, a named contact instead of a policy pack.
The techno-feudal turn
Insurance is drifting toward what some economists call techno-feudalism: platforms controlling the infrastructure everyone else must rent access to, extracting value through control of data and defaults rather than open competition. Comparison sites, embedded insurance at checkout, and algorithmic underwriting increasingly decide what a client is offered before a human ever enters the conversation.
The pitch is personalisation. The reality is often standardisation wearing a personal-sounding wrapper, a telematics score or postal code standing in for a person, sorting the client into a segment rather than actually knowing them. These platforms don’t win because their personalisation is more sophisticated than a good broker’s. They win because their defaults are stickier. The pre-ticked box and the auto-renewal do more work shaping outcomes than any algorithm, because most people never scroll past the default to see what else was on offer.
Some platforms go further, using discounts and “rewards” to shape behaviour so their risk pool performs better. That’s a legitimate tool in the right hands, but it’s also a quiet transfer of authorship, where a client’s habits start being written by the platform’s incentives rather than their own judgement. The broker’s counterweight isn’t nostalgia for paper forms. It’s insisting that data inform advice rather than replace it, and that any default a client is placed into serves the client’s goals, not the platform’s loss ratios.
Why this matters now more than ever
Regulatory focus on fair value and treating customers fairly has made this more than a nice-to-have. The expectation that customers understand what they’re buying, and that it delivers value for them specifically, is essentially a regulatory codification of the Coriander Principle. A product can be correctly priced and still fail a fair value test if the client never understood it, never trusted it, or was steered there by a default they never consciously chose.
Personalisation here isn’t about bespoke policy wording for every client. Most books of business don’t have the margin for that. It’s about personalising the conversation: the diagnosis, the story, the pacing, the defaults, the channel. That costs nothing but attention, and it’s the single biggest lever brokers have over retention, complaints and trust.
Chromosome 11 doesn’t care whether coriander is objectively good or bad. It simply determines how a given person experiences it, permanently and involuntarily. Insurance works the same way: the policy is fixed, but the client’s experience of it isn’t, and pretending otherwise is how good cover ends up feeling like a bad meal.
The brokers who thrive won’t be the ones with the widest product range or the most rigorous fact-find. They’ll be the ones who accept an uncomfortable truth: clients aren’t buying indemnity; they’re buying the removal of a specific anxiety. Read the table before you cook. Tell the right story, not just the right number. Watch the defaults. They decide more than the pitch does. And remember: the client who never calls you again didn’t necessarily get bad cover. They may simply never have believed, in their gut, that they’d been looked after.
Taste test before you serve. Your clients will thank you for it — even the ones who never liked coriander.