The availability heuristic is the tendency to judge something as more frequent or more important the more easily an example of it comes to mind — regardless of how often it actually happens. In a business where nobody logs what actually happened each night, that's not a curiosity from a psychology textbook: it's how you quietly base hundreds of decisions a year on the wrong number.
You open Google in the morning and there's a new review: two stars, a long complaint about a cold starter and a slow waiter. Underneath that one review sit forty others, most of them four or five stars, most of them two sentences long and long forgotten. And yet it's that one review that stays with you all day, that you mention to another owner, that you reread again that evening. Not because it represents your business — because it's easy to remember.
Psychologists Amos Tversky and Daniel Kahneman named that pattern in 1973: the availability heuristic. People estimate how often something happens based on how easily examples of it come to mind, not on how often it actually occurred. It's one of the most thoroughly documented patterns in the research on how people judge under uncertainty — and running a business is nothing but making continuous judgements under uncertainty, all day, with no logbook ever kept to check them against.
A guest remembers one evening. An owner has to carry hundreds of evenings in their head at once — every review, every busy Saturday, every supplier, every employee — with no system that automatically keeps the running average. Lacking that system, your brain falls back on what it always does: using the most vivid, recent or emotional example as a stand-in for the real number. That example is rarely the average. It is almost by definition the exception — because exceptions are what stick.
This guide first explains the mechanism, with the most famous experiment in the research itself. Then seven places where it steers your business in concrete ways: reviews, ordering, no-shows, complaints, suppliers, marketing and risk. At the bottom sits a calculator that uses your own numbers to show how much weight one striking event gets in your head versus what your own data actually says. Everything runs in your browser: nothing is sent or stored.
Why easy-to-recall feels like common
Your brain uses a shortcut rather than 'count every case and divide by the total'. It asks itself, usually without you noticing: how quickly and how vividly does an example come to mind? If it comes fast and sharp, it feels frequent. If you have to work to think of one, it feels rare — even when the real numbers say the exact opposite. That shortcut works surprisingly well most of the time, because what happens often is usually also easier to recall. It fails exactly where something is rare but vivid, or common but dull enough to never stand out.
Tversky and Kahneman demonstrated this with an experiment that has since become one of psychology's classics. They asked people: in English, are there more words that start with the letter K, or more words with K as the third letter? Most people answered: more words starting with K. That's wrong, and wrong by a wide margin — English has roughly twice as many words with K in the third position. The reason for the mistake is exactly the mechanism above: a word starting with K (kitchen, king, keep) comes to mind effortlessly; a word with K in third position (ask, bike, lake) has to be actively searched for. The easier search was mistaken for the larger count.
Translate that to running a business and you get three amplifiers that make an event 'available', regardless of how often it truly occurs: how recent it was, how emotionally charged it was, and how often you've retold it to someone else — which drills it in even deeper. A review from this morning, a no-show that ruined your Friday night, a supplier you were still angry at on the phone: none of the three needs to be representative to steer your next decision, simply because it comes back to mind so easily.
The classic Tversky & Kahneman (1973) experiment: what people guessed versus what's actually more common in English.
Illustrative depiction of the reported pattern from the original research, not the exact raw figures. What matters is the shape: people guessed the opposite of reality, purely because one example was easier to summon than the other.
The 7 places your memory is steering your business
They run in the order they hit a business: first your reputation and your revenue, then your floor and your purchasing, and finally your marketing and your risk budget. The calculator at the end turns it into your own numbers.
1. One bad review feels like a trend, sitting on top of forty good ones
Your average sits at 4.6 stars across two hundred reviews. That figure barely moves when one new two-star review lands — arithmetically, the average shifts by a fraction of a fraction. And yet it's that one review you bring up at the briefing, that you type and delete a reply to twice, that has you wondering by evening whether quality is slipping. The forty happy guests from the same week wrote nothing, or something short and generic — and that vanishes from memory without a trace.
That's exactly the asymmetry the availability heuristic predicts: a long, detailed, emotionally charged complaint is easy to remember; a satisfied guest who said nothing leaves no trace at all. So you systematically remember an unbalanced sample of your own reputation — not because you're not paying attention, but because the brain simply holds on to what stands out.
The fix isn't 'ignore the review' — it still deserves a reply, and a good reply wins back more on average than the review itself cost you. The fix is basing your judgement of your reputation on the log, not on memory: the real average over the last hundred reviews, not the last one you read. A review reply belongs to answering; a reputation-management check belongs to reading the real number.
An illustrative comparison: how rare something actually is versus how much mental space it takes up — five recognisable moments from a hospitality week.
An illustrative reconstruction of the pattern, not measured data about your own business. The lesson is the shape of the gap: the smaller the real share, the bigger the mental space it usually takes up — the exact opposite of how a good decision should weigh it.
2. One slammed Saturday makes next Saturday feel obvious — until it isn't
Last Saturday the place was packed: a local event, good weather, a full calendar. This week you order to the same level and staff the same number of people, because 'that's how busy it was'. Two weeks later half of what you ordered is still sitting in the fridge and you have one cook too many on the floor — because that one packed Saturday was the exception, not the norm, and your memory kept it as if it were.
This is exactly why 'same weekday as last week' is the most commonly used and least reliable rule of thumb in hospitality: it literally weighs a single data point, and that data point happens to be the most recent one — so the most available one. A system that actually weighs every previous Saturday, giving less weight to a one-off and more to the pattern, produces a very different number than your memory does.
That's exactly why the covers forecast tool on this site never calculates from a single previous night: every logged evening gradually replaces a slice of the first guess, instead of the most recent number overwriting everything. Weather plays a role too, and not the way you remember it — see weather and revenue for why rain does something very different to a terrace business than to a cosy winter room.
3. One dramatic no-show feels like 'no-shows are getting worse'
A table of twelve, booked for a birthday, simply doesn't show — no call, no cancellation. That night the table costs you most of your revenue for the slot, and you remember it for weeks. The next morning it feels like 'no-shows are really getting worse', and you consider a stricter policy for everyone.
What your memory doesn't tell you here: how big and costly one missed table was says nothing about how often it happens. A rare but heavy event feels, through its sheer impact, more frequent than it is — the same reason a rare but spectacular plane crash makes people more afraid of flying than of driving, which is statistically far more dangerous.
Before adjusting the policy, look at the real percentage over the last six months, not the one evening you can still retell in detail. A deposit policy is often the right answer — but calibrated to the real percentage, not the panic of one birthday table. Reducing no-shows goes deeper into what actually works.
4. One guest walking out over a complaint convinces you service is slipping
A guest stands up, halfway through the main course, after a complaint that wasn't handled well. It's visible to the whole room, it feels personal, and it's the first thing you tell your partner that evening. The fifty guests who left happy that same night go untold — there's nothing to tell about an evening that simply went well.
One visible, emotional failure weighs heavier in memory than dozens of invisible successes, and that shifts how you view your entire team and your entire service after a single night — while the complaint rate over a longer stretch barely moved at all.
The question that makes the difference isn't 'did this feel bad' but 'how often does this actually happen, and how well did we recover'. Complaint handling is about that recovery — and about tracking complaints as a number, not a feeling, so one evening doesn't overwrite the whole picture.
5. One botched delivery outweighs fifty clean ones from the same supplier
Your regular fish supplier has delivered correctly, on time, in order for three years. On one Friday the order arrives two hours late and partly wrong, right in the middle of mise-en-place for a busy night. That one Friday is what you remember going into the next contract negotiation, not the hundred and fifty problem-free deliveries before it.
One incident that hit you at the worst possible moment — right in the rush, with real consequences for that night — weighs heavier in your head than its real share of the relationship. That's not a reason to ignore it, but it is a reason to keep it separate from the question 'is this supplier reliable overall'.
An ordering and delivery schedule that logs every delivery gives you the real percentage instead of the most recent incident. Pair it with a stock-control tool that tracks discrepancies per supplier, and the next negotiation is about numbers, not about the one Friday you can still feel.
6. One viral moment convinces you that's what marketing looks like now
One Instagram reel of your dessert hits a hundred thousand views and produces a busy week. The months before and after it, your ordinary, recurring posts produced a steady, smaller stream of guests — less spectacular, less worth retelling, and therefore easy to undervalue next to that one viral moment.
The risk is that your marketing budget and time end up fully aimed at chasing 'that again', while the average return of your ordinary, boring, consistent channels — the ones nobody brags about at a party — was already higher over a full year.
One outlier is not a strategy, it's a data point. Social media for restaurants is about what works consistently over months, not about chasing the next lucky hit.
7. The rare, vivid risk gets more budget than the boring, likely one
After hearing about a kitchen fire at a nearby colleague's place, you immediately consider an expensive extra fire-prevention investment. Meanwhile the risk that actually hits your business every year — a slippery floor, a lapsed fire-extinguisher check, a staff member without current food-safety training — is boring enough to never become a story, and therefore easy to underfund.
That's the same distortion behind fear of flying: the vivid, rare scenario feels more urgent than the routine, likely risk, precisely because one has a story and the other doesn't. That's why insurers and safety inspectors work from percentages and incident rates, not from the last anecdote they heard.
Weigh your own risk budget against what actually goes wrong often in a business like yours, not against the most recent story you heard. HACCP in practice and restaurant insurance are both built around what statistically happens often — not around the most alarming story of the month.
Calculate how much weight that one moment gets in your head
Enter how many observations you have (for example, reviews), your average excluding the most recent one, the value of that most recent one, and how much it's stuck with you.
The calculator shows what your own data says against what your memory says once you let that one event weigh disproportionately — and how many 'ordinary' observations that one moment is quietly standing in for.
Memory Weight Calculator
Your own numbers, not a general rule of thumb.
—
This model is an illustrative rule of thumb for how recall weighting works, not a measurement of your own brain. Everything runs in your browser; nothing is sent or stored.
Two things to keep in mind while reading. The 'how much it's stuck with you' slider isn't a scientific measurement — it's a way to make visible that the more emotional or recent something feels, the more weight you unconsciously give it, regardless of how often it actually happened. And the number of observations is deliberately kept low for a supplier or an employee and high for reviews: the smaller the group, the faster one event takes over the picture.
Repeat the calculator for each of the seven places above with your own numbers. Wherever the gap is large, you now know why your gut got it wrong — and which log to check from now on instead of your memory.
What to do this week, this month and this quarter
Fixing all seven places at once doesn't work for anyone. This order does, because each step makes the next one measurable.
This week — find your own log
- Fill in the calculator above for the event that's still sharpest in your mind, and note the gap it produces.
- Look up the real number for that same event: your star average, your no-show rate, your delivery history — not from memory, from the system.
- Write down which of the seven places above you decided by feel most often in the last month.
This month — turn one log into a habit
- Pick one place — reviews, forecasting, no-shows — and start checking the real number weekly instead of the most recent moment.
- Tie it to an existing tool: the covers forecast for ordering and staffing, the reputation check for reviews.
- Bring one decision you made by feel to the next briefing, and check it against the real number.
This quarter — build the log for the rest
- Set up a simple counter per supplier of deliveries versus incidents, so one bad Friday never again defines the whole picture.
- Repeat the same exercise for your risk budget: what statistically goes wrong often in a business like yours, versus the most recent frightening story.
- Put it next to your decision-fatigue approach: that article protects when you decide, this one protects what you decide on.
Your memory isn't a logbook, and that's fine — as long as you know it
None of the above means your gut is worthless. Years of experience build an instinct that's often right, and weighing that instinct against the numbers is exactly what a good owner does. The problem only starts when one vivid moment overwrites the whole instinct, instead of being one data point inside it.
The fix doesn't take discipline, it takes a habit: for every big decision, ask once 'is this what I feel, or what the log says' — and check the log when in doubt. That costs a minute and prevents decisions built on the exception instead of the norm.
Pair it with decision fatigue — whether you're sharp enough to spot the difference — and delegating as an owner — who else is deciding by feel besides you — and you have three pieces of the same problem: how to run a business without the exception constantly overwriting the norm.