Two tables both wait exactly ten minutes for their seats. One party stands there relaxed, gets handed an aperitif menu, and tells you on the way out that the place "has a great vibe". The other stands in silence, checks a watch every thirty seconds, and leaves a review about "waiting forever". The clock says ten minutes both times. The experience says something else entirely.
That gap has been studied for forty years, and the answer isn't in a hospitality handbook — it's in the operations-research literature. In 1985, David Maister, then a professor at Harvard Business School, published a short essay laying out eight propositions about how people experience waiting. None of the eight is about the length of the wait itself. All eight are about the conditions around it: whether you have something to do, whether you know how long it will take, whether it feels fair, whether you're standing there alone.
Maister's essay is now one of the most cited pieces in the service-operations literature: a 2025 retrospective counts nearly 1,500 works citing it and 201 peer-reviewed studies that have tested, extended or challenged his eight propositions. Richard Larson, an MIT operations-research professor, added an influential companion piece in 1987 on fairness in queues. And in 1991, Katz, Larson and Larson showed in a bank branch that customers who had something to watch or do while waiting were noticeably more satisfied — with the exact same wait on the clock.
This guide walks through seven of those principles — translated into what literally happens at your front door — with a concrete fix for each one you can act on tonight. Below, you'll put your own queue through a calculator: how many minutes your guests actually wait, and which of the six levers you personally control is costing you the most right now.
Everything runs in your own browser. Nothing is sent or stored — this isn't a measurement of your specific guests, it's a model built from the direction the research above points, so you know tonight where to start first.
Why a wait costs revenue, not just a bad review
The easiest thing to overlook about a queue is that it costs something. A guest standing in it isn't seated yet, isn't ordering anything, and the impression they form right there decides whether they come back. That's not a guess: De Vries, Roy and De Koster followed 94,404 real visits to one restaurant in the 2018 Journal of Operations Management and found that a longer wait predicts three things — a higher chance a guest leaves before being seated, a longer gap before they return, and a shorter meal once they do sit down.
Their simulation on that same data showed the restaurant's revenue would rise by almost 15% if there were no wait at all — a theoretical ceiling, since a restaurant with zero wait doesn't exist. The more useful number is the recoverable slice: 7.5 to 7.7% of that revenue was already reachable through better table allocation or a bit more seating capacity, without turning away a single guest or adding one table. That's not an investment, that's planning.
This connects directly to what you already know about overbooking and no-shows: some of the guests who "don't show up" did in fact stand at your door and left again because the wait felt like it would never end. That departure never shows up in your no-show rate — but it shows up in that night's revenue.
The ultimate guide Managing Restaurant Reservations: 6 Steps to a Full Room From waitlist to no-shows: everything your reservations can do for you, in one guide. Open the guideThe 7 principles — and what to do about each one tonight
Six of the seven are entirely in your hands — those six also make up the calculator further down. The seventh is different: it explains why some restaurants get away with a longer wait than others, and why that's not a free pass.
1. Unoccupied time feels longer than occupied time
This is the first and most-cited of Maister's eight propositions, and the reason is simple: a brain given nothing to process falls back on its own internal clock, and that clock runs slower than a watch does. Two guests waiting exactly ten minutes — one with an aperitif menu and a glass of water in hand, one standing empty-handed at the door — walk away with two completely different memories of the same ten minutes.
Katz, Larson and Larson confirmed this in a 1991 bank-branch study: the moment there was something to watch or do during the wait, satisfaction rose noticeably at the exact same clock time. For a restaurant, the translation is concrete: a menu in hand, a water or aperitif service at the door, or simply a view of the kitchen instead of an empty hallway.
The rule of thumb: work out what it costs to put something in a waiting guest's hands from minute one — a glass of water is the cheapest version there is — and compare that with what a guest who leaves costs you. That arithmetic wins almost every time.
Two tables, exactly the same wait on the clock. The difference is what happens between minute zero and minute ten.
The unoccupied row has exactly as many minutes as the occupied one — just nothing to hang them on. That's precisely what Maister's first proposition and the Katz, Larson and Larson bank study measure: not the duration, but what fills it.
2. Waiting before the experience starts feels longer than waiting during it
Maister's second proposition: a wait that falls before service begins — before you have a table, a menu, or a greeting — weighs more heavily than an equally long wait once service has already started. Ten minutes at the door before anyone even looks at you feels different from ten minutes between the starters and the mains, even though both show exactly ten minutes on the clock.
The reason is that a guest who hasn't been greeted yet doesn't even know if they've been noticed. The moment someone says "I see you, you're next" — without a single step closer to an actual table — that same wait shifts psychologically from before the experience to inside it, and that shift alone shortens it.
The practical translation: the first twenty seconds at the door outweigh the next ten minutes. A host who makes eye contact and takes a name within those twenty seconds moves the wait from "outside, unnoticed" to "inside, in progress" — without a single table opening up any faster.
3. Uncertainty stretches a wait
Anxiety and uncertainty are two of Maister's eight propositions, and in a restaurant they largely overlap: will our table still be there, will we make the babysitter or the show, has the place simply forgotten us. Every second a guest spends on that question is a second they're not spending waiting — they're spending it worrying, and worrying feels slower than waiting does.
The most common, most avoidable moment of uncertainty is a host who gives an estimated time and then never comes back to update it. Fifteen minutes that quietly become twenty feels heavier than twenty minutes that were estimated correctly from the start.
What helps is strikingly small: a brief mid-wait update, even when the news is just "five more minutes" and tells them nothing new. The signal that someone is still thinking of you matters more than the content of the message itself.
4. A wait with no explanation or end point feels longer than an explained, bounded one
Maister's fourth and fifth propositions — uncertain waits are longer than known, finite ones, and unexplained waits are longer than explained ones — tend to overlap in practice, because an explanation almost always comes with a time estimate attached. Hui and Tse showed in a 1996 Journal of Marketing study that information about wait duration meaningfully reduces guests' tendency to overestimate that duration, especially for medium-length waits — precisely the range most restaurant waits fall into.
"We haven't forgotten you, a table's about to open up, about ten more minutes" contains two things that "we'll let you know" doesn't: a reason and a boundary. The reason removes the uncertainty about whether something went wrong; the boundary gives the guest their own clock to measure against, instead of the endless clock of "it could be anything".
A digital waitlist with an estimated time on a screen does exactly this, automatically, without a host having to say it out loud every time — which is also why a good waitlist system earns so much more than a name scrawled on a piece of paper.
5. An unfair wait feels longer than a fair one
Larson's 1987 paper added a dimension to Maister's list that's especially visible in hospitality: fairness. A guest who sees a later arrival, a regular, or a well-connected group get seated ahead of them doesn't just experience a longer wait — they experience an unjust one. And by Larson's research, an unjust wait weighs more heavily than an objectively longer, fair one.
This is the principle a restaurant cuts itself on fastest, because the exception feels entirely reasonable from the host's side — "they're regulars", "they'll only need two minutes" — while every other guest in line just sees it happen.
If you want to make exceptions, you can, but do it invisibly: hold a table back that never appears in the visible queue, rather than seating someone in front of a waiting party. The line your guests see has to be the line they actually experience.
6. Waiting with company feels shorter than waiting alone
Maister's eighth proposition is the simplest: a wait shared with someone feels shorter than the same wait spent alone. A conversation is itself a form of occupied time — the first principle above — but it adds something more: the sense that the wait is a shared experience, not something happening to you personally.
For a restaurant, this matters most for the solo guest and the small party: don't put them where they're most alone with their wait — put them somewhere with a view of the room, of other waiting guests, or of the bar, where a conversation starts easily.
A bar near the entrance happens to do this well already — not because it pulls in extra aperitif revenue, though it does that too, but because it turns a solo wait into a wait with company, even if that company is just the bartender.
Each bar below is one principle you control yourself. Fix them and ten minutes feels like ten. Leave them all unfixed and ten minutes feels like twenty.
This is the same arithmetic as the calculator below, with all six levers at their worst. In practice rarely more than two or three are active at once — which is exactly why the calculator shows your own combination instead of this extreme.
7. The more a guest wants a table, the longer they'll wait for it — and that's not a free pass
Maister's seventh proposition sits outside your direct control, but earns a place on this list: the more valuable the service, the longer a guest is willing to wait for it. A buzzy new opening or a terrace table on the only sunny evening of the month earns patience an ordinary Tuesday lunch never would.
That explains why two restaurants with identical waits get completely different reactions — it doesn't mean you can spend that reserve without limit. The same 94,404 visits from De Vries, Roy and De Koster show that even at a popular restaurant, a longer wait raises the chance a guest leaves; the patience is bigger, not infinite.
The practical lesson: use this principle to understand why a strong reputation buys you a longer wait, not as an excuse to skip the six principles above. They keep adding up, even at the most sought-after table in town.
Run your own queue through the calculator
Enter how long your guests actually wait and which of the six levers are working against you right now. The model is deliberately an illustration, not a measurement of your specific guests: it translates the direction of the research above into one comparable number, so you know where to start tonight.
By default the calculator opens on an ordinary Friday evening: fifteen minutes waiting at the door, no menu in hand, and no mid-wait update. Flip the six switches to match your own situation.
Perceived-wait calculator
Six switches, your own wait time, and what it actually feels like.
—
This model translates the direction of Maister's eight propositions, Larson's work on fairness in queues, and the Katz, Larson and Larson bank study into one illustrative number per lever. It isn't a lab measurement of your specific guests. Everything runs in your browser; nothing is sent or stored.
The perception gap isn't just a curiosity — it's the part a guest remembers and repeats, while the actual wait time is long forgotten. Two restaurants with the exact same wait on the clock can end up with completely different reputations because of it.
And it's also the part where you win fastest: all six levers cost little to nothing to close. A glass of water, a quick word at the door, a table that's never visibly pulled ahead of someone else — that's not an investment, that's attention you're already giving, just at the right moment.
What to do tonight, this week and this month
Fixing six levers at once in one service isn't realistic. This order is, because every step is free and sets up the next.
Tonight — the two free levers
- From the first waiting guest onward, put something in their hands: a menu, a glass of water, an aperitif suggestion.
- Make eye contact and take a name within the first twenty seconds — that moves the wait from "outside, unnoticed" to "inside, in progress".
- Pair every estimate with a reason: "a table's about to open up" outweighs a bare number.
- Keep any exceptions to the queue invisible to the rest of the waiting guests.
This week — measure your own queue
- Track the actual wait time per guest, from arrival to being seated, for one full service.
- Fill in the six switches in the calculator above for an ordinary night and for your busiest one.
- Pick the single heaviest lever and agree who owns fixing it starting tomorrow.
- Ask your floor staff whether they recognize the uncertainty and unfairness signals guests go through.
This month — build it into the system
- Make a mid-wait update a standard part of your waitlist process, not an exception when it's busy.
- Review your waitlist system: a screen with an estimated time handles the explaining and bounding automatically.
- Set your queue numbers next to your no-show rate — part of it is a departure during the wait, not a forgotten booking.
- Repeat the measurement after a month and compare the perception gap, not just the actual wait time.
The clock doesn't lie, but it doesn't tell the whole story either
Almost every restaurant that runs these numbers finds the same thing: the actual wait time is rarely the problem. Ten or fifteen minutes is perfectly normal for most guests. What makes a wait unbearable are the six things around it — standing empty-handed, no greeting, uncertainty, no explanation, a visible exception, standing alone — and every one of them costs almost nothing to fix.
That's the good news: this isn't an investment in more staff or a bigger dining room. It's a sequence. Start with the two that cost nothing tonight — something in hand, a name taken — and work through the rest this week with the calculator as your guide.
And remember the seventh principle when things go wrong anyway: a good reputation buys patience, but never unlimited patience. The 94,404 visits behind this research show that even the most popular restaurant loses a guest when the wait runs too long. The six levers above are what decide how much room you have before that happens.