What Is Occupied vs Unoccupied Waiting Time?

An illustrative scene. A tyre-fitting bay on a Saturday morning: six chairs, a coffee machine, and a wall-mounted screen playing the news with the sound off. Nobody looks up at it. It’s there because a wait with something in it is supposed to feel shorter than a wait with nothing in it.

That idea has a name. Occupied vs unoccupied waiting time is the difference between a wait with something in it and a wait with nothing in it. The screen is an attempt at the first. Bare chairs and a clock are the second.

The distinction is more than forty years old, and it came out of a business school teaching note rather than an experiment. It also came with two conditions attached that almost nobody repeats.

What follows checks that rule against research published in the last three years. It also pulls apart two things the rule runs together. How fast time seems to pass is one of them. The number of minutes a customer would give you if you asked is the other, and the two don’t always move together.

Key Takeaways

• Occupied waiting time is a wait with something in it; unoccupied waiting time is a wait with nothing in it. The distinction comes from a 1984 Harvard Business School note by David Maister, offered as a proposition for research.
• In a 2025 Swiss laboratory study, 90 people waited seven and a half minutes with a phone, a sudoku, or nothing. What they did changed how fast time seemed to pass and how bored they felt (both p < .001). No two conditions differed significantly on the minutes estimated (Scientific Reports, 2025).
• Which occupation barely mattered. Phone and sudoku differed on neither felt speed (p = .87) nor boredom (p = 1.00) (Scientific Reports, 2025).
• Maister set two conditions on a filler that almost nobody quotes: it should offer a benefit in itself, and relate to the service coming next. Hold music was his counter-example.

What is occupied vs unoccupied waiting time?

Occupied waiting time is a wait with something in it. Unoccupied waiting time is a wait with nothing in it. A customer reading a menu in a queue is occupied. A customer watching the counter and nothing else is not.

The pairing comes from David Maister, who wrote it up as a background note for Harvard Business School in April 1984, item number 684064 (Harvard Business Publishing). The text appeared the following year as a chapter in The Service Encounter (D.C. Heath, Lexington Books, 1985). Maister set the same material out again on his own site in 2006.

Its status is worth stating, because most of what gets written about it skips this part. The note reports no experiment. Maister said he hoped “to identify testable propositions offering the opportunity for future research”.

He also told the reader how to take his headings. Each section title, he wrote, “should be considered a proposition concerning the psychology of waiting”. The first of those headings reads “Occupied Time Feels Shorter Than Unoccupied Time.”

So the rule quoted everywhere started life as an invitation to go and test something. It is the first of eight propositions in the note, and the only one this article is about.

What did Maister offer in support of it? A philosopher’s observation and a set of restaurant practices. He cites William James: “Boredom results from being attentive to the passage of time itself.” That is an argument about attention, and it is a reasonable one. It was never put forward as a measurement.

Does filling a wait actually make it feel shorter?

It changes how the wait feels. The number of minutes a customer would give you is a harder claim to support, and the evidence behind it is weaker.

The most direct recent test ran at the University of Fribourg in Switzerland and was published in Scientific Reports in March 2025. Ninety young adults, mostly students, waited for 7 minutes and 30 seconds. Nobody was told how long the wait would be: the cover story said nothing about time, and the on-screen clock was hidden. Thirty went into each of three conditions: free use of their own smartphone, a sudoku on paper, or nothing at all.

Two measures moved clearly. Time felt slower to the people doing nothing than to the sudoku group (p < .001) and the phone group (p = .03) (Scientific Reports, 2025). Boredom split the same way, with the do-nothing group significantly more bored than either of the others (p < .001 in both comparisons).

The estimates behaved differently. The condition effect on them was only marginal (p = .054), and “Post-hoc comparisons showed no significant differences between conditions” (Scientific Reports, 2025). That is a weak result rather than an absent one, and the paper doesn’t present it as absent.

The setting bounds all of it: a laboratory in Switzerland, young adults who knew they were being studied, seven and a half minutes. What waited at the end was a paid lab task, not a service. The paper puts the sample first among its own limits: 76 of the 90 were women (Scientific Reports, 2025).

One result from the same study cuts against the tidy explanation. The usual account says busy people notice time less. Fribourg asked participants how often they thought about time during the wait and found no significant effect of condition (p = .08, Scientific Reports, 2025).

Boredom is not a soft outcome either. What tips someone out of a queue altogether is measured in how long people wait before they give up.

Two 2025 studies found people underestimating short waits

The advice to fill a wait usually rests on an assumption. It says customers overestimate how long they’ve waited, and that giving them something to do pulls the estimate back towards the clock. Two studies published in 2025 found short waits running the other way. Both were controlled studies, so neither watched people who had simply turned up to wait.

In the Fribourg experiment the wait lasted 450 seconds. Across all three conditions, participants put it at 343.73 seconds on average. The paper says they “underestimated the duration of the waiting period” (Scientific Reports, 2025). That is roughly five and three-quarter minutes called from seven and a half.

The second study used a real lift, but it was still arranged. Researchers from a Japanese research institute and an electronics manufacturer ran it as an experiment in a four-storey building. Thirty participants did 48 trials each, 1,440 trials in total, pressing buttons, getting out at set floors and skipping lifts when told to.

The finding, in the paper’s words, was that “the perceived waiting times were significantly shorter than the actual waiting times” (PeerJ, 2025). Across the thirty participants, the two readings still tracked each other (r = 0.510, p = 0.004). The paper is not consistent about the setting. Its introduction calls it a commercial building; its methods place the lift in the authors’ own laboratory.

Both are short waits, and that limit does real work. There is no single multiplier for the gap between felt time and clock time, and the answer changes with the setting. Nothing here says a forty-minute wait behaves like a seven-minute one. Neither was a queue for a service, and neither wait happened outside an experiment.

The argument underneath is the part to carry forward. If a short wait is already being estimated below the clock, then filling it isn’t correcting an overestimate. Whatever occupying a wait does, that is not what it’s doing.

What kind of filling actually works?

Maister answered this himself, in the sentence nobody quotes. The activity used to fill time, he wrote, should “(a) offer benefit in and of itself, and (b) be related, in some way, to the following service encounter”. Two conditions, both about the customer rather than the clock.

A 2026 paper in Scientific Reports had 50 online ride-hailing drivers in China wait for a passenger in a simulation. Entertainment fillers and passenger progress information were both varied. Only the entertainment moved the estimate: 66.79 seconds with it against 82.78 seconds without (Scientific Reports, the ride-hailing study, 2026). Drivers in a simulation are not customers in a queue.

Filling that is part of the service

Maister’s own example is the menu handed to people still standing in line. His claim for it is double: “Apart from shortening the perception of time, this practice has the added benefit of shortening the service time”. Customers arrive at the table ready to order. It meets both of his conditions at once, being useful in itself and being the first step of the thing being waited for.

Filling that has nothing to do with the service

Hold music is his counter-example. Music played to callers waiting for an agent is, to many people, “an added annoyance rather than a benefit”. His explanation is the second condition failing. The activity, he wrote, “is totally unrelated to the service activity to come”.

Filling the customer brought with them

Most people waiting have a phone on them already. In the Fribourg study the phone and sudoku groups differed on neither how fast time felt (p = .87) nor boredom (p = 1.00) (Scientific Reports, 2025). Having something to do was what separated both of them from having nothing.

That group also produced an oddity. Five of its thirty participants used the phone for less than half the wait, and a sixth read a book instead. That condition was analysed at 24 rather than 30 (Scientific Reports, 2025).

Picking the right one for a particular room is the subject of what actually shortens a wait.

Does it change anything an operator can measure?

Anxiety, yes. Satisfaction moved too, though not for every filler. A 2026 quasi-experimental study in Clinical Ophthalmology ran three waiting-room set-ups at an eye clinic in Alberta, allocated by day of the week. The three were a room with no media, a room with instrumental music at 40 to 70 decibels, and a room showing slow ocean footage over music. Each condition got 125 patients, 375 in total, with no dropouts.

Anxiety was rated on a 0 to 10 scale. It averaged 5.69 with no media, 3.59 with music only and 3.74 with the multimedia screen, and the difference across the three was significant (P<0.001, Clinical Ophthalmology, 2026). Two points of anxiety is a substantial move in a room full of people about to have their eyes examined. Satisfaction moved only for the multimedia room, above both no media (P=0.023) and music alone (P=0.041).

Then comes the sentence that decides what all this is worth, and it belongs to the authors. It opens: “Although perceived wait duration was not measured”. The rest of that sentence is a hypothesis offered in a discussion section, not a result. The paper also records that “Individual waiting times were not recorded” (Clinical Ophthalmology, 2026).

So the best real-waiting-room evidence available shows filling a wait moving anxiety, and moving satisfaction in one of the two rooms. It doesn’t show the same thing moving the clock in anyone’s head, because that clock was never read.

Which is a measurement problem before it’s a psychology problem. Anxiety, satisfaction and estimated duration are three separate readings, and a room can move one of them while leaving the others where they were. A venue that redecorates and then surveys only satisfaction has measured the change it made, not the wait. Taking the readings apart is the practical half of measuring wait times in your own venue.

Where the “filled time feels longer” finding fits

There’s a laboratory result that appears to say the exact opposite of Maister’s proposition. It’s called the filled-duration illusion. Filling an interval makes it seem to stretch out, rated as lasting longer than an equally long stretch left empty. Both can’t be true of the same thing, which makes the question whether they are about the same thing.

Start with what “filled” means in that research. A 2026 review in Psychonomic Bulletin & Review puts it directly: “filling the gap between two tones with more tones creates an impression of temporal extent”. The review gives the other standard version too. “Periods filled by a single continuous sound are also said to be perceived as longer in duration than periods between two discrete sounds” (Psychonomic Bulletin & Review, 2026).

A filled interval there is a stretch of time carrying tones or a continuous sound, set against a stretch marked only by its two endpoints. What fills it is sound, not activity.

“Occupied” in Maister means something else. A person doing something while waiting for a service they came for. One word is doing two jobs, and the two findings are not in contact with each other.

That reading is this article’s own. No paper reconciling the two turned up in the research behind this piece. What is offered here is what each literature says about its own scope, set side by side.

A 2025 paper in Attention, Perception & Psychophysics adds a reason to keep them apart. How long something is judged to have lasted and how fast time seems to pass are separate judgements. In that study they moved together, but far from perfectly. Across brief tones and a blue square, the two correlated between .37 and .59 (Attention, Perception & Psychophysics, 2025). None of it was a wait. It shows only that the two questions are not the same question.

Why the menu works and the hold music doesn’t

The rule survives, but not for the reason it’s usually given. Filling a wait doesn’t reliably shrink the number a customer would give you in minutes; what it reliably changes is whether the time drags and whether they’re bored. Both of those are worth changing for their own sake, and the in-date evidence supports them far better than it supports the minutes.

Maister’s two conditions predict which fillers manage it. The menu offers a benefit in itself and belongs to the service coming next, so it passes both. Hold music offers neither, and he said so in the same note that produced the rule everyone quotes.

Boredom possibly pushing somebody out of a queue altogether is a claim tested in why customers abandon a queue.

So the test worth carrying isn’t whether the wait now has something in it. It’s whether the customer would have chosen that something if there had been no wait at all. A menu passes. A muted screen on the wall of a tyre bay does not.

Frequently Asked Questions

Who came up with occupied and unoccupied waiting time?

David Maister. The note was written for Harvard Business School in April 1984, item 684064 (Harvard Business Publishing), and published as a book chapter the following year. He set out the same eight propositions on his own site in 2006. His closing position is easy to miss. In the full text of the article he writes: “The main point of this chapter is that the waiting experience is context specific.”

Does giving customers a menu really make them wait more happily?

The menu is Maister’s own example, from the 1984 note, and his claim for it has two halves. One is about time perception. The other is about throughput: customers who have read the menu in the queue are ready to order when they sit down, which shortens the service itself. The second half doesn’t depend on anything happening inside a customer’s head, which makes it the sturdier of the two.

Is hold music better than silence?

Maister thought not, calling it “an added annoyance rather than a benefit” because it has nothing to do with the service the caller is waiting for. Music in a physical room is a different case. A 2026 Alberta eye clinic study played instrumental music at 40 to 70 decibels. Patient anxiety fell from 5.69 to 3.59 on a 0 to 10 scale (Clinical Ophthalmology, 2026). That study did not measure how long the wait felt.

Does the filled-duration illusion mean keeping people busy backfires?

Not on the evidence available. In that research, “filled” means an interval carrying tones or a continuous sound, measured against an interval marked only by its endpoints (Psychonomic Bulletin & Review, 2026). What fills it is sound, not activity. A person doing a crossword while waiting is not that. One word, two situations.

What’s the difference between how long a wait felt and how long people say it was?

They’re two separate judgements and they can move apart. “How fast did that go?” is one question, “how many minutes was that?” is another. A 2025 study had people judge both after hearing a tone or seeing a blue square lasting between 200 milliseconds and 5 seconds. Tones were judged to last longer, and time was judged to have passed more slowly during them (Attention, Perception & Psychophysics, 2025). The two only correlated between .37 and .59.

Sources

• Attention, Perception & Psychophysics, Bratzke, “Do “auditory” and “visual” time really feel the same? Effects of stimulus modality on duration and passage-of-time judgements”, published 25 July 2025, retrieved 2026-08-29, https://pmc.ncbi.nlm.nih.gov/articles/PMC12568886/
• Clinical Ophthalmology, Mohamed and colleagues, “Effects of Background Music and Multimedia Exposure on Patient-Reported Anxiety in Ophthalmology Waiting Rooms: A Quasi-Experimental Study”, published 12 May 2026, retrieved 2026-08-29, https://pmc.ncbi.nlm.nih.gov/articles/PMC13180329/
• David Maister, “The Psychology of Waiting Lines”, the full article text, carrying the 1985 D.C. Heath and Company copyright line, retrieved 2026-08-29, https://davidmaister.com/articles/the-psychology-of-waiting-lines/
• David Maister, “The Psychology of Waiting Lines”, the author’s own restatement of his eight propositions, posted 8 September 2006, retrieved 2026-08-29, https://davidmaister.com/blog/201/the-psychology-of-waiting-lines.html
• Harvard Business Publishing, “Psychology of Waiting Lines” by David H. Maister, the publisher’s record for Harvard Business School background note 684064, publication date April 02, 1984, retrieved 2026-08-29, https://store.hbr.org/product/psychology-of-waiting-lines/684064
• PeerJ, Shiomi and colleagues, “How long is too long? Examining waiting times and stress in human-elevator interaction”, published 18 August 2025, retrieved 2026-08-29, https://pmc.ncbi.nlm.nih.gov/articles/PMC12369606/
• Psychonomic Bulletin & Review, Yousif and Sherman, “Distortions of space and time in and around objects and events”, published 26 February 2026, retrieved 2026-08-29, https://pmc.ncbi.nlm.nih.gov/articles/PMC12945933/
• Scientific Reports, Liu and colleagues, “Optimizing waiting experience: how passenger progress information and entertainment fillers jointly influence online ride-hailing drivers’ time perception and emotion”, published 8 January 2026, retrieved 2026-08-29, https://pmc.ncbi.nlm.nih.gov/articles/PMC12868781/
• Scientific Reports, Meteier and colleagues, “Effect of task nature during short digital deprivation on time perception and psychophysiological state”, published 26 March 2025, retrieved 2026-08-29, https://pmc.ncbi.nlm.nih.gov/articles/PMC11947215/