What a purchase actually buys

A hundred and twenty places in this record — fifty-eight per cent — expect money across a counter before a stranger may sit.

Of those, ninety-two set no limit of any kind on how long the sitting lasts.

The figures

Seventy-seven per cent of the places that take your money then let you stay as long as you like.

Not one of the ninety-two says so.

The silence is the subject

This is the hinge of the whole record, so it is worth stating slowly.

A person buys a coffee. Nothing in the transaction states how long the table is theirs. No sign, no menu line, nothing said at the counter. In seventy-seven per cent of cases the honest answer is as long as you want, and in no case at all is that answer given. When observations span many locations, Monitask project reports for remote staff can add time and project context without turning a single activity signal into a performance verdict.

So the person has to supply one. The third section of this site counts what they supply, and the answer is: an hour, usually, and wrongly. Readers comparing this walk with broader public-space evidence can also see Tripadvisor.

The commonest arrangement in the record is one where both parties know the money part and neither states the time part, and only one of them is guessing.

Against the places that ask nothing

ask for money (120):    92 no limit   (77%)
ask for nothing (71):   63 no limit   (89%)
a fee (15):             13 no limit   (87%)

Paying across a counter is the entry condition most likely to come with a limit, and it is still four cases in five.

The gap between seventy-seven and eighty-nine per cent is twelve points on groups of a hundred and twenty and seventy-one. It is in the expected direction and it is small, and the honest summary is that almost everywhere in this city lets you stay, whatever it asked for at the door.

What changes is the assumption, not the rule

Set the same three groups against what was assumed before checking:

ask for nothing:  22 of 63 with no limit were assumed to have one  (35%)
a fee:             6 of 13                                          (46%)
ask for money:    77 of 92                                          (84%)

Twelve points of difference in what places actually do. Forty-nine points of difference in what gets assumed about them.

That is the clearest result in the entry section and it is about people rather than places. The counter does almost nothing to the rules and nearly everything to the expectation.

Why a purchase would do that

Three readings, and the record cannot choose between them.

It feels like a contract. Having paid for a defined thing, a person reasons that they have bought a defined amount of something else as well, and time is the obvious candidate.

The cup is a clock. A coffee is finite and visibly so. When it is empty the arrangement feels expired, in a way that a library chair never does.

Everybody has seen somebody moved along. The expectation may be learned from rare events — four signs about turnover in the whole record, eleven tables cleared during the walk — and generalised far beyond them.

The second is the one the observer found most plausible while walking and the record contains no evidence for it at all. It is named here as a hunch rather than printed as a finding.

What would settle it

Asking people, which this walk did not do.

The assumption field in this record belongs to one observer. It is not a measure of what anybody else thinks, and every page that uses it says so. A version of this walk that asked three people in each place what they believed would turn the central finding from an anecdote with a number attached into something that could be defended.

That is the first of the six questions on the page about what a second walk would settle, and it is the one that matters most.

The free chair, which is where the counter really shows

Of the three fields crossed with entry, the seat field moves furthest.

ask nothing (71)       free chair in 67    94%
a fee (15)             free chair in 6     40%
ask for money (120)    free chair in 16    13%

Ninety-four per cent against thirteen. There is no comparison on this site with a wider gap.

That is nearly a definition rather than a finding — a business with tables sells the use of them — but it is worth setting beside the limit figures, because the two point opposite ways.

On time, the counter barely matters: seventy-seven per cent of paying places limit nobody, against eighty-nine per cent of free ones.

On a seat, the counter is everything: thirteen per cent against ninety-four.

So a transaction buys a chair and does not buy time, and the thing people assume it buys is time.

The sixteen that take money and seat you anyway

laundrettes and shops with chairs    5
transport concourses                 3
cafés                                3
hotel lobbies                        2
pubs                                 2
bookshops                            1

The laundrettes are the only structural case: waiting is the product, and a chair for it is part of what the place is.

The concourses and lobbies have seating that serves something unrelated to selling — a train, a guest — with the business happening elsewhere in the building.

The three cafés and two pubs are front benches, two of them explicitly for collecting orders. A stricter reading would give those five kinds zero free chairs between them, and a reader who prefers that should know the figure is five.

What the hundred and twenty do not tell us

Whether the silence is deliberate. No place was asked why it says nothing about time, and several would presumably answer that the question has never come up.

Whether stating a limit would help or hurt. This record has no view. It counts what is said and what is assumed, and the page on the recommendation it cannot make explains why it stops there.

What happens when the place is full. Almost the whole walk was weekday daytime. The arrangement that holds at eleven on a Tuesday is the one measured here, and it is the easiest version of the question.