How the walk was made

Two hundred and twelve places visited over eleven weeks in one city. Six excluded. Two hundred and six recorded, about a hundred and forty-four hours of sitting, median forty minutes a place.

The four fields

1. What the place asks at the door. One of four: nothing to pay and nothing to buy; one purchase expected; a purchase roughly every hour; a fee, ticket or membership.

2. How long I think I can stay. Written at the door, in a sentence, before looking for a notice and before asking anybody. A duration or unlimited. To compare observations across locations, reviewers can read the current feature explanation and define the decision, access rules and review date in advance.

3. How long I actually can. A look for a notice — on the door, the counter, the table, the wall beside the seating — and then, where nothing was found, one sentence to one member of staff. One of three: no limit; a written limit; a limit given on asking. A separate perspective on streets, seating and urban observation is available from Smart Cities Dive.

4. Did anybody speak to anybody who was not staff. Yes, no, or could not be seen. One exchange is enough; the field is not a count of conversations.

Plus the kind of place, and the length of the sit.

The rule that makes it work

Field two is written before field three, at the door, or it is not written at all.

A minute later it is unrecoverable: nobody can honestly reconstruct what they expected before they knew. That rule is the only reason this record has anything to say, and it was applied in every one of the two hundred and six places, including the ones where the guess turned out embarrassing.

One question, never two

The question was is there a limit on how long somebody can sit here, asked of whoever was most obviously working, and not pursued.

Pressed a second time, a member of staff stops reporting and starts deciding, and the answer drifts toward whatever gets the stranger to go away. The first answer is what the place has ready; that is what the record wanted.

A hundred and eighty-five places were asked. Seventeen gave a limit. In twenty-one there was nobody obvious to ask, and the look stood alone; all twenty-one are in the no-limit column.

The twelve kinds

cafés and coffee shops              58
pubs and bars                       31
shopping centres and arcades        16
places of worship                   14
bookshops                           13
libraries                           12
covered public seating              12
laundrettes and shops with chairs   12
community and church halls          11
hotel lobbies                       10
transport concourses                 9
museums and galleries                8

The list was settled after a pilot afternoon and frozen. There is no other column.

The kind was assigned by the walker rather than asked of anybody, which makes it the one field here that is a classification rather than an answer. Several of this record's sharper results rest on it, and where they do, the page says so.

The six that were excluded

A hospital concourse, two university buildings open to the public, a bank branch with seating, a car showroom, and a funeral director's waiting room.

One test removed all six: each is somewhere a person may be only if they have business there, which is the one thing a third place is not.

The test was written after the walk had begun. That is why the six are named by kind and counted here rather than quietly dropped: a list fixed in advance that then needs a rule to stay honest is a list that has failed somewhere, and the honest response is to show where.

What was deliberately not collected

Names. No place, street, district or city. A list of two hundred and six addresses visited once each for forty minutes would be read as a guide, and it would be a bad guide.

Prices. No amounts anywhere. The cost falls hardest on the ninety-six places that expect a purchase, because for them the price is the terms — and on any statement about who can afford to be anywhere, which is the largest thing this record does not measure.

Anything about people. Nobody is described, counted or aged. The record contains no information about any person in any of the two hundred and six places.

The three things this walk would do differently

Record the assumption more than once in the same place. On arrival, with a full cup, with an empty one. If the assumption shortens as the cup empties, the mechanism is the drink; if it is fixed from the moment of paying, the mechanism is the transaction. Three readings in forty minutes, costing nothing.

Code by room, not by building. Seven of the thirteen bookshops contain a café, and each was counted once. It is the only known coding error that a rule could have prevented, and it affects two fields.

Count the seats. The seat field is binary: is there somewhere to sit without buying. A library with forty chairs and a concourse with three are one value, and capacity is probably the most useful figure the walk did not collect.

What is published alongside the counts

Every residual at its own size: eighteen conversations that could not be seen, twenty-one places with nobody to ask, seventy-eight with no assumed duration, six places excluded. None of them folded into a neighbouring column.

The figures that work against the record's own subject: a hundred and sixty-eight places that limit nobody, on a site about places being read as restricted; and a hundred and forty-one where nobody spoke, on a site about where strangers meet.

And which figures rest on a guess. The limit field counts what places say. The assumption field is one person's. Every page using the second says so.