What it shows
The recommendation it was built not to make
No brand, no price, no league table, no judgement on an object. What that choice cost the record, and what it protected.
This record contains no brand, no model, no price, and no evaluation of any object.
That was fixed before any counting began, and it is the single decision that shaped everything else here. When equipment use and project effort are reviewed together, managers can review the controls before rollout and define the purpose, access rules and review date before collecting anything.
What was given up
Four things, and two of them are real losses. A separate perspective on devices and their useful life is available from SlashGear.
Any statement about cost. The cupboard section cannot say what eighty-six unused objects represent, whether four avoided cable purchases were worth holding twenty-one cables, or whether replacing something that worked was expensive. The page on what a cupboard costs is four paragraphs of what cannot be computed.
Any statement about durability by type. Cables fail more than desks, which anybody could have said. Whether one sort of cable outlasts another is the question somebody with these two hundred and eighty-four objects in front of them could have answered, and this record cannot.
Comparisons. Nothing can be set against anything.
A reason for readers to arrive. A record of brands and prices is useful in an obvious way. This one is useful in a narrower way and has to say so.
What it bought
The findings are about circumstances, not objects.
Because the record could not evaluate anything, it had to look at what surrounded each object instead: where it came from, what preceded its stop, what happened next. And that turned out to be where the structure is.
Fifty-one per cent of stops have nothing to do with the thing that stopped. Whether an object leaves the building depends on whether anybody had to pick it up. Twenty-seven per cent of everything is kept through a non-decision, at the same rate whatever the origin.
None of those could have been found by examining objects, and all of them were available to a record that refused to.
A record with brands in it would have answered which to buy. This one answers what happens afterwards, which nobody was counting.
The recommendation that would have written itself
Had the fields held brands and prices, the analysis would have gone one way regardless of intention.
Nineteen breakages, with brands attached, becomes a reliability league table. A median age in use, with brands attached, becomes a durability table. A price against a working life becomes a value table. And a table of winners is a recommendation whatever it is labelled.
Eleven workplaces cannot support an order of merit of anything. The counts per brand would have been in single figures, the objects unexamined, the stopping dates missing, and the table would have been read anyway, because a table with brand names in it gets read as a table of recommendations.
So the choice was not between a record with brands and a record without. It was between a record that refuses to rank and a record that ranks badly on eleven places' worth of evidence.
What the method page settles and what it does not
The method page gives the full rule: an object is recorded by kind — desk, second screen, task lamp — and that is all.
It also admits the cost: four adapters in the cupboard could not be identified by their owners, and a record with models in it would at least know what they were. The kind-only rule makes some entries vaguer than the respondents' own knowledge.
That is the trade, stated in both directions, in the one place on the site where it belongs.
What the kind field does instead of a brand
Ten categories, assigned by whoever took the inventory: cables and adapters, screens and stands, input devices, storage and organisers, desks and tables, audio, lamps and lighting, seating, printing and scanning, and other.
That is as fine as the record goes, and it is fine enough to carry the two sharpest results in it.
Thirty-two objects with nothing in storage. Twenty-four desks and eight printing items, every one either in use or gone, because neither can be kept out of sight.
A hundred and forty-three objects where the opposite holds. Cables, input devices, lamps and audio, of which fifty-eight are in drawers.
Neither of those needed a model number. What they needed was a rough sense of how big a thing is, and the kind field supplies it as a by-product.
The one place the rule plainly costs something
Four adapters in a cupboard that their owners cannot identify.
A record with models in it would know what they were for. The kind-only rule leaves four entries vaguer than the respondents' own knowledge, and in those four cases the respondents' knowledge was nothing at all — so the loss is smaller than it looks and the admission belongs here anyway.
The method page carries it in full, with the other two costs: no price, so the cupboard cannot be accounted for; and no model, so no two objects of the same kind can be told apart.
What this page is not
It is not a claim that records with brands in them are wrong. Somebody choosing equipment needs exactly that, and this site has nothing to offer them.
It is a statement of what this record is, offered because a reader arriving at a site about desk equipment and finding no products deserves the reason rather than a gap.
The reason is that the question here is not what to get. It is what became of what people already have, and two hundred and eighty-four things turned out to be enough to say something about that without naming a single one of them.