UKSpend (@ukspend)

In depth: UKSpend: Making London's Council Payment Data Searchable

Written from UKSpend's own profile data — claims, answers and testimonials — in a measured human voice via ScriptGrain · last written 2026-10-02

Every London council publishes what it pays out. That's the law, or close enough to it, and it means the information exists. Finding anything useful inside it is a different matter entirely.

Picture a journalist with a hunch about a supplier. She knows the company name, she suspects it has been paid by more than one borough, and she has an afternoon. What she faces is a separate file from each authority, each laid out a little differently, none of them built to be compared. The data is public in the strictest sense of the word. It is also, for practical purposes, buried.

UKSpend exists for that afternoon.

What It Actually Is

UKSpend is a free site that makes the published payment data of 33 London authorities searchable. The headline figure is £118.4bn, spread across more than 16.6 million transactions, with the most recent dated 31 March 2026. There is no sign-up, no account and no paywall. You arrive, you search, you get an answer.

It is part of Adapt Progress Evolve Limited, a UK software studio, and it does one job: it collects spending data and presents it so anyone can search it. That narrowness is the point. It isn't trying to be a campaigning platform or a think tank with a dashboard. It is a search tool for a large, awkward pile of records.

Where the Data Comes From

As for the source material, UKSpend builds on nothing else. Everything in it comes from the councils' own published payment data, released under the Open Government Licence v3. No leaked files, no scraped third-party databases, no estimates. If a figure appears on the site, it traces back to something an authority put into the public record itself.

That matters because it makes the whole exercise checkable. A reader who doubts a number can go back to the council's own publication and look. The licence, in turn, is what allows the data to be reused and republished at all, so the site's legal footing is as open as its inputs.

How You Use It

The two main routes in are vendor search and category search. Vendor search answers the question the journalist above was asking: who got paid, by whom, and how much? Category search works from the other direction, asking what a borough spends on a given type of thing, and lets you look at spending by purpose rather than by supplier.

Both routes sit on top of the same 16.6 million records, which is the real work of the project. Turning dense, inconsistent spreadsheets into something you can query in plain language is unglamorous. It's also the core value of the whole thing, because a dataset nobody can search is, basically, a dataset nobody reads.

For people who would rather not use a website at all, there is a public JSON API. That opens the data up to anyone who wants to pull it into their own analysis, their own tooling or their own reporting pipeline, without scraping the front end or starting from the raw files.

Who It Is For

The profile names its audience plainly: journalists, researchers, and anyone who wants to analyse UK government spending without wrestling with raw published data. The phrase to notice is "without wrestling". The users aren't assumed to be data engineers. They're assumed to have a question and limited patience for file formats.

That framing also explains the decisions around access. Free, no sign-up, openly licensed output. Each one removes a small barrier that would otherwise send a curious reader back to the spreadsheets, or away altogether.

What It Doesn't Cover Yet

Now for the caveat, and UKSpend states it openly. The Greater London Authority's own data isn't loaded yet. The 33 authorities in the headline figure are the ones currently in the dataset; the regional body that sits above them is not.

Anyone reading the £118.4bn figure should hold that in mind. It describes the coverage that exists, not the whole of London's public spending. The site's data-coverage page shows exactly which authorities are in and what date ranges they span, so the gap is something a user can inspect rather than something they have to take on trust.

It is a modest kind of honesty, and a useful one. A search tool that overstated its reach would send people to the wrong conclusions with great confidence. One that says where its edges are lets a reader decide how far to lean on it.

The journalist's afternoon, in the end, comes down to a single box and a company name.