Money arrives from platforms and broadcasters with a title, sometimes a spelling of a name, and rarely an identifier. Turning that into a payment to the right person is a data problem, and the platform is built for that problem rather than around it.
The platform was built with companies that already work in the sector, and it exists to solve a specific set of problems rather than to be a general one.
A month of digital usage is millions of lines, and a work can appear in any of them under a slightly different title. The platform processes the whole file rather than a sample of it, because a sample is an estimate and an estimate is how money ends up unidentified.
Collecting is well understood; attributing is not. Every stage of the platform serves the moment a line of usage is matched to a registered work — code generation, duplicate detection, conflicting-claim resolution. Everything else is upstream or downstream of that.
CWR goes in, ISWC and IPI come out. Nothing is stored in a shape only Unison can read, which is what makes a catalogue portable — including away from Unison. A platform that locks a catalogue in is competing on the wrong thing.
Four stages. The second is the one the other three exist to serve, and it is the one the rest of the industry treats as administration.
A catalogue arrives as the industry files it, or as a spreadsheet. Usage arrives from the sources that measure it, in their own formats.
A line of usage carries a title and rarely an identifier. Attaching it to a registered work — the right one, with the right shares — is the whole job, and it is the step that decides whether money is paid or held as unidentified.
Digital is collected under direct agreements. Everything else runs through representation agreements with the entities that collect locally.
Each distribution is assembled from the lines that justify it, so a statement can be taken apart and put back together by whoever receives it.
None of these are Unison inventions. That is the point: a work identified with them is identified everywhere, by everyone, including in a system that has never heard of Unison.
A client portal and an internal back office, reading and writing the same data. The portal is not a report generated for clients afterwards — it is the same system, with the parts that belong to somebody else hidden.
Earnings, uses of works, repertoire and catalogues, cue sheets and recordings, statements, and collection mapped by territory.
Unison IDs and duplicate-conflict resolution, as their own part of the system rather than a step somebody remembers to do.
Usage reports, shares, bundles — and errors. A pipeline with nowhere to put a failure cannot tell you when one happened.
Claims, distributions and statements, so what was claimed and what was paid are the same object seen from two sides.
Rates, agreements and invoices for the licensing side of the business, in the same system as the works they license.
Business intelligence over the whole dataset, which is what makes a question like “which territory stopped paying” answerable rather than a hunch.
Formats, identifiers, how a conflicting claim gets resolved, what happens to a catalogue that arrives without codes. Those are the conversations this was built for.