What Metadata Actually Does

The Invisible Architecture

Somewhere right now, a song is being streamed. The listener found it through an algorithm, added it to a playlist, let it play through twice. The rights holder — the person who made that song — will receive nothing. Not because the platform is stealing. Not because the distributor failed. Because the song has no idea who owns it.

Metadata is the information attached to a recording that tells the music industry’s infrastructure what a piece of music is, who made it, who owns what share of it, and therefore where money should flow. It is not paperwork in the bureaucratic sense. It is the actual mechanism by which ownership is asserted and royalties are routed. When it’s wrong, or absent, the money doesn’t disappear — it sits in a collection society’s account, unclaimed, eventually redistributed to other rights holders who did register correctly. Your silence is someone else’s windfall.

Most independent musicians engage with metadata the way most people engage with terms and conditions — skimming it, clicking through, assuming it will sort itself out. It won’t. Understanding what each identifier actually does, and why each one exists, changes the relationship from passive to active. It makes the difference between being in the system and being invisible to it.

The Recording and the Song Are Different Things

Before getting into the identifiers themselves, one distinction matters above almost everything else: the recording and the underlying composition are legally separate entities. They generate separate royalties, flow through separate systems, and are identified by separate codes.

When a song is streamed, two types of royalties are typically generated. One goes to the owner of the master recording — usually the artist or their label. The other goes to the owner of the composition — the songwriter and their publisher. A solo artist who writes, records, and releases their own music may be entitled to both. But both claims need to be registered separately, with different identifiers, in different places. Miss one, and you’ve effectively waived that royalty stream, at least until you sort it out — and sometimes retroactively, the unclaimed money is already gone.

This distinction is not a technicality. It is the structural logic that the entire royalty system is built on, and every identifier in use exists either to describe a recording, describe a composition, or describe a product. Knowing which is which is the foundation of getting paid correctly.

ISRC: The Recording’s Fingerprint

The International Standard Recording Code is a twelve-character alphanumeric string assigned to a specific recording. Not a song — a recording. If you record a new version of a track, even with only minor changes, that’s a new recording and warrants a new ISRC. A live version is different from the studio version. A radio edit is different from the album cut. The acoustic demo is different from the final mix. Each distinct recording gets its own code.

ISRCs are embedded in audio files and submitted alongside recordings to streaming platforms, digital distributors, and performing rights organisations. When a platform reports plays to a collection society, it reports the ISRC. The society looks up that code, finds the registered owner of the master, and routes the mechanical or neighbouring rights royalties accordingly.

Most distributors will auto-generate an ISRC for you when you upload a release, which is convenient but has a subtle risk: if you later switch distributors and re-upload the same recording without preserving the original ISRC, you create a duplicate — the same recording now has two different codes in circulation. Platforms may recognise it as a new asset, play history fragments, and reporting becomes inconsistent. The correct approach is to keep a record of ISRCs the moment they’re assigned, and to carry them forward whenever a recording moves between systems.

In territories with strong neighbouring rights collection — Germany, the Netherlands, Scandinavia — ISRCs are also used to collect royalties when recordings are broadcast on radio or played in public spaces. For artists with any international presence, an unregistered or incorrect ISRC is leaving broadcast royalties uncollected.

ISWC: The Composition’s Identity

The International Standard Musical Work Code does for compositions what the ISRC does for recordings. It’s assigned to the underlying musical work — the melody, harmony, and lyrics — not to any particular version of it. One ISWC can apply to hundreds of different recordings of the same song, from the original demo to every cover version ever made.

ISWCs are assigned through performing rights organisations — ASCAP, BMI, SESAC in the United States; PRS for Music in the UK; APRA AMCOS in Australia; and their equivalents everywhere else. To get one, you register the composition with your PRO. When you do that, you also declare the splits: who wrote what percentage of the song, and who holds the publishing rights to each writer’s share.

This is where many independent artists quietly lose money. A PRO can only pay what’s been registered. If you co-wrote a track with another artist and neither of you registered the split, the PRO holds the full royalty in suspense until the composition is claimed — and if both writers are signed to different PROs in different territories, the international flows get complicated quickly. Getting ISWC registration right, with accurate splits, is not about bureaucratic completeness. It’s about the fundamental act of asserting that you own something.

UPC: The Product in the Market

The Universal Product Code — or EAN in most non-North-American markets — identifies the release as a commercial product. Where an ISRC identifies a recording, a UPC identifies the bundle: the album, the EP, the single. It’s the barcode system that retail and distribution uses to track product movement, and in music it functions as the top-level identifier for a release in digital storefronts.

UPCs are typically assigned by distributors, and for most independent releases this is handled automatically. But the same logic applies: if you’re moving a catalogue between distributors, pulling UPCs with you (or at least documenting them) prevents fragmentation in streaming platform data. Some platforms use UPC as a linking mechanism between different format versions of the same release, so inconsistency causes operational problems downstream.

UPCs matter more visibly in sync licensing and physical distribution, where a product literally needs to be scannable in a system. But in the streaming economy, their role is structural rather than front-facing — think of them as the release’s address in the supply chain rather than its public identity.

Publishing Splits and the Split Sheet

None of the above identifiers solve the problem of who owns what percentage of a composition. That’s determined by the split agreement between collaborators, and it’s enforced by whatever documentation exists at the time of a dispute — or doesn’t exist.

A split sheet is a simple document that records the agreed ownership breakdown of a song: who wrote it, what each person’s percentage is, and what PRO each writer belongs to. It has no formal required structure. It simply needs to exist, be signed by everyone involved, and match what gets registered with each writer’s PRO.

The absence of a split sheet is one of the most common causes of royalty disputes in independent music. Two people write a song in a session, assume they’ll sort the paperwork later, and then one of them registers the track with their PRO as a solo write — intentionally or because they forgot there was a co-writer. The co-writer’s share sits unclaimed or is absorbed. By the time anyone notices, the conversation has become legally fraught.

The professional standard is to establish splits before a track is released — ideally before it’s even mixed. This is not because music professionals are paranoid, but because it’s the only moment when everyone is aligned and the conversation is easy. After release, after the track does well, every conversation about ownership becomes harder.

What Happens When It Goes Wrong

Metadata errors compound over time. A missing ISRC means broadcast royalties go unmatched. An unregistered composition means performance royalties accumulate in suspense. A split registered incorrectly means one writer is overpaid and another is underpaid — and untangling that retrospectively requires the cooperation of everyone involved, including potentially their labels and publishers.

Streaming platforms also use metadata to power discovery. Incomplete or incorrect genre, mood, language, or contributor credits affect how an algorithm categorises a track and therefore who it surfaces it to. This is a secondary effect, but it’s real — the same data layer that routes royalties also signals to recommendation systems how to place a piece of music in the ecosystem.

Collection societies run periodic data-matching exercises to clear suspense accounts and pay out unclaimed royalties. The window for claiming varies by territory. In some cases it’s years; in others, unclaimed money is redistributed after a defined period. Artists who register late don’t always recover everything they were owed.

The Practical Infrastructure

For an independent artist, the minimum viable metadata practice looks like this: every recording gets an ISRC, tracked in a master document alongside the release it belongs to. Every composition is registered with a PRO, with the ISWC noted once assigned. Every co-write has a signed split sheet before release. Every release has a UPC documented in the same master record.

This is less work than it sounds, and it compounds in your favour. A catalogue of fifty tracks with clean, consistent metadata is an asset. A catalogue of fifty tracks with fragmented, inconsistent, partially registered metadata is a liability — and the problem gets harder to fix with each passing release cycle.

The music industry’s payment infrastructure was not designed with independent artists in mind. It was designed around labels and publishers who had dedicated teams to handle registration, matching, and collection. The identifiers exist, the systems exist, and the royalties flow through them constantly — the question is simply whether your work is visible to those systems, registered correctly, and positioned to receive what it’s owed.

Metadata is not the least interesting part of making music. It is the part that determines whether making music is financially sustainable. Every unclaimed royalty is a small argument that you didn’t quite believe in the work enough to follow through on it — which is almost certainly not true. The paperwork just got in the way.

The Resonillator Metadata Organiser — available at resonillator.com/modules/metadata-organiser.html — is a free browser-based tool for building and maintaining a clean metadata record across your catalogue.