ISWCs & ISRCs- Making (or Breaking) Music Publishing
- Henry Marsden
- Sep 30, 2025
- 6 min read
The music industry runs on identifiers. Without them, the complex ecosystem of recordings, compositions, rights, and royalties would quickly collapse into chaos (... more so than it already is). The two most critical identifiers are ISRC and ISWC- the codes that uniquely define distinct audio recordings and song compositions respectively. Their existence looks simple on the surface, but they represent both different philosophies and underpin key workflows and challenges in attributing revenue accurately to rights holders.
While ISRCs have become near-ubiquitous and relatively robust, in my experience ISWCs have remained stubbornly patchy, slow to issue, and unreliable in practice.
Is one system “better” than the other- or are they simply two sides of the same coin? What does the future hold as industry bodies, publishers, and DSPs work to connect the dots between works and recordings with a goal of eradicating unattributed digital revenue?
What They Are and How They Work
The International Standard Recording Code (ISRC) is the identifier for sound recordings and music videos. Every time a track is recorded and made commercially available it will receive an ISRC. The format looks like this: CC-XXX-YY-NNNNN.
CC = country code of the registrant
XXX = registrant code (three characters assigned to a label or distributor)
YY = the last two digits of the year of reference
NNNNN = a five-digit number unique to the track
ISRCs can be generated relatively easily. National agencies (such as PPL in the UK or RIAA in the US) allocate registrant codes, but once an entity has that three-letter code**, they can create ISRCs themselves for each new recording. Aggregators and distributors handle this automatically on behalf of artists, and DSPs will not accept recordings without an ISRC. In practice, this has made ISRCs universal across digital catalogs.
The International Standard Musical Work Code (ISWC) is the parallel identifier for compositions. A typical ISWC looks like: T-345246800-1 (the final digit a check to ensure validity of the other characters). Unlike ISRCs, ISWCs are centrally governed by CISAC, the international body for collective management organisations (CMOs). The historic rules for their allocation are more stringent:
An ISWC can only be issued once a work has been formally registered with a CMO or publisher.
The registration must include the work’s title, the contributors, their roles (composer, lyricist, etc.), and their IPI numbers.
Only one ISWC should exist per unique work, regardless of how many recordings are made of it.
This centralised governance theoretically makes ISWCs more reliable and authoritative than ISRCs. But in reality, the opposite sadly has become the norm.
Opposing Philosophies: Flexibility vs Rigidity
The ISRC system is intentionally decentralised and flexible. Once a registrant has their three-letter prefix, they can issue as many codes as they like. That’s why virtually every digital recording you encounter has an ISRC. DSPs demand them, distributors automate them, and artists can even generate them themselves if they have access to the right tools.
The ISWC system is the opposite: centralised and rigid. Codes are only issued through CISAC’s framework, with each national society controlling registrations. The theory is that tighter control ensures accuracy and reduces duplication.
In practice, though, the rigidity has created bottlenecks. Works often circulate in the industry without ISWCs attached. From my own experience reviewing large-scale catalog deliveries, I’ve seen up to 50% of works missing ISWCs entirely. When industry stakeholders don’t trust or use them consistently, their utility collapses- exactly like the required trust for currencies to function.
By contrast, recordings without ISRCs are rare- in fact near non-existent in the digital diasphora. Because DSPs enforce their use, and distributors embed them in their workflows, ISRCs have become mandatory. Even though errors do occur they remain the default, globally recognised identifier for recordings.
The Problems With Each System
Neither code is perfect.
For ISRCs, the biggest historical issue has been duplication. There have been cases of bad actors recycling codes- some prominent DIY distributors being known to have recycled codes in the past. In those instances, two different recordings ended up with the same ISRC, creating chaos in royalty reporting and attribution. But today, duplication is relatively rare. The checks at DSP ingestion points and within distributor systems have made the ISRC landscape much cleaner.
For ISWCs, the problems are broader. Sometimes, the same work is assigned multiple differing ISWCs, typically because it has been registered at different times by different societies. Elsewhere, ISWCs are applied liberally to derivatives or adaptations of the same work (typically because they have the same or a similar set of writers), blurring the boundary between what should and shouldn’t qualify as “unique.” The result is fragmentation. A single work might have two or three ISWCs floating around in different databases, while another equally important composition might have none.
The industry consequence is profound. Without consistent ISWCs, catalogs cannot be reliably reconciled. A song might exist in five different datasets under five slightly different titles, credited to contributors with differently formatted names, and attached to two or more ISWCs- or none at all.
New Developments in ISWC
Recognising these problems, CISAC and its partners have introduced new tools and frameworks to make ISWCs more timely and reliable.
Provisional ISWC allocation: Record labels can now provisionally allocate ISWCs at the point of release, even before a work has been fully registered with a CMO. This is a significant step toward embedding ISWCs earlier in the value chain, ensuring that works aren’t left floating without identifiers.
ISWC Resolution and ISWC Allocation Service: Through CISAC’s publisher services, publishers can now query and allocate ISWCs directly, bypassing traditional CMOs. This is designed to reduce duplication, speed up allocation, and give publishers more direct access to the system that governs their livelihoods.
Preferred ISWC: In addition to “provisional” codes, CISAC has introduced the concept of a preferred ISWC. This allows societies and publishers to understand a single authoritative code where duplicates exist. The aim is to reduce confusion and duplication by providing a canonical reference point for each work, even if multiple ISWCs have been issued in the past.
These changes are promising, but are still in rollout so not yet universally adopted. The reality on the ground remains that ISWCs are patchy across most catalogs, with trust and usage subsequently inconsistent.
Connecting the Dots
The holy grail of publishing data is early and reliable recording–work connectivity. Every recording (ISRC) should be linked to the composition it embodies (ISWC). If this were achieved, the post-release matching effort currently replicated across CMOs, publishers and administrators could be eliminated. Royalties could flow more accurately and more quickly, with fewer disputes and less leakage.
But there are hurdles. Who is responsible for establishing the link? Should DSPs require ISWCs alongside ISRCs at the point of upload? If so, are labels equipped (or incentivised) to provide them? What happens when a work doesn’t yet have an ISWC, or when multiple ISWCs exist for the same song?
The promise is clear, with the new CISAC services going a long way to address the issues- pursuant to adoption. With ISRC–ISWC links embedded in DSP metadata, publishers and creators could finally begin to trust that usage is being reported against the right works. But until ISWCs are consistently applied and universally trusted, the dream of seamless connectivity and attribution remains stubbornly out of reach.
A Case in Point: The Challenge of Agreement
I once worked on a catalog project with a seemingly straightforward goal: ensure every song was registered 100% accurately, with recordings matched, everywhere. We received datasets from six different CMOs, from the publisher, and from the creator themselves alongside other publicly available data to compare these against. I’d assumed the first task would be to match reported recordings against the works- but it was simply to even agree on what the list of works actually was!
Titles were inconsistent. Contributors were listed differently and IPIs were either missing or didn’t line up. ISWCs were often missing, duplicated, or contradictory.
Working closely with publisher, the creator’s PRO and the CISAC ISWC resolution service, it took over 12 months to reconcile the catalog into a unique list of just over 1,000 songs. The outcome was a clean, agreed-upon catalog, with each work uniquely described by its title, contributors, IPI numbers, and a single, preferred, ISWC. But it was a massive undertaking- work that then needed to be disseminated again to all parties/societies, before we could even get to the task of matching recordings to these works.
Building Trust
ISRCs and ISWCs both matter. ISRCs have succeeded because they are easy to generate, universally required, and embedded in the digital distribution ecosystem. ISWCs have struggled because they are centrally governed, slower to assign, and inconsistently used- resulting in the opposite effect that their governing rules were intended to prevent.
The good news is that reforms are underway. Provisional allocation and resolution services are promising steps forward. But trust will only come when ISWCs are universally applied, universally agreed upon, and reliably connected to ISRCs.
Ultimately, these identifiers are not just codes. They are the foundations on which royalties flow, rights holders get paid, and the creators themselves being able to sustain themselves.
Without clean identifiers, the entire edifice of music publishing wobbles. With them, we have the chance to build a system where efficiency replaces duplication, and where datasets serve the purpose they were designed for: ensuring that the right people are economically paid for use of their music.
Comments