Exchanging Data among Publishers with CWR

CWR was not designed for sharing of publishing data between publishers. But this is a common practice, and quite often it is done in a wrong and harmful way. Will it happen to you? Or has it already?

In many texts on this web site, I rant against CWR as a ‘common’, ‘standard’ or even ‘format’. But in this text, I will try to stay on the constructive side. The part that is obvious is that CWR was conceived as a process of registering works with societies on massive scale. It is a process where the publisher sends the registration and the society (aka PRO) sends the acknowledgment(s).

The Bad Way

What has been happening for years now is that the same registration file that was sent to a society is being sent to sub-publishers. There is another way, which is terrible in a completely different way, it will be covered later.

The file being sent does not contain the sub-publisher. Or their sub-publishers. They are supposed to “add themselves”.

They import it into their systems, add themselves to the chains (or some equivalent thereof) and then new CWR is exported, with different work IDs, publisher IDs and writer IDs, which is sent to their local society. The society then sends the acknowledgment(s). This acknowledgment may be useful for the sub-publisher, but if there is any issue with any works within the registration, bad data, conflict, whatever, there is no good way to send it up the chain. And in many cases the cause of the issue is up the chain.

There is a way to convert acknowledgment files in a way that sub-publisher replaces there IDs with the ones from their client. While this will help in solving most of the issues, it can introduce new ones, particularly those that are caused by sub-publisher. However, this is actually the best (as in least bad) way to do it.

The horrible way

Some publishers prepare “world-wide” registration files with all their sub-publishers inside, or even, specific files for every sub-publisher. And they expect theirs sub-publishers to send these to societies.

Although this may seem like a good idea, it actually is far worse than the one with modifying acknowledgment files. Let us presume (naively) that such file will be accepted by all societies. A sub-publishers sends the files to the society, receives acknowledgments and sends them to their client. They do not really have to worry about issues reported in acknowledgment files, the client will inform them if any action is needed, in theory at least. Sup-publishers can still import CWR, or even better, acknowledgments into their system, but they can not really do much with it.

Let us presume the following scenario. Two writers of a work with unequal shares are represented by two original publishers and two administrators etc., but then, in some territory, the same sub-publisher represents both chains. The first registration will pass, but lets presume that the second is done with a file that was received from the client and passed to the society unmodified.

Compared to the first registration, it will have same title and writer names, but different work, writer and publisher ids. Different societies will treat these differently, but most interpretations will lead to serious problems:

  • Society will accept the registration, they will treat it as if it was sent by two different publishers. Quite often sub-publishers have issues with dealing with payments, but this is the only case where everything can work.
  • Society will accept the registration and postpone the issues until the first payment. Which will then be postponed until resolved. And it will never be resolved. Yes, the point is for them to sit on your money. Indefinitely.
  • Society will merge the two registrations. Payments will (most likely) be on time and in full, but sub-publishers rarely have the tools to separate these payments correctly.
  • Society will accept the registration and overwrite the first one, since in the second you are claiming that you only control one writer. The bottom line is that only a single writer’s percentage of the work will be payed for.
  • Society will reject the registration and require you to register it with the IDs from the first one. Good luck with that. This is actually the only right way to do it.
  • There might be more…

The reason why some majors have chosen this way to do things is related to the fact that in most of the cases explained, the first registration beats the second one, and this way is faster since the processing on sub-publishers software is not needed for registrations. Not to mention that when there are payments, due to the way some of the software out there works, they often get their share plus their share of the other shares. Or, rarely, everything.

For them the advantages of this system are obvious. In total, they collect much more than by using any other system, although many works will not get any money. If you are up the chain, ultimately author, or the sub-publisher who is not the exclusive sub-publisher, you may be on the loosing end.

The culprits

IDs, most importantly, work, publisher and writer IDs, are the core of all these issues. In the “bad way”, the ones assigned by sub-publisher is used, but it is hard to report things up the chain, as the IDs have been changed. In the “terrible” way, the ones assigned by someone up the chain, usually a major sub-publishing company, so from the point of sub-publisher none of these IDs may be unique (which is one of the most important features of any “ID”).


There are ISWCs and IPI Name/Base numbers that are truly unique. ISWC can, at least in theory, be acquired through CWR registration. IPI Name numbers are automatically assigned to society members. So these number are always available, or, at least, acquirable. By making them required, and a part of all CWRs, societies will use them, instead of “internal” IDs.

EBR (used by ASCAP, BMI and SOCAN) requires either internal ID or IPI Name for controlled writers and publishers. If both are present, IPI Name has precedence. While there are no such rules for ISWC and work ID, there is no reason not to use ISWC as internal work ID as it stands.

A next step would be to use IPI Name for reversibly creating writer and publisher IDs. IPI Name has 9 original and two check digits. There is absolutely no reason not to use the 9 original digits for IDs. And this is just one of the possibilities.

As in many issues in music publishing, tech is ahead of mindsets.