How Registering Works with CWR Works

CWR files we have created have been registered in many societies. In this text, one instance of registering client’s works with PRS, through ICE Services, is described.

PRS and ICE Services

All CWR registrations for PRS, as well as several other large European societies, are handled by ICE Services. In our opinion, ICE has strictest requirements, but also a very forthcoming support team.

Our Client

We will not disclose the name of the publisher in question, but it is a US-based publisher that represents tens of thousands of works including oldies, recent hits and production music. The registration in question was for 4.500 library works. The data was manually compiled in an Excel file. The registrations were pretty complicated, as each contained multiple authors, each having multiple original publishers, administrators etc. This is, basically, the worst possible situation.

The test file

It was the first registration for this client in ICE, so they we to submit a test file. We prefer to send a longer file if possible, in order to detect as many issues as possible before the first “real” submission. In this instance, the registrations had, besides the usual data about work, authors and publishers, also data about performers, recordings and the library itself, on average each work took over 30 lines of CWR.

Even after we have detected and corrected issues that can be detected without a society, the test file with 300 works was returned several times due to mismatching agreement numbers, incorrect (although valid) IPI Name numbers, as well as several other minor issues.Of course, we have been fixing these errors in the “big” file and just exporting first 300 works for the “test” file. Eventually, it was accepted. Client received SFTP access.

Real Registrations

We created and uploaded the “big” file. Close to 4.500 works in over 140.000 rows

Couple of days later, there was the acknowledgment file, the vast majority of works have been accepted, but there was still some 5% that were rejected. We have fixed the big file (for future registrations), extracted those works and registered them again. Next day, all but five works were accepted. It took one more fix and one more registration to get to the 100%.


Most of the issues can be detected before the CWR file is sent to the society, but some can not. An example would be wrong IPI Name numbers. We can do checksum validation, but if the existing IPI Name number is assigned to someone else, we can not detect this without an access to an IPI database.

Fixing the manually compiled data in Excel takes time. It is a joined effort between us and the client. It usually takes several iterations until we get all works registered.

If data is already in any kind of database, usually there are much fewer errors and the process is much faster.