The Copyright Office is soliciting public comments about how things are going with the MLC to help the Office decide whether to permit The MLC, Inc. to continue to operate the Collective (see this post for more details on the “redesignation” requirement). We are impressed with the quality of many of the comments filed in the “Initial Comments” at the Copyright Office. As there will be an opportunity to comment again, including to comment on the comments, we will be posting selected Initial Comments to call to your attention. You can read all the comments at this link. If you are hearing about this for the first time, you have until June 28 to file a “reply comment” with the Copyright Office at this link.

You will see that there is a recurring theme with the comments. Many commenters say that they wish for The MLC, Inc. to be redesignated BUT…. They then list a number of items that they object to about the way the Collective has been managed by The MLC, Inc. usually accompanied by a request the The MLC, Inc. change the way it operates.

That structure seems to be inconsistent with a blanket ask for redesignation. Rather, the commenters seem to be making an “if/then” proposal that if The MLC, Inc. improves its operations, including in some cases operating in an opposite manner to its current policies and practices, then The MLC, Inc. should be redesignated. Not wishing to speak for any commenter, let it just be said that this appears to be a conditional proposal for redesignation. Maybe that is not what the commenters were thinking, but it does appear to be what many of them are saying. Perhaps this conditional aspect will be refined in the Reply Comments.

For purposes of these posts, we may quote sections of comments out of sequence but in context. We recommend that you read the comments in their entirety.

Today’s featured comment is from John Guertin, the highly knowledgeable independent publishing administrator who operates ClearRights in Austin, Texas. He works with many Texas artists whose music represents generations of Texas music vital to the Texas economy such as Marcia Ball, Guy Forsyth, Vallejo, Quiet Company and the South Austin Moonlighters.

Like other commenters, Mr. Guertin focuses on The MLC, Inc.’s failures to adopt world-class metadata standards. He offers insight to the Copyright Office similar to information the Office could get if they actually did a proactive deep dive on the MLC standards and practices rather than wait for commenters to get so disillusioned that they will sit down and write up their grievances when their frustration exceeds their fear of retaliation.

If Mr. Guertin is correct about bad old HFA data populating the MLC’s data, one consequence arises when the MLC, Inc. distributes its data feed to dozens of users. Does this mean that anyone who uses the MLC’s mediocre HFA data also has error-ridden data? What is the plan to unwind that one?

Lack of transparency
How does the automated matching process work and what is the logic for a match? We submit quite a bit of data to The MLC, yet titles go unmatched. It is hard to understand how a match does not happen when the system has been provided the song title, writers, isrc and supplementary data such as iswc, recording artist etc. It begs the question, what is the matching logic? If the song title, isrc and songwriter match 100%, how is a match not created? Having worked in the digital music space in the early 2000s at the onset of online digital subscription and download services, there was a fuzzy logic matching employed to help clear thousands of songs at a time. A fuzzy logic matching criteria would have to require a certain percentage of a given data field to match and thus enable matches to be made when there was punctuation or additional wording in the sound recording title such as “Live”. It’s hard to understand how so many line items go unmatched at The MLC when there are small variations in titles etc. Is a fuzzy logic protocol being employed, and if so, is it too tight?

New System , Same Old Player
The forward-facing organization we see is The MLC and its staff, however the vendor(s) used by the MLC is the same player, The Harry Fox Agency. The MLC data is often powered by and supplied by HFA. The HFA system, being a for profit, proprietary system, has been known for years to have old, outdated and/or incorrect data. One can often find the same song registered two, three or more times in the system. In most cases the publisher/owner is different or variant. This “bad data” has been allowed to proliferate the MLC system and has basically resulted in the same issues of old.

Having said vendor(s) also operating as match makers raises several concerns/questions, especially when incorrect matches are made based on this bad or outdated data. When an incorrect match is made (again how does this happen if the titles and songwriters don’t match yet publisher submitted data matches 100% and a match isn’t made?), the publisher is paid royalties.

The burden then falls upon the recipient to find the incorrect match, and then take action to remedy it by either returning monies to the MLC or having it deducted from future payments for other, non-related publishers and songs.

In some cases, the dollar amount of monies is significant and results in the publisher and/or songwriter being debited for the amount all at one time and unable to earn future royalties until the debited amount has been recouped. This can result in financial burden and distress for the publisher/songwriter. The publisher/songwriter may be dependent on these royalties to live on and due to no fault of their own, are subjected to a recoupment process for something they did not initiate. Why is this and why do we think this methodology works? Additionally, we are often told to contact the other party and get the money from them.

Lots of matches, yet even more unclaimed monies
An 80-85% match rate seems impressive until you look at the amount of money that remains unmatched each month. Approx $20 million in monies each month go unmatched and/or unclaimed. That’s over $200 million in a year. How and when is this going to be addressed? Yes, it’s much easier to ignore that and simply distribute that money via market share. But does artist/songwriter X really need more limos and vacation homes when the large majority of these royalties are indie songwriters that either don’t know about this, don’t understand it, or have been frustrated over the years and trained to think that they get micro-pennies for their efforts? We can’t blame this segment for not being totally engaged or not being educated on the complexities of the music industry. If we can put a man on the moon, why can’t we figure this out?

Lack of innovative strategy to clear the back-log of unmatched line items
What exactly is the process used to currently address this [old mediocre HFA data] and how is it being measured? We are told that outside vendors are contracted to perform this function, yet we see approx. $20 million each month in unmatched royalties. Clearly this strategy is not reducing the amount of “black box” monies at a fast enough rate and raises several concerns.

The first is that our senior songwriters and publishers are not getting younger by the day. They do not have time to wait 5 or 10 years for this to be straightened out. Many depend on the fruits of their past labor to live on. They deserve better.

With regard to the apparent inability to make matches and reduce the unmatched royalties, there seems to be other ways to approach this, which may currently be employed but we don’t really know due to the lack of transparency. Many of these unmatched recordings are songs that are registered at PROs. Those PROs have the songwriters and publishers, along with any recording data submitted by the songwriters and publishers. This is a good source of data which also has the contact info for those entities. A strategic partnership with other industry organizations, such as the PROs, should be made to help share and communicate data to bridge the gap with missing data which would allow matches to happen.

Also, where is the data that is being used to match coming from? Most indie artists use aggregators such as CD Baby, TuneCore, Distrokid etc. to distribute to dsps. This is the source of data that feeds to dsps. Such aggregators allow the input of inaccurate data without verification. All one must do is write something in the required data column (i.e. songwriters) and it goes through the system and starts populating everywhere. So bad data in results in bad data going out and reducing the likelihood matches can be made. Industry wide cooperation is required if we are to streamline these processes and make things efficient.

Read the entire comment at this link.