Open Data follow-up: Q&A with Musicbrainz founder Robert Kaye
1. In your presentation at Music 4.5 Open Data you mentioned how you unwittingly created a free open data product that could be monetised and people wanted to pay for, how did this happen?
I set out to create a self sufficient free open data system in the context of a non-profit. This act was 100% intentional. The unintentional part was that it ended up being 100% profit, which is rather strange in the context of a non-profit.
The take-away lesson from this is that even though the core data may be in the Public Domain and therefore considered by some as free of monetary value, there are ancillary components to the data that might have value and allow a business model to be built around it. In the case of MusicBrainz there are two aspects that allow us to capture value:
1. The community of people that curate the data the data collectively create value on an hourly basis. Deriving financial value from this directly is not wise — communities and money have a very tenuous relationship that is easily damaged.
2. The stream of data that is the collective output of #1 can have actual monetary value. As long as this monetary value is directly re-invested into the project so that the community gets better tools, faster service, more resources, this arrangement can work. Any attempt to “capture all of the value of the community”, especially if the benefits are not re-invested, will almost assuredly damage the community, perhaps irreparably. And in my talk I covered that without a community, MusicBrainz would be worthless. And as Michael followed up in his post, this point is not to be taken lightly. In fact, I would say that this has to make up the foundation of any community that attempts to work in such a capacity.
2. What can others learn from it?
That they better understand how online communities and, in particular how open source works. Ignoring or glossing over even small details can send new endeavours straight to the rubbish bin.
3. What do you think the future is for open data online?
We have a long way to go before we can make real strides in this field. There has been much talk about the Semantic Web, which arguably will be powered by open data. Yet, after over a decade of talk, there still isn’t a killer application that exploits this data. At some point, if the Semantic Web is to have any real impact, a killer application will need to come along that will bring amazing features to people like my Mum. People whose eyes glaze over when you say the word “metadata”.
That said, open data is critical for the success of the Semantic Web. If data is constrained to data silos (or data islands) then the data is useless. If the public can’t access the data, either because of a lack of permissions or a lack of connection then this data is irrelevant to the Semantic Web.
Another point is that the world of Open Data has taken on this feeling that Open Data comes from government organizations, which opens governments slightly so that citizens can add value to the data. Personally, I see Open data much more comparable with Open Source. Initiatives created by individuals or groups of individuals (not government or corporate entities) that create data sets like MusicBrainz, Open Street Maps or Wiki Data. Not data that a government tosses across the fence for citizens to pick apart. I really wish to see the concept of Open Data shift towards citizen curated and citizen owned data.
4. How do you think the music industry will need to adapt?
The music industry has had many years of shoddy IT practices. When someone in the industry says “We’ll send you our database!” and then an Excel spreadsheet arrives, the Computer Scientist in me dies a thousand deaths. Fortunately there are great tools being built and the IT departments in the labels are working hard to improve things. One stumbling block I see to progress is that each player in the industry demand that they control 150% of everything. Clearly that cannot work — the labels and publishers will need to give up some control in order to fix the problems that plague the industry. Without some transparency with respect to licensing and where the money flows, money will continue to flow into the wrong hands, leaving out the (usually small) bands that are not adequately represented.
Labels will need to recognize that their data isn’t really as valuable as they think. They print it on their CDs — we have the data, so how can they think that they will keep in control of the data? They will have to get used to the fact that their business needs to become a little more transparent; if they refuse then they will be replaced by companies that understand that some level of transparency is necessary. Labels and publishers will need to learn how to work with people — rather than erecting walls at every step and turn, they need to start building bridges. The people who build bridges will win in this business and those who build walls will languish behind those walls.
5. In an ideal world how should open data be used and treated?
It should be examined, picked apart, analyzed and scrutinized. People should call it crap, make inflammatory statements about and then roll up their sleeves and work to make it better. Open data needs to be ubiquitous, accessible, up to date. Open data needs to be free so that people can begin to take ownership of a small corner of the data and curate it as they would tend their garden. Until we enable all of these things, open data cannot succeed.
6. You talked a bit about transparency when it comes to your salary as well. If you look at things from an open source point of view, how should content creators earn their living?
I think its unfortunate that we’ve arrived at a point where artists have been starved. The stubbornness of the labels in the last 20 years has irreparably hurt the industry. The fact that Napster was created showed the demand for digital music and then the collective efforts to shut Napster down showed the closed mindedness of the industry. Napster and everything that followed, drove the price of music down to zero. And I think no one really thinks that that is fair, yet that is what everyone expects.
MusicBrainz works hard to establish as many revenue sources as possible. While some of them don’t add up to 10% of our income, it still is income and this income is diversified. We fight for every dollar that comes in and we are as aggressive as we can be given our structure. I still consider us a scrappy start up, even though we’ve been around for 9 years and have been in the black for 8 of those 9 years. As a whole, our income enables us to do what we love to do: Make MusicBrainz better.
Content creators need to embrace a very similar model. A recording contract is indentured servitude and will not allow you to give up your day job. Creators need to treat themselves as a business, complete with business sense, marketing, accounting, business development and an honest voice towards their fans. They need to run their creative endeavour as a business and then identify and exploit every possible revenue source. Then re-invest the money to create better content and reach more fans.
And any artist who is unwilling to do this, needs to realize that their endeavour will be limited to “playing for beer money”.