Digital distribution has been a mixed blessing for the music industry - greater access means less value. As David Reed discovers, leverage the extra benefits data can bring to the user experience could be one way to restore the fortunes of this sector.
What do you do if your product used to command a premium because it was scarce but is now seeing its value eroded as a result of abundance? How do you manage a business whose product can be rented, bought, licenced, consumed for free (and is frequently stolen)? And are there better ways to manage your metadata and information when each product sale may require you to pay up to 50 interested parties in its production?
Those are the challenges facing the music industry. It used to have tight control over the distribution and pricing of its output, even if its back office processes were (and often still are) manual, paper-based and slow-moving. With the Internet age, all that has changed. Consuming music may be easier than ever for individuals, but making that pay has got harder. Digital distribution still does not yield the same revenues as physical products used to.
But there may be new, alternative sources of value from this new model. “Data by itself is not meaningful, but it has an inherent value in creating analytics, business intelligence and metadata,” said Gregory Kris, CEO of music metadata specialists Decibel at an industry-focused event in April called “Music 4.5 Data is Sexy 2”
One area of potential value for data that has largely been overlooked in this sector is what the industry calls “label copy”. Knowing who the songwriter is or which artists are playing on a track is of interest to many consumers and could be a source of incremental revenue. Kris gave the example of Jimi Hendrix who played backing guitar for Little Richard. “That didn’t cost much to find out, but it generated tens of thousands of pounds in revenue through additional sales to Hendrix fans of Little Richard product,” he noted.
This example also points to ways in which digital distribution has actually destroyed value which used to be part of what consumers paid for. “We are seeing label copy diminution - that data is not available on iTunes. That is shocking,” said Jeremy Silver, chairman of MusicMetric, a real-time music analytics provider. In the era of vinyl, cassettes and CDs, it was easy to discover who wrote that song you love or to find out that the conga player on “Your So Vain” produced the first Captain Beefheart album.
Silver recalls that just 18 years ago when he built the first website for Virgin, he asked the IT department to build an application to analyse the information it was generating. “They said there was too much data to do it,” he said. With big data tools like Hadoop, volume is no longer an obstacle.
Alongside this, music lovers are leaving digital trails and sharing huge amounts of information about their likes and dislikes which the industry is starting to get interested in. The challenge, according to Silver, is “how to avoid creepiness.” He pointed to the mash-up app “Girls Near Me” which told male iPhone users about women nearby based on open source data from Facebook and the like. It was rapidly pulled from the Apple App Store after concerns were raised.
That does not mean there is no place for “fanalytics” which draw on similar social network sources. Indeed, a range of initiatives are underway that should make the whole process of managing and mining data and information easier. Global Repertoire Database is one example, as is the Digital Copyright Exchange.
Metadata standardisation and co-ordination will help with the internal processes that track where music is being consumed, whether appropriate fees or payments have been made and sending that money to the right people. At the same time, consumer-facing data-driven services are extending the ways in which music gets used. Shazaam, Decibel, Echonet and NueMeta are all using the massive volumes of data about music, its usage and who likes it in entirely new ways. “They are taking a new approach to the masses of data the industry throws up and making it more dynamic and efficient,” said Silver.
The hope is that this will yield added - and incremental - value. “We have five billion rows in our database,” explained Chris Carey of EMI. “I love good metadata and hate metadata because, for example, if I can prove the appeal of an artist based on what it tells me and meet the consumer with that product in the right place, we will make more money.”
Carey pointed out that many in the music industry have made a flawed assumption that data is the new value driver and the product itself is valueless. “How do you monetise a mailing list later on if you have given away your content in exchange for an email address?” he asked.
What data can do effectively for labels and publishers is improve their marketing by finding the right markets for artists. Soundout is an example which uses crowdsourcing to help “sort the world’s music”. It generates reports about new releases which are used by radio stations and labels to identify where a new track will connect with an audience.
“We process every new release and send reports to radio stations. It is being used by a major name and label to help decide how much to spend on marketing based on radio airplay,” explained founder David Courtier-Dutton.
If Soundout uses social networks to make the industry more efficient, Last.FM reverses the process to help consumers get better at finding new tracks they like. By asking which tracks a user listens to, it makes recommendations - a process the company has termed “scrobbling”. It now has 60 billion scrobbles in its system, which makes this matching process ever more refined.
“We even get scrobbles about tracks before they have been released. If we look at who did that, we can identify the ‘mavens’ and see what they are listening to now. That could be of great value to record labels,” pointed out Adrian Woodhead.
At the moment, users can enjoy a free service in return for this information. “We are nothing without our users,” said Woodhead. “It is about taking the data they provide, aggregating and enriching it with information from labels and affiliates to enrich the user experience to make it compelling enough that they will pay for it.”
Getting paid has become a significant challenge for the music industry, from artists through their managers and labels right across the distribution chain. In an era where the combination of free streaming services and unchecked privacy make music available whenever or wherever consumers want it at little or no cost, how can the industry sustain itself?
“The business is fundamentally fucked by the fact that the old business model was based on scarcity and we are not dealing with abundance. We are stuck in the old paradigm,” said Peter Jenner, former manager of Pink Floyd and The Clash and a director of the UK Music Managers’ Forum.
This is where improving the quality of product and copyright data associated with music tracks will help by optimising revenue collection and payment. Evidence that global standards can do this is easy to find - the International Telecommunications Union created the standards which mean a mobile phone user in the UK can call a friend in Singapore. ITU data standards also ensure each network charges correctly for that call and pays the counter party what it should.
In the music industry, the level of incorrect, inaccurate or simply missed payments for using copyright material is very sizeable. This is an easy win for the business if it can agree on and then use some of the metadata standards and information exchanges which are being developed. New initiatives like the Digital Licensing Framework, funded by the UK Government Technology Standards Board, are working to enable new, simpler ways for companies to licence music for use, for example.
“There are huge inefficiencies in copyright licencing at the moment because of legacy structures and mismanagement,” pointed out Nic Garnett, a solicitor consulting on the new framework. It is not uncommon for record labels to rely on filing cabinets full of paper contracts to work out who owns the rights to a track, for example.
Other changes will also be required which could be yet more challenging to this industry. “Data creation is becoming a fundamental part of creation,” said Mark Isherwood, lead consultant on Rightscom in the support of the digital supply chain consortium DDEX. “We will need to make sure the metadata is in place to to give us half a chance of getting your money in.”
That will mean capturing artist and songwriter information in the studio as tracks are being created, for example, something which might appear to run contrary to the creative process. Labels might also have to challenge artists about some of their other decisions. As he pointed out, “you can’t search for the symbol used by Prince instead of his name.” (Interestingly, there is already a movement of bands choosing names specifically because they can not be easily found via Google.)
In many respects, the music industry sits where FMCG and retail did twenty years ago when they wanted to improve the efficiency of their supply chain and billing systems by standardising data. Barcoding and EPI systems drove those improvements through and have created a highly-efficient logistics chain as a result.
Music has many more dimensions that a tin of beans, of course, not least because it is created in a unique way. But it still comes with rich associated information and business critical metadata. Working out how to leverage that might be critical to this sector’s future in the digital world.
As Patrick Towell, joint chief executive of Golant Media Ventures told the event: “Data and content has no value, only the verbs - what people do with that data and content. That releases the value. We need to identify that and focus on what releases the greatest value early on.”