Following Apple's purchase of Semetric, David Balfour considers the role and importance of data analysis for various types of music companies

23 January 2015 - Press release

The news this week that Apple has apparently paid around $50m for UK tech company Semetric is worth investigating. It’s a fantastic result for Semetric, which many of us know well for its Musicmetric data analytics platform. It’s also a pretty eye-watering price tag, depending of course on how much money you have, which is not generally much of a problem for Apple. Compared with the estimated $3bn Apple reportedly paid for Beats Electronics, $50m is not so much. Nevertheless, Apple must have identified some really meaningful value in the technology Semetric has built 

Our experience of Musicmetric has been as a data tool which has been primarily aimed at labels and artists. The platform enables them to collate, analyse and act a range of data around their artists such as download sales, streaming stats, social media trends and more. For many artists this is data which they already have free access to, so what does MusicMetric add? The company would argue that its tools help to identify notable trends and ‘significant happenings’ from the available data – enabling artists to react and plan accordingly. At its best, such data analysis might inform tour routines, social media posting strategy and even release planning and digital platform strategy. To play devil’s advocate however, one might say that any decent artist or label should have a firm grip on all this data anyway. For small to middling artists there are a key number of data sources to follow as an essential part of being the artist and label – is a shiny tool really needed, and does it actually deliver that much additional value, taking costs into consideration?

It’s clear that for big artists with many thousands or millions of fans, interpretation of such big data can pay clear dividends by providing true, actionable insights. Has the general usefulness of such tools to artists and labels been proved however? Analytics tools such as Musicmetric, as well as many others, have been selling themselves hard at conferences for years now. Despite claims that ‘data is the new oil’ et cetera, such tools don’t seem to have had a transformative effect on the way that artists and record companies go about their jobs yet. A good or bad record is still a good or bad record, and no amount of analytics can alter that.

To what extent has data informed the marketing and promotion of such records?

One frequently reads that Radio 1 playlist policy is increasingly driven by data – and social media performance in particular. Is there really any sophisticated analysis of these platforms happening, or is it mostly a ‘check the numbers’ game? Recent research would suggest that all this apparent data analysis – in the case of Radio 1 at least – last year to a place where the playlisted artists are almost without exception those from major labels, which are probably unsurprisingly also those with the strongest social media stats. This is not to criticise Radio 1 in particular, it’s more a way of raising questions about the importance given to such data and whether that has compromised the principal of gut feeling in the modern industry. Have we become so reliant on data that we’re now unable to take risks on artists?

All this leads back to trying to estimate what value Semetric brings to Apple. In a year which should see Apple launch a global on-demand streaming service, as well as roll out its iTunes Radio service into new territories, Apple is itself entering into some uncharted waters. Whilst there’s little doubt that Apple could develop its own analytics packages, perhaps it simply doesn’t have time. So which data does Apple want? Musicmetric’s reach is wide. Is Apple looking to acquire its ability to follow and analyse individual artists and their performance on the key music services? One must wonder whether companies such as Vevo, YouTube and Spotify will continue to provide feeds to a service owned by a competitor, it seems unlikely. The value for Apple therefore probably lies elsewhere. Many journalists and commentators have quickly assumed that Apple wants Musicmetric’s data to help it improve the quality of its recommendations on its new platforms. Producing high quality recommendations does seem to be a logical priority for any modern streaming service - not least since suggestions have historically been pretty bad on many platforms. Spotify purchased The Echo Nest partly with this thinking in mind. It’s logical that Apple would place similar emphasis on producing high quality recommendations. 

Providing a well-researched recommendation can make the difference between a user loving a platform and leaving it. This applies equally to on-demand and radio-type services. Can Musicmetric’s platform make the difference here? It seems it would require some further development to take it beyond merely identifying links – as far as we know there is none of the actual sonic fingerprinting and analysis that happens on The Echo Nest, for example.

Perhaps Apple will have an interest in tracking interest around artists to which it is considering adding some editorial push? iTunes has a strong track record of putting its weight behind artists and it employs some of the best and most experienced human music programmers in the business. To continue this editorially-driven approach, backed up by an additional layer of data, would be entirely sensible. Apple might also have a strong interest in tracking interest and activity around the launch of its own upcoming music products – such data tools need not only be used for analysis of music and artists.

As ever, the business of speculating on Apple’s plans can be as never-ending as it can be pointless. We will however continue to be interested by the developing balance between the use of data and the value given to gut feeling, by labels services, radio stations and beyond. Whilst it makes clear sense to back up decisions with real-world statistics, over-reliance on such statistics can also lead to a very homogenised music experience. Can artists only break through in this age if they focus as much on their social media as they do on their music? Or does the key role of tastemakers who aren’t afraid to take a risk continue?