Net, Blogs and Rock'n'Roll

Latest discussions on music discovery, recommendations and social networks

There have been a couple of interesting panel discussions this week, on opposite sides of the US, about how people discover music and the growing role of, respectively, recommender systems and social networks in helping them do this.

At the SanFran MusicTech Summit on Monday Paul Lamere led a discussion (pictured right) between two techies who build automated recommendation systems and two 'curators' who make recommendations based on human skills and knowledge. Paul has provided his own summary and reflections on the discussion, as well as a video of the full proceedings — which runs for just under an hour from 14 minutes in. And thanks, Paul, for mentioning my book in the course of the discussion: here's more on the categorisation of listeners that Paul attributes to me (though I took it from Emap research).

In New York yesterday, the Digital Music Forum East ran a session on "Social Networks and Music Discovery: What It Means for Music Businesses". I can't find any recording of this, but Eliot van Buskirk of Wired has blogged the contributions of the panellists, who included people from US National Public Radio, the iLike social network and peer-to-peer monitors BigChampagne.

Promoting events and festivals in a 2.0 world

It's been a while since I posted one of the future scenarios that got cut out of Net, Blogs and Rock'n'Roll the book, so here's the fourth in the series.

The premise is that social networks, customer relationship management and audience profiling will converge, creating new opportunities for obscure and niche events to develop a good audience.

As with my earlier TV listings scenario, this one is presented in the form of a fictional interview. I must thank my friend Eric Namour, who, among a wide portfolio of jobs, is half of the London experimental music promoters [no.signal]. Eric took time to have lunch with me and explain a bit about how he works and builds relationships with performers and audiences. The speculative projections from his experience, and any lack of plausibility in them, are my responsibility.

The photo is of Tony Conrad performing at one of [no.signal]'s biggest events to date last summer, and is licensed under Creative Commons by Yeled.

---

How did you get into this line of work, and how have you seen it change?

Live performance is the arena where you get to see an artist evolve before your eyes, as they experiment or improvise. They interact with the audience, and there's an adrenalin rush you get from being in that moment. You don't get that from recorded media, and there's a demand for it: I don't think it will ever go away. So that's what attracted me to working in this field, and what I've tried to do throughout is create the right circumstances where that chemistry between performer and audience can happen.

When do you want to hear 30-second recommendation clips?

I know the question I'm about to ask must have been posed many times before, but I'm not sure I've ever seen an answer. I've been revisiting my thinking about the different 'modes' in which we listen to music. Sometimes we're open to discovering new artists and songs, sometimes we just want to listen to our favourites. Sometimes we want listen to music in 'lean back' mode where we surrender control of track sequence to a DJ, a playlist we made earlier or even a playlisting algorithm; and sometimes we're more prepared to interact and intervene, 'leaning forward' to sequence the music track by track.

Into this spectrum of listening behaviours come recommender systems like MyStrands (above in the image on the right) and iLike (below) that offer me suggestions for branching my listening based on what I'm listening to in iTunes at this moment. But they offer only 30-second clips of each suggestion, and the suggestions disappear at the end of the track I'm listening to. So my question: in what context would listeners break off in the middle of listening to a track they had chosen to hear a clip (usually fairly lo-fi) of something they hadn't chosen — and quite likely hadn't heard of?

Is this user experience just a reflection of licensing constraints that will disappear as soon as the services concerned have licensed full tracks? Or was it designed that way on purpose, irrespective of legal parameters?

When I was talking about criteria for recommender systems a few weeks ago, I suggested that the key pragmatic measure of a good system is the degree to which people keep coming back for more recommendations. So how much do users actually engage with the 30-second recommendations? Does anyone have any figures they could share — or even anecdotal evidence?

And if people do break off their current listening to attend to recommendation clips are they more likely to follow up on those recommendations, or to forget about them?

Going beyond music, radio producers insert trailers for other programmes in their features (especially on the BBC!), but would you ever insert a trailer for a film in the middle of another one? TV does that, but then people pay a premium price to watch films without such interruptions, suggesting that this isn't seen as adding value.

How community builds audiences and makes discovery stick

At a music marketing event in London a couple of days ago, Scott Cohen, Co-founder of The Orchard (and on the left in the collage), presented a nice thought experiment that demonstrates the importance of community and narrative in making sure that discoveries stick in the mind of audiences. What if Simon Cowell (on the right in the collage) just cut to the chase on The X Factor and presented only one show in each series, saying "we've done the research, we've done the auditions, we've consulted audiences, and here's the winner"? The result would be the same — a winner like Leona Lewis (middle of collage) — but would this winner sell as many records? Answer: No, because it's the backstage access, the community that grows around the competitors, and the story of the winner's rise that engages the audience interest. They make an emotional connection with the artist and the song, and buy it to help it succeed.

Yet, as Scott went on to say, record labels do all the research and auditioning that X Factor and its equivalents do, but they hide it from view, and just say, "here's the winner we've identified: please buy their records". The solution to this is not to create some contrived competition for every new signing (personally I hate it when art gets turned into sport), but it may be to find other ways to involve the audience and give them "backstage" glimpses of artists that help people engage with them. Even a simple blog is a start.

Which are the best social networks for promoting music?

Last week I spoke to students at Westminster University's MA in Music Business Management (I'm always happy to speak to students about digital discovery and social media, so let me know if you run a course that you'd like me to take part in). Lots of good questions in the discussion session, including one at the end that I didn't have time to answer, which was along the lines of "How can we identify which are the best social networks for promoting music?" Here's the bulk of the answer I sent as a follow-up — please let me know if you can improve on it (or challenge it), since I know my knowledge of web metrics is less than comprehensive.

The way I see it you can answer this question with crude figures, more fine-grained measures or with hand-waving arguments. Let's take each in turn.

1 The Crude

Crudely, there are measures from the companies that specialise in website rankings that tell you how which of the social networks get the most traffic. Nielsen and Alexa (here are some old data reported here, but you can get up to the minute data yourself direct from alexa.com)