Net, Blogs and Rock'n'Roll

Audio whiplash from personalised ads with free music downloads

In the last month or so I've been dipping my toes into the ad-supported music download service We7. Their catalogue has expanded significantly since they launched, and there's now enough there that I like for me to have got 40-50 tracks that look interesting without having to spend too much time digging around on the site (the catalogue needs to expand further or the navigation and recommendations need to improve, however, for this to remain the case).

We7's model is based on their MediaGraft technology, which 'grafts' personalised ads onto the front of the track. Generally these are short enough — at 10-20 seconds — not to be too intrusive. Neither were the registration questions bothersome in terms of pumping me for details of my lifestyle. But the consequence of this is that you can end up with song intros like this clip:

Two points about this. One: the personalisation isn't working too well. I have never paid money for games hardware or software. I asked for Sim City for Christmas in 1992, but in the month or two immediately afterwards, I got so annoyed with myself for wasting whole weekends on it (sometimes not even eating properly) that I swore off it and have never gone back.

And, two: even if you leave that aside, listen to that audio whiplash between the ad and the track itself. When you're listening to French chansons from the '20s and '30s, just the tone of the ad is enough to make it come across like an aural invasion, and thus counterproductive in getting across the message.

Given that my selection of tracks is probably a better guide to my personality, lifestyle and buying habits than my age and gender, why not tailor the ads to the music rather than the listener demographics? No need for the clever software (patent pending) then, but the 'creatives' at the ad agencies might have to work a bit harder to think about the context in which their work will be heard.

Detecting exceptions and fraud on recommender systems

I stumbled across this iLike profile when I noticed someone who'd listened to David Bowie over 2 billion times. "Patrick S" has, according to his profile, been alive less than a billion seconds (and iTunes has been available for only the later part of his life), yet he's managed to register over 2 billion plays for 32 separate artists. The data concerned seems to be imported by iLike from iTunes.

Who knows whether this is just idle mischief, a software glitch, or a concerted effort at gaming or 'shilling' the iLike recommender system? Whatever the cause, recommender systems clearly have to detect such incongruent profiles — which can't be hard in cases as gross as this — and make sure the data is excluded from their recommendation algorithms. Cleaning up the profiles for this user and the 32 artists concerned would be a good idea, too, lest they undermine others' confidence in what they see on the site.

Swarmteams: capturing the buzz of trusted fans

If you've read my book, seen my presentation slides, or have just been following this blog reasonably closely, you will know that I'm a sucker for biological metaphors for discovery. I use these metaphors particularly in connection with information foraging and with the idea of people 'pollinating' social networks by spreading the word about their favourites on their profiles and blogs with widgets and playlists.

So I was especially excited to be introduced to Ken Thompson of Belfast-based Swarmteams (thanks to Steve Moore), because he takes biological metaphors even more seriously, and more rigorously, than I do. Take a look at The Bumble Bee, Ken's essays and blog at www.bioteams.com for proof: it's packed with resources and insight into teams and groups in nature (a research theme that goes back at least as far as Peter Kropotkin's Mutual Aid in 1902).

The extra good news is that Swarmteams have just launched an application of their mobile- and online-messaging technology to help bands tap the energies of their most committed fans. The idea, which also closely resembles some of the themes that Howard Rheingold developed in Smart Mobs, is that the top 'alpha' fans of a band will mobilise other fans and friends by spreading the word about new gigs/promotions/offers via SMS, email or instant messaging. The Swarmteams technology has feedback to establish trust in these fan networks, so indiscriminate spammers will quickly be filtered out. But trusted and proactive fans will get rewards for their advocacy.

This application of Swarmteams is a real world realisation and test of some of the armchair future speculation I posted a few weeks ago, so it's really interesting to see how it will turn out. The alpha fans that Swarmteams aims to appeal to are also almost identical to the Savants and Enthusiasts described in other research I've drawn on. This is definitely a space worth watching.