For years now, I've listened to Yahoo Music but I've recently switched to Pandora because it does a better job of selecting the music I want to hear.
For those of you who don't know, Yahoo and Pandora are online radio stations that select music based on the listener’s preferences. Despite the similarities, they approach music selection (or sourcing) from different philosophical perspectives. A Yahoo user rates artists, songs and genres and, based on their selection, Yahoo plays music liked by other similar listeners. It is musical democracy in action. A Pandora user, on the other hand, selects artists and songs they like and Pandora plays music with similar attributes such as instrumentation and rhythm. It is musical genetics in action.
Recently, my curiosity got the better of me and I decided to determine if there was an identifiable difference in the music recommended by the two systems. I set up a new account in each service and, after selected the Dandy Warhols as the single piece of information the services knew about me, played 100 songs on each station. I then assessed the popularity of the songs in each playlist by comparing the lists to a sample of Billboard Charts.
Subjectively, I concluded that the range of music delivered by Yahoo was much broader than that delivered by Pandora and that there was a positive correlation between my subjective assessment of the inappropriateness of the Yahoo songs (for a Dandy Warhols fan) and the number of weeks the song appeared on the Billboard charts. For example, Spice Girls' Wannabe (26 weeks on Billboard) was recommended to me by Yahoo and not played at all on Pandora (thankfully). A Yahoo song was 7 times more likely to appear on a Billboard chart than a Pandora song.
So what does this have to do with sourcing?
Lately, I've been mulling over the sourcing “long tail” as described by Dave Stephens. Dave's contention is that, as the costs of interacting with a vendor falls, buyers will be able to engage exactly the right vendor for each task and that this will increase the number of vendors a company deals with. However, for this to happen, procurement departments must not only decrease the cost of interacting with the vendors, they must also improve the efficiency of identifying potential vendors.
Currently, most sourcing is done using a yahoo-type system: the firms with the biggest reputations get invited to tender for work even if their skill set is less relevant to the task than other smaller providers. This creates a tipping point in vendor sourcing where the market tends to consolidate around a few large players.
In order for Dave's long tail to work effectively, the smaller players need to be given a chance to play. If this is to occur, the method of sourcing suppliers needs to change from Yahoo sourcing to Pandora sourcing – that is, from reputation-driven sourcing to functionality-driven sourcing. If we continue to source based on reputation, buyers will continue to be driven to the most popular suppliers rather than to the most appropriate.
By the way, Election Day by Arcadia is exactly the type of song I want to hear when I'm in the mood for Dandy-Warhols-type music, and Pandora played it for me.
Technorati Tags: outsourcing, vendor management
[...] Doug is a Procurement consultant in AUS. I'm enjoying his latest posts. I thought he did a wonderful job exploring & giving a real world (and fun) example of Procurement's long tail here. And I've thoroughly enjoyed his exploration of vendor relations, especially his latest post on contract management & fairness. Good work Doug! [...]