Shared independent music sites could live up to their potential if they just had a better way of determining which of the songs uploaded to them are actually the most worth listening to.  There’s a way to accomplish this—but it requires some resourcefulness in overcoming obstacles that have so far deterred the major players in the field.

The Pandora streaming service has addressed the quality problem in a straightforward and fairly standard way, by hiring people with formal musical credentials to screen the independent songs submitted to them, and admit only the best.  The results have generally been good, as far as they go.

But unfortunately, this approach doesn’t scale well.  There’s so much independent music being created today that it’s difficult for anybody to hire enough staff to sift through it all.  And not having enough evaluators ends up replicating the problems of the traditional music industry, where people are so overwhelmed they can’t listen to all the demos that come in.

One of the strengths of the Internet, however, is that it enables large numbers of people to collectively perform large amounts of work in a crowdsourcing effort.  Why not use it to surface the best independent music being created?  People already prowl beaches with metal detectors in hopes of finding lost watches and pocket change.  Why couldn’t they be induced to sift through songs by currently-unknown artists to discovere a different kind of treasure?

A certain number of people would probably be willing to take part in such a social content curation endeavor purely for the fun and social value of it.  But to be realistic, they might not be the best suited to perform the function.  To begin with, a lot of us can’t even tell whether a singer is on key or not.  Also, we should be wary of creating a musical equivalent of the cat pictures currently inundating the Internet, which have attained their peculiar prominence over meatier content simply because everybody “gets” them.

Since one of the reasons good music gets overlooked by the traditional music companies is that it isn’t well enough suited to mass tastes, adopting a crowdsourcing system for independent music based on lowest-common-denominator appeal isn’t apt to have much of a positive transformative effect on the industry.  More likely, it would just serve up songs that the established mass marketers of the music business have already heard, and rightly passed up for not meeting their quality standards.

A more promising approach would involve providing incentives to people who others consult for their knowledgeability and taste to participate in the discovery effort.  Examples of individuals who might be well suited to perform this function include people who book bands for clubs, independent music bloggers, music writers for newspapers or magazines, and radio DJs.

Rather than simply presenting songs based on their overall popularity among these “taste mavens” in a one-size-fits-all approach, a shared music site could enable visitors to optionally select specific mavens whose recommendations they’d like to check out.  (After they’d found a song they liked, visitors would be able to see everyone who recommended it, so they could quickly learn which recommenders’ tastes best matched their own, and then tailor their subsequent browsing accordingly.)

In an arrangement of this type, the shared music site would play the role of a content “wholesaler,” while the maven-recommenders would in effect become content “retailers,” whose success would be based on their ability to create smaller, more targeted subsets of the whole with maximum appeal to a particular set of tastes—in essence, like successful clothing stores.

To help mavens succeed in this retailing role, the shared music service could provide them, at essentially no cost to itself, with their own free personal online recommendations galleries, where they could arrange songs within the various shared categories in whatever order they felt made for the best listening experience. 

Each gallery could also be given its own unique URL, which the recommender would be able to link to from his own blog or column—giving followers a more convenient way to check out the songs he’d recommended, without having to slog in reverse chronological order back through every post he’d ever made.  

To further boost visitor appeal, a suite of the site’s own sophisticated content discovery tools could be built into each recommendations gallery—again, at no real cost to anybody.

If the shared site also offered a streaming service, people with trusted musical tastes could perform another retail-type function as programmers of their own individual channels—effectively taking on the additional role of radio stations.