But, but but... It is just that discovery is everything now. In an age of viral media, tastes flowing across social networks and recommendation engines, serendipitous encounters with new content in a media-saturated universe, tracking our tastes to their source is endlessly fascinating. Wired editor Chris Anderson is absolutely right, the tail is made longer by digital technology.

After broaching the long tail theme last week, the off-deck mobile content leader Thumbplay contacted me with some interesting usage statistics. That site, which charges a $9.99 a month fee for access to a package of games, wallpapers, ringtones and video, lets users manage their choices mainly via the Web site. The site gets about four to five million uniques a month, says CEO Are Traasdahl, and even the WAP portal gets up to 20 million page views.

The most interesting stat at Thumbplay is the level of diversity among tastes. Unlike the phone deck, where up to 80% and 90% of sales revolve around the top ten titles that appear in a catalog, about 83% of Thumbplay sales come from outside the top 20. Only 17% of sales occur in the top 20. Traasdahl tells me that only 2% of orders are coming from the best selling item, and the users are spreading their purchases across 11,000 or more unique orders every week. This follows the patterns that feed aggregator Buzzwire also sees among its users. Orb Networks reports this week that they have over 2 million users of their remote access tools. The Orb system lets users go fully personal and access their own library of media from any device.

Clearly there is a tension between the content serving mechanisms of the deck and what people really want when given the freedom to explore and discover. But what tools really work best here? Which ones leverage the complex discovery methods we cultivate in a media saturated environment?

And so I return to my irritating party question: how are we hunting media in these always-on, always-there times? Traasdahl tells me that there is a search box on the Thumplay WAP portal but "it is never good enough." The search box on the Web site is the most popular entry point for content discovery, but then simple browsing and "re-discovery" takes over. People start moving laterally across the network and often remind themselves of things they used to like. I see more mobile apps with "friends' recommendations" features, which let other members push content to you, but unless a service has real critical mass like a Facebook or MySpace I am always skeptical about how many friends a user actually bothers to amass in the tenth such socially powered application they use. For my money, Amazon still has the best recommendation engine I use. It pushes content at me based on my own history but also lets me move more liberally into other people's tastes.

As much as I like the expansive Web interface as a way to discover new content, I will not rule out the mobile platform for these duties. With a snappy enough network, I find myself snacking on samples from Verizon and Sprint's music libraries. I think that content sampling on a phone is a natural break-time entertainment. The new Wi-Fi-enabled iTunes store is wonderful for this, except it lacks the main store's recommendation engine to facilitate discovery. As younger users start relying on their phones as primary PCs, it will be incumbent on us to solve this problem.

The phone is the ultimate platform for personalized media experiences, the perfect place for a long tail economy, but its interface is worst at accessing it. The Web-to-phone model at Thumbplay is one solution, but I am not sure whether the marriage of Web and phone will be the final answer. Several mobile developers for the college segment have already told me that their young clients are asking them to put it all on the phone. We need to find ways to make the long tail wag in people's faces.

This commentary is insightful. I recommend it to others.