Layout 1When Chris Anderson delves into the “power of collective intelligence,” he asserts that the web-based amplification of word-of-mouth communication has serious implications for the trading of recommendations and popularizing of content. In his model, aggregators and filters place genres, categories and ideas into neat contextual buckets that give their relative recommendations meaning and substance. This is instead of making an apples to oranges comparison between, say, a bag of tube socks and a power drill, the relative ratings of which would tell you nothing about either. But this position omits the key fact that – more than ever – content exploration is not driven by a linear, taxonomy-powered long tail experience but powered by search and recommendation engines that instead of driving us to the margins, more often than not consolidate users around popular, “mainstream” products to satisfy unavoidable economic constraints.

Popularity and obscurity, as Anderson points out, are not necessarily mutually exclusive. Even the most obscure music genre, for example, can have relatively popular artists or albums that have generated the strongest user affinity for that particular subculture. But this popularity, which is usually defined as a discrete value such as a rating or rank in the online space, is not merely an agnostic tool for the user. It represents an incredibly valuable business tool that is increasingly used to turn users into customers, to take someone from an entry point to a sale as quickly and as authoritatively as possible. The seminal long tail example of Touching The Void illustrates this functionality and is touted as highlighting the surfacing potential of these new tools.

More than additive sales for Amazon and the publisher instead of for the similarly themed book Into Thin Air, however, a much clearer reading is that Void was able to piggyback on marketing dollars to enter a niche that had simply not been properly exposed. In succeeding despite a major push to sell another book, Void represents a ‘corner case’ whereby the long tail becomes the ‘head.’ What it also demonstrates is the power that recommendation engines can have in homogenizing instead of distributing the spread of traffic throughout the tail so that, once an item is truly represented as trusted among users, it is already part of the ‘head,’ regardless of its origins or time spent languishing.

Ultimately, the business needs of content distributors – whether hawking digital or physical media – revolve around the bottom line. Boasting a large, long tail catalog is a selling point to bring in users to be sure. But from this point on, conversions are the goal and turning users into customers is only complicated by the breadth and depth of freely available content on the web. Targeted, trusted, peer reviewed content carries the premium that demonstrates itself to be successful, and a discriminating recommendation system ensures that the music, book, or hammer that can come prepackaged with this type of equity is placed front and center. Occurring over and over, feedback loops appear around these core ‘head’ products and even fans of the lesser-known niche are picked off as sets of products consolidate their authority in rank and ubiquity as a recommendation for other products.

Why wouldn’t content catalogs use their recommendation system to generate less popular but presumably still highly relevant experiences if these could lead to major sleeper hits? Google knows not only what most people are interested in from a particular search, but also how important a page is viewed from across the entire web. The algorithms that drive Amazon know what users ‘like’ you are interested in and know who has ranked what the most times. But more than anything, the gatekeepers to content have a keen awareness of what people actually buy. Regardless of their interest in presenting every option under the Sun, steady purchasing means steady income and a predictable quarter.

Economic forces – regardless of the efficiency and quality of search and the leveling factors of low-cost production – still favor high volume commodities. Even formats that are ostensibly free to deliver to the consumer such as digital music still benefit from 90% sales from 5% of content in creating a predictable model in which to sustain a business. The other 95% of content may indeed be long and wide in its breadth, but the tail will continue to be incredibly flat regardless of user affinity as long as these very easily documented forces continue to drive the implementation of content discovery and reinforce the popularity of that which provides revenue.