One of our colleagues and a frequent contributor on this weblog, Andrew Jaquith, has a signature file for his e-mail with a quote that reads “the plural of anecdote isn’t ‘data.’” It’s appropriate when considering the nature of a digital universe in which readily-accessible information can be easily assembled in such a way as to identify an economic groundswell or a cultural zeitgeist.

All too often, we mistake analytical models for empirical data and blur the lines between positive and normative. In a word, we forget the nature of science itself. As industry analysts, we regularly run into this issue ourselves. Sitting on the front end of technology markets, we find the occasional researcher hoping to turn our analysis into empirical proof of something that is neither measurable nor yet to be. And turning those models into actionable recommendations adds another layer of complexity.

I identify these challenges as a way to say that technology markets are works in progress, and we almost always start with models for how we think things will turn out. Once those markets mature, we can then test our hypotheses.

Digital media is one such new market with many unproven aspects. There’s the promised mobile media market. New, yet-to-be-defined behaviorally targeted advertising businesses. And the Long Tail.

The Long Tail is interesting not for what it is, but rather for what is has become. This analytical model, based upon anecdotal information, transitioned from simple hypothesis to proven business theory in a few short months.

Now that the data is in, it looks like things are a little less promising for niche media. According to an article recently published in the Harvard Business Review, mass media is alive and well. In the piece, Anita Elberse asks the monumental question Should You Invest in the Long Tail?

There remain numerous questions about the validity of the Long Tail model, such as: has it been there all along in used book stores and flea markets? and is it simply more visible because it’s been digitized and made searchable?

But more importantly, virtually every digital media business model that I’ve seen in the past several years has incorporated some level of Long Tail thinking. Personalization and recommendation engines for media sites come to mind. So do the collections of niche media tied together neatly in a digital bow.

The digital media market has pinned significant hopes on another, more complex hypothesis — that people want to receive targeted advertising, that this advertising will perform better than current media products, and that targeted advertising will support profitable levels of media production and distribution. This complementary, revenue-side model for the Long Tail, has drawn considerable investment and must now prove itself in the marketplace.

If we accept Elberse’s conclusions (and McPhee’s long-standing hypothesis) — that people are actually drawn to blockbusters and best sellers, rating them more highly than obscure and niche media, then what does that do for the idea of targeted digital advertising and other performance-based approaches? Is the “Long Tail” synonymous with “Below the line,” or can one exist without the other?

Because there’s clearly something to the idea of targeting, but questions about the validity of the Long Tail help to better define the nature of targeted digital advertising. People are social. They want to share. And targeted advertising creative not only lacks economies of scale and scope, but it may be less effective than mass-market creative.

Stay tuned. We’ve recently published the first of many reports about behavioral targeting, and we’re defining the contours of this growing industry. Response has been strong, so we’ll continue building on the excellent work that we’ve seen coming out of the academic community. Thanks again to Anita Elberse for carrying the ball forward.