The Probabilistic Magazine Brand in the Social News Age
A working hypothesis for what a brand is in today's media fractured and pulsating ecosystem
But in a world where magazine are a lot more than their printed artifacts, what's a brand? I'll go through my evidence in a minute, but here's my current hypothesis.
Brands are probabilistic now. The primary power of a brand is to increase the probability that someone clicks on, upvotes, or links to a story associated with your brand. It is not calculated primarily on a per issue basis as in the past. Instead, it's a kind of implicit regression based on all the stories a publication has produced.
This isn't a completely foreign idea. In the past, a brand's strength would have been measured by how likely people were to buy or subscribe to a magazine. Today's probabilistic brand, though, is less coherent and much broader than before.
The web contains a wider and more diverse universe of readers than the traditional magazine audience, most people will have at best a hazy knowledge of any individual brand. But in this media world, where people don't have to purchase anything, people don't even have to be conscious of liking a brand to contribute to be more likely to click on its stories. It's all about the marginal edge in the war for clicks, votes, and links.
That's my working hypothesis. I've gotten there by looking at how publications are actually interacting with the Intertubes. Here are the three threads that I'm trying to bring together:
- 1. For any site I've ever seen, most traffic comes in through the sidedoor to individual stories. While everyone has a few sites they go to regularly -- The New York Times or The Awl or The Atlantic, say -- but in aggregate, most visitors to a website are not arriving that way.
1a. Homepages matter less and less. Because so much traffic goes to individual stories, the homepage -- the part of a web publication that nominally gives it coherence -- continues to fall in importance.
1b. Most people are reading broadly across the web, picking up links from various sources and piecing together ad hoc publications they assemble in their minds.
- 2. Most sites' monthly or quarterly traffic is delivered by a very small percentage of their total stories. Our traffic world is spiky.
2a. Traffic gets pushed around by increasingly diverse and powerful social news sites. Reddit, Hacker News, Digg, StumbleUpon, Twitter, Facebook, and LinkedIn can all drive powerful traffic waves that dwarf an individual site's native traffic.
2b. These social news sites are increasingly conscious of themselves as communities with editorial power.
2c. Only four pieces of information are available about stories on most social news sites: Headline, dek (the one-sentence that describes the story), popularity, and the domain on which it resides. This makes domains more important than people think.
- 3. All magazines used to compete for roughly the same audience. Affluent people between the ages of 25 and 60 who might be looking to buy a car or some other expensive item. Our addressable market is much larger now: it's everyone on the web.
3a. The wider web people are less likely to be intimately familiar with any individual media brand.
The toughest thing for the probabilistic magazine brand is to find some kind of coherence. In the traditional sense, coherence as a package of interrelated content is gone. The story is the unit that matters, after all. But a big part of the value we add *is* structuring the world in a consistent way. So, the question becomes: what can form the basis for a new coherence for magazines?
One answer that is specific to The Atlantic but extensible is very old: moral purpose. This magazine was founded as an abolitionist publication and that helped structure the varied voices that ran in its pages through the early days.