Forget Joe Average, he’s dead. Ten or twenty years ago, analyzing audiences was much easier. Medias enjoyed well-defined and relatively unchanging target groups. For television, networks had a precise idea on who was watching what, and specialized cable outlets knew their viewers pretty well. Newspapers had their content structure sliced to fit various audiences by center of interests, age groups and opinions. At the time, contents were bundled together, delivered on a unique platform for a flat fee, on a per copy or subscription basis: the popular sport section, or classifieds did subsidize the expensive but more elitist foreign section, all for a dollar or the equivalent of a euro.
In today’s marketplace, every single piece of information lies the open, naked, stripped of a set value. People don’t buy contents by the bulk, they peck at it, leaving to a third party (the unstable advertising market), the burden of financing it. As the content scatters on the internet, so does the audience. The money has shifted as well, with an expense of $260 per US household per year for digital services (cell phones, cable, broadband, satellite) that didn’t exist a generation ago. (Even the poorest families still spend $180 a year). This, in itself, makes it hard to hope for an extra $20 a month for news content that is widely available for free.
But the real competition is now for time and attention. Last December, in the United States, people spent 64 hours online, but stayed only 57 seconds on each web page, according to Nielsen. OK, it’s an average, and I’m about to kill this very notion in a minute. But still, it points to the time allocation challenge we face. Again, last December, American web users spent 6:24hrs on Facebook, 2:56hrs on Yahoo properties, 2:21hrs on various Google sites and 2:03hrs on Microsoft sites. As for the time spent on newspapers, it remains stable: around 20 minutes a month, whether you look at US or the European markets.
The shift seems to accelerate towards Facebook, which is becoming the absolute internet attractor: the amount of time spent per person on Facebook has tripled in just one year; in the meantime, Google gained only 10% and Yahoo and Microsoft slipped slightly. Interestingly, the Facebook time explosion even occurred at the expense of online video: with 3:13hrs per user last December, it remains quite high but it is down slightly, by 3.4%. And the Facebook effect probably explains why people are visiting a smaller number of web sites: 83 domains visited last month, a surprising 23% drop in just a year.
The advertising spending is shifting as well. In the US market, between August 2008 and August 2009, the amount spent online by brands decreased by 2%, as the money spent on top social networks and top blogging sites increased by 119%. Unfortunately, this is done on the cheap: based on 2009 revenue estimates, Facebook is grossing about $1.5 per user and per year in ad revenue. Just to put things in an unpleasant perspective, this compares to the $647 a newspaper such as the Washington Post gets from its print advertising for each of its buyers or subscribers. Make that $215 for each of its readers assuming a rate of three readers per copy. (This is based on the full 2008 year).
Coming back to the title of this column, analyzing trends has become more complicated: audiences are no longer monolithic, their breakdowns are hard to ascertain. This uncertainty makes an average a less and less relevant notion. For an online newspaper, what is an average reader? Consider two readers and focus on their different level of engagement. One is glued to the New York Times, Le Monde or Aftonbladet on his/her iPhone during a 30 minutes daily commute. The other, at 7 pm, casually glances at headlines while sipping a glass of chardonnay with TV providing the ambient noise. In this particular example, the level of engagement makes a crucial difference to the value of a reader.
As a result, the notion of heavy users becomes a critical one. The top 10% or 20% can be tied to a type of platform, a time of the day, in addition to the age group, socio-demographic profile, etc. Here are some examples:
Broadband consumption: according to Cisco, the average worldwide household consumes 11.4 gigabytes of data per month. Right. But 10% of those account for 60% of the total bandwidth consumption; and the top 1% of the heaviest users consume 20% of the whole bandwidth. The difference lies in the peer-to-peer volume that accounts for 38% of the traffic. (Interestingly, this number is down from over 60% two years ago.) Another factor is the rise of network-hungry video connection tools that cause the demand for bandwidth to explode during internet peak hours (which are 21:00-01:00 local time).
- Cell phone usage. According to AT&T, the top 3% of smartphone users – most of them iPhone users – account for 40% of the data usage. Again, video and audio exchanges and downloads are responsible for the imbalance.
- Twitter and Facebook. For the first one: 5% of users account for 75% of all Twitter activity and 10% for 86%; one half of all users never tweets, and the same proportion doesn’t not follow anyone.
As for Facebook, a research conducted by BLiNQ Media found out that, for a specific viral application used as a test, only 6% of the user’s accounted for 56% of the traffic associated with the app. In addition, it appeared that this group of heavy users was centered on 37-year olds vs. 25 for typical Facebook addicts.
- Time spent (again). Remember the 64 hrs spent on the net by the average user according to Nielsen? Well, the Digital Future Report came with roughly the same figure (which is reassuring), but with a slight distinction between the lightest users who spend 11 hrs a month online and the heaviest users who spend about 168 hrs – fifteen times more.
When assessing a business model, the importance of knowing the top grossing 10 or 20% of an audience is key. For traditional media with bundled-contents delivered on single platform, it used to make sense to consider the average group of users. In those days, the structure of the media led to an even distribution of revenue, regardless of contents and users. This no longer works for internet audiences: they are more scattered, segmented than ever and their specific value can make a critical business model difference. —email@example.com