A Billion Pennies, or Show Me the Abrahams

Every time someone on the web visits a Demand Media website, the company earns two-thirds of a penny.

DM had 9.7 billion page views in the first half of 2010, bringing in 6,629,000,000 pennies. For the arithmetic-challenged, that's $66.29 million.

Two-thirds of a penny per view. That's the part of the Demand Media business model I just don't get.

Two-thirds of a penny ain't much! It's the type of ad revenue a decent blog gets. For my own pages, a collection of websites, blogs and online content, I generally shoot for -- and usually get -- income of a penny-per-view.

Even the articles I've written for eHow, which are all part of DM's vast empire, are earning me about 1.5 cents per view, meaning DM is getting a good deal more -- probably about 3 cents per view, assuming there's a 50/50 split of the revenue -- at least for that little chunk of content.

So why is the company's overall RPM (revenue per thousand views) seemingly so meager? And what does that mean for their IPO prospects?

DM's prospectus divides their content revenue into two big chunks, income from site like eHow that are "owned and operated" by DM, and income from their "network of customer sites" like NFL.com and the San Francisco Chronicle.

Their owned and operated sites get a decent 1.2 cents per view (first half 2010 earnings of $46.06 million on 3.9 billion views), and both views and RPM increased about 20% over the year before.

Their customer network adds up to a lot more views, probably an indication of the popularity of the sites they're teaming up with, but due to shared revenue, a much smaller RPM. Demand Media earned $19.66 million on 5.8 billion views, or 1/3 of a penny per view. These views, too, grew by a big chunk (24%) over the prior year, but RPM only increased 6%.

The bottom line here is that DM's less-than-a-penny-per-view revenue puts their earnings on par with about a zillion bloggers and website owners running Adsense on their pages. Where's the earning boost of the much-vaunted secret algorithm that DM uses to select titles and maximize income?

I suspect that what's happening here is a large dilution effect of some sort. DM probably has a billion or so core views that are earning very well...more like a nickel per page than a penny per page. But at the same time, the company is spending effort and cash on bringing in billions of other views that, so far, aren't doing much on the earnings front.

This is where the crap-shoot part of the whole IPO comes into play. Unless DM can substantially increase revenue to something like two pennies per view, it's hard to see their content machine cranking up a profit. Down the road, as an ever-larger proportion of their views comes from established (and essentially cost-free) content from prior years, hopefully that's exactly what the company can make happen.