I spent a few months of full time work investigating the viability of an eHow like site - trying to measure likely profit margins. I put the effort in because of this Demand Media article in wired: http://www.wired.com/magazine/2009/10/ff_demandmedia/ and because it was the opinion of some people whose opinion I value highly that such a site would be profitable. The idea actually went against my own gut feel because I was wary of all the SEO that goes on. Although typically localized in topic, I thought in aggregate the global phrase space would be pretty densely covered by now.

Anyway...

I optimized on both the demand side (data from keyword search tool) and supply side (number of results returned from searches, number of exact term matches, and page rank of each site in the top ten). I did this for about 200k phrases (yes, it was difficult to get the supply side data). I set up 3 separate test sites to test the most promising search phrases.

I couldn't make the numbers work. I was actually nearly an order of magnitude off being somewhere reasonable and even allowing for benefits of scale (which are admittedly hard to quantify) I still couldn't rationalize that it would be a good business.

So i'm quite happy to see this article in that it some what validates my own findings. I for one will not be taking part in the IPO!

I briefly tried this too, data mining thousands of high-paying keywords and looking for available exact match domains, so I didn't even have to worry about SEO to get to #1.

I still couldn't make this profitable, even to cover domain/hosting costs, much less my own time put into the project.

But you or I are not Demand Media, and that's why we're not making the big bucks. The fact remains that they are INCREDIBLY good at acquiring and monetizing low-quality traffic.

This has less to do with their SEO skills and more with the fact that, for example, they can produce content that will hit the Digg front page 80% of the time.

gaming the Digg FP for traffic ala Cracked.com is now (for better or worse) basically over, I'm sure Cracked is seeing a massive loss of traffic.

I think you are sharply underestimating three things: winners win in SEO, the importance of DM's search prediction ability (their only durable source of advantage), and "buying a poorly monetized site for women that has links from 1995."

Google keyword data is directionally accurate, but cruddy absolutely. Pretend it is quoted in terms of visits per Gestational period of a mythical creature: useful for comparison, not for calculation.

You have to ask yourself, what would Warren Buffet do? If you look at Demand Media not as an IPO to speculate on, but as something to own because you understand how they profit and you believe in what they do, then all I see is a big red stop sign.

Queue patio11.

I remember Patrick commenting on this issue a few months back. His comment was insightful and, from what I remember, he concluded that Demand Media was indeed profitable.

His SEO knowledge lends his conclusion credibility.

"His SEO knowledge lends his conclusion credibility."

No. What lends a conclusion of profitablity credibility is numbers that make sense and not much else.

That's definitely true. The grandparent poster was referring to Patrick's top-ranked comment on http://news.ycombinator.com/item?id=1599181 , which dealt mainly with how amortization of Demand Media's acquisitions requires them to amortize each acquisition's costs as a quarterly charge over the next ten years. A similar point is made by John Batelle in the Wired article, and is another reason why non-GAAP measures like EBITDA that exclude amortization are widely used.

This doesn't say very much about whether Demand Media is profitable once you exclude amortization, but it is an important thing to consider.