When I was in the midst of preparations for the Torino 2006 version of DFL, I privately predicted that I would get less traffic than I did in 2004, but possibly more revenue — and that I would get more traffic in the first week, but less in the second.

I figured less traffic on the basis of less global interest in the Winter Olympics and, as a result, less media interest to drive traffic to the site. I figured more revenue because I thought my AdSense ads would perform better despite the reduced traffic — I’d placed my ads very poorly in 2004, so despite huge media interest I made very little money. (One wants to be rich and famous; one would prefer to be rich rather than famous; I ended up being famous without being rich.) So I anticipated better results.

The question was, would I be right, and by how much?

Well, the numbers are in; let’s take a look. First, here are number of visitors for both the Torino and Athens Games.

(The 2004 Olympics are in blue, the 2006 Olympics are in green. The graph is to scale, but I’m not telling you what that scale is, to stay out of trouble with Google insofar as divulging AdSense stats are concerned. On the traffic graph, I’ve also highlighted some events that had an impact on traffic, more on which below. Both graphs begin on the day of the Opening Ceremonies and run to the end of the month; DFL did not originally start until the third day of the Athens Games.)

DFL Traffic, Aug. 2004 vs. Feb. 2006

It’s easy to see that traffic was considerably less this time. There are probably several reasons, not the least of which is the smaller global audience for the Winter Games. Something like half as many countries participated this time, and many countries that did sent only a handful of athletes. Bottom line, it mattered less to fewer people.

As a result, there was much less media interest in DFL this time around. All told, I was interviewed five times for DFL’s Torino edition (covered here and here) — less than I was on just one day during the peak of the Athens frenzy. Not that I minded overmuch. Experience taught me that offline media had a much lesser impact on my traffic (and hence my income) than online sources: what spiked my traffic last time was when wire-service stories were picked up by Yahoo! and its ilk; while I made the front page of the National Post, the URL of the site was mentioned nowhere! And live radio and television interviews were particularly stressful and took up more time than you might think: I might lose two hours for every five-minute radio interview. So really, I wasn’t bothered by the reduced media interest.

I also figured that much of the novelty had worn off: guy doing blog about last-placers at the Olympics is news; guy doing blog about last-placers at the Olympics again, not so much.

As a result, most of my traffic was driven by blogs picking up links from other blogs. A link from Dave Winer at the outset was picked up by several other blogs, but Jason Kottke’s link was the biggest single source of traffic and links from other blogs. Word of mouth spread much more organically this time: I was amused to see links to DFL proliferate amongst the LiveJournallers. It was interesting to see how links to me proliferated; too bad it could only take place over a couple of weeks.

(I did get a couple of links from media sources without being interviewed by them: on PBS’s MediaShift blog and Tagesschau.de, a German TV site. Both were able to do their articles based on material on the site, which I’d prepared for just that purpose.)

Now let’s take a look at earnings. (Again, I’m keeping exact numbers secret to comply with Google’s terms of service.)

DFL Earnings, Aug. 2004 vs. Feb. 2006

Isn’t that interesting? As I hoped, I did a little better financially. It turns out that raw traffic isn’t everything: on a per capita basis, advertising revenue was nearly thirty times higher this time around. Whether that’s due to better ad block positioning, “better” ads (more on-target or more lucrative), a “better” audience, or simply a more mature advertising network, I have no idea.

Looking at the big picture, though, it was still lousy pay for the amount of work put into the site, but, of course, remuneration wasn’t the only reason to do it. Neither was fame, and neither was having a nice item to put on the résumé, though those were all factors. But it was a lot of work, and a lot of stress. I’m not sure I want to do it again.

We’ll see how I feel — and how busy I am — in 2008.