Week -2 (Sept. 11-17)

• 30 comments
• 2200 words
• 300 words per day

Week -1 (Sept. 18-24)

• 30 comments
• 2500 words
• 350 words per day

September 25: We add the Recent Comments widget.

Week 1 (Sept. 25-Oct. 1)

• 225 comments
• 16,000 words
• 2300 words per day

Week 2 (Oct. 2-8)

• 200 comments
• 16,000 words
• 2300 words per day

Week 3 (Oct. 9-15)

• 250 comments
• 23,000 words
• 3300 words per day

Week 4 (Oct. 16-22)

• 225 comments
• 18,000 words
• 2600 words per day

Observations and Speculations

1. If you love interacting with your audience, you’ll probably enjoy a Recent Comments widget.  We are getting considerably more comments both from commenters that had never before left a comment and those that had.
2. In the month since, pages with many comments have gotten considerably more traffic.
3. The proportion of our readers that are coming directly to our site (rather than using a search engine or referral link) has risen considerably.  That encourages me.  If I had to guess who was most likely to buy a book from our website, I’d say it would be someone that came directly to us rather than someone to stumbled upon our website.  Why are we getting more direct traffic?  Perhaps the site feels more interactive and readers feel more engaged.
4. Site traffic increased considerably in the weeks after we added the widget.  There are many factors involved here, but there’s some circumstantial evidence that the widget was relevant.  In the two weeks before we added the widget, we had 1150 and 1250 visitors respectively.  In the four weeks since, we had 1350, 1850, 1550 and 1550 visitors.

For example, we’ve tabulated our numbers for January 2008 and found that Atlanta and Toronto currently have almost as many Superhero Nation readers as NYC.

I think this provides several insights that go beyond the “NYC-LA-Chicago” strategy.

1. Our top six cities (in North America) are Los Angeles, Chicago, Houston, Toronto, New York, Seattle and Atlanta. I think Vancouver, Dallas, St. Paul/Minneapolis and Madison also stand out because they cram significant numbers of fans into a tight geographic region.

2. Watch out! Analytics makes California look deceptively insignificant by splitting Californian cities (and Toronto) into deceptively small geographic areas. For example, Los Angeles looks like it has far fewer hits than New York City, which is only split up into three zones (New York, Brooklyn and Astoria). By contrast, Los Angeles is split up into at least twenty zones.

3. We’re seeing some evidence to support B. Mac’s expectation that military readers would be statistically significant. (Did you know that Marvel Comics prints comics specifically for the military? Comic books are apparently kind of a big deal on military bases). I’m still somewhat skeptical, but we’ve seen heavy return traffic from several Air Force bases, particularly in Nebraska and California. B. Mac speculates this is because the USAF is relatively techy. I’m just hoping it’s not because the USAF’s Office of Special Investigations is researching an unfair portrayal case against us because Superhero Nation makes the OSI look so wacky.

4. Georgia’s on my mind! Atlanta generated 80% as many site views as the New York metropolitan area. However, unless we want to have angry Bulldog fans riot and kill us, we should probably get BM to tone down his Florida Gator loving… if you look at Northern Florida, you can see that he made a slight adjustment to the color scheme around Gainesville (the home of the University of Florida’s Gators).

5. Canada has been exceedingly kind to us and, demographically speaking, Canada seems favorable because its cities are pretty clustered together. It’s got a few cities in its southeast, a few in its southwest, and Winnipeg* in the middle. The southeastern cluster is especially logistically helpful because it fits most of Canada’s population into an area about the size of Montana.

*Winnipeg is a small Canadian city that is notable for having consistently devoted Superhero Nation fans. It leads the world in the number of viewers that have spent more than 30 minutes on our site.

6. Not too surprisingly, we do well in most college towns. Madison and Gainesville made our map and we got so many hundreds of hits out of South Bend/Notre Dame– mainly tied to B. Mac’s writing classes, I think– that we had to filter them out. I think our main underperformance among college towns is Boston.

7. Demographically speaking, we perform worse than expected in the Northeast, particularly New York, and and drastically better than most American authors in Canada. I’m not too surprised about underperforming in New York. You don’t make too many friends by mocking them. And, besides, so much comic book stuff– pretty much all comic book stuff– comes out of New York already.

8. We overperform in Washington, DC and the northern Virginia area. That’s a bit surprising, given that we mock Washington specifically and US politics generally (suddenly I’m seeing the attraction for Canadians).

9. We only average about one Hawaiian viewer a day. As much as we’d love to take a trip out to the Aloha State, somehow I don’t think the publisher will spring for that. [Cadet Davis adds: what if we started blogging on Hawaiian politics to spike traffic?] In response to Davis, I would say… You dog! I like your thinking. … Likewise, we receive virtually no traffic from Alaska. Thank God. Having no Alaskan traffic guarantees that our publisher will not ask us to go there. (Not that I have anything against Alaskans. After a millennium or two of global warming, they might have a state that’s worthy of them).

Yesterday was otherwise unremarkable statistically…

Incidentally, three quarters of our traffic yesterday came from search engines. So, if we take the 40% number at face value, a lot of returning visitors (at least 15% of total site traffic) returned by using a search engine (instead of, say, just typing in the name or using a bookmark). Looking through the search queries, I didn’t see very many that looked like they would come from a returning visitor, like agent orange alligator or “captain carnage.”

This might just be a one-day fluke, but the change is so drastic that I wonder if Google Analytics has changed the way it measures the percentage of new visitors.

UPDATE: I checked the statistics again and now it says that we had 84% new users yesterday.