< Thoughtcrime Experiments
Next >

[Comments] (5) Thoughtcrime Update: Day One: I've put up a zipfile containing high-resolution versions of all the TE artwork, suitable for making desktop wallpaper or whatever. I kept this in reserve because I was afraid that my server wouldn't be able to handle the anticipated traffic, and putting up a huge image dump would make things worse. But they make servers a lot tougher now than they did in 2004, which was the last time I personally had to worry about a server handling a lot of traffic, and we're fine. It helps to be serving static files.

Matthew McClintock of manybooks.net has made it possible to get TE in the weird-ass ebook format of your choice. Thanks, Matthew!

The process of creating a POD book is extremely slow by Internet standards, but if all goes well you should be able to buy a print copy by the end of the week. We'll be selling the print copies at cost.

Now, a bunch of boring numbers. TE got a huge amount of traffic today, almost all of which came from three big-name weblogs and one I'd never heard of. Boing Boing posted first and has sent us about 900 hits, John Scalzi posted later and so far we've got a huge 3600 hits from him [Update: this number is totally inaccurate because Scalzi included our image inline. The real number of hits is more like 200], and tor.com posted even later in the day and has sent us about 100 hits. We got 60 hits from Grasping for the Wind, a weblog I'd never heard of but which sent us more hits than higher-profile sites like Twitter.

Also of note are 1200 hits of unknown origin through Google Reader, and 900 from Livejournal friends pages. Add in other sources and the total number of hits is about 69003500.

Why did Scalzi send us so much traffic despite not having the first-mover advantage? Probably because his audience has a higher concentration of print SF fans than BB. [Update: this speculation is moot; see above.] BB readers are more likely to use feed aggregators, which hides their allegiance. Scalzi also simply posts less, so we'll stay on his front page longer. And I bet most of the people reading tor.com had already read BB or Scalzi by the time they saw the link on tor.com. Needless to say we appreciate all links and publicity and appreciation, those being our only payment. I'm not trying to min-max you guys.

How do hits translate into readers? As of this writing the PDF has been downloaded about 1150 times, the master OpenOffice document has been downloaded 120 times, and the zip archive of the website 90 times. By the crudest metric imaginable, about 1/6 1/3 of the people who visited were interested enough to download the anthology.

People seem to prefer downloading the whole anthology for later, but a significant minority are reading the stories online. Stories nearer the top of the TOC get more hits. The first story has 350 hits and the last story only has 50.

The art pages are much more popular than the story pages, which I was expecting. The most popular artwork (with the attention-grabbing title of "Pirate vs. Alien") has 450 hits. "Bio Break" (my personal favorite) has only 200 hits, but that's more hits than most of the stories.

The "random story" feature has only been used 75 times, which I find a little disappointing.

Comparison: usually the most popular thing on my website is Beautiful Soup. During the same period, the most recent version of BS was downloaded about 300 times, and version 3.0.7a (the non-lame-parsing version) was downloaded about 750 times. The documentation was retrieved 500 times. I find this hard to believe, but I never paid attention to BS downloads before, so I gotta go with the data. I think most of these are automatic downloads done while installing some other piece of software with a BS dependency.

Filed under: ,

Comments:

Posted by Fafner at Wed Apr 29 2009 22:42

I just emailed the mobipocket version to myself so I could read it on my phone, and the Google Ad, rather cleverly, linked to this: http://weadmire.net/product.aspx?productid=13396&gclid=CM7y4tnJl5oCFQFHFQodSXTo-w .
I was pleased.

Posted by Nathaniel at Thu Apr 30 2009 02:29

That's interesting about the early stories being hit more, since (it seemed to me) you followed the print convention of putting the best story last. Of course, that convention assumes that readers are reading straight through, not sampling a la carte, and in retrospect that's obviously less true on the web than in print. But I hadn't considered that particular effect of the different medium before, and now I'm wondering if there's a better convention that'll emerge for web anthologies.

Posted by Leonard at Thu Apr 30 2009 06:57

The random story thing was supposed to fix that discrepancy, but on the day we launched it was underneath the table of contents, so nobody used it. I moved it to the top, it's now getting heavy use, and the hit counts for the stories are a bit more even.

Posted by Nathaniel at Thu Apr 30 2009 19:50

Ah-hah, an excellent and sneaky solution.

It works by completely removing the hand of the editor in directing the reader on their path through the anthology, though; I'm still curious whether there's any way to preserve that aspect of the experience (even in part) for the flighty web audience. Perhaps not. (You could give different stories unequal weights in the RNG! What an opportunity for drama that would be!)

Posted by Leonard at Thu Apr 30 2009 21:12

My thinking is that ordering is for physical media and random selection is for electronic. Same for music.


[Main] [Edit]

Unless otherwise noted, all content licensed by Leonard Richardson
under a Creative Commons License.