Every May, newcomers to LA inevitably ask, “What the hell are those purple trees?”
They’re called jacarandas, and it’s okay to pronounce the j. You can also find them in places with similar climates, including Australia and South Africa.
(Photo by [autumn leaf](http://www.flickr.com/photos/_autumn_leaf/299450076/))
My short story [The Variant](http://johnaugust.com/variant) has been on the market for a week. As promised, here’s an update on how the 99-cent experiment has gone.
Short version: I sold more copies than I expected, with fewer technical issues. I had picked the Friday of Memorial Day weekend precisely because I hoped it would be slower-paced, allowing me to fix whatever disasters struck without a crush of weekday traffic. But I could have been more ambitious, and a mid-week launch would have made more sense.
I get 35 cents on each Kindle sale, versus 89 cents on each download.
I’d be less grumbly about Amazon’s 65 percent cut if their reporting were better. Their DTP publisher tells you almost nothing about your sales. It only shows how many total units, with no breakdowns at all — not by day, not by state, nothing. Fortunately, I had embedded my Amazon tracking number in links from my site, so I do know that 458 of my Kindle sales came from people who clicked through from the launch page. That’s only a third of the Kindle sales, so many people were getting it in one of three alternate ways:
* Following a direct link from an outside site, such as [Daring Fireball](http://daringfireball.net/linked/2009/05/22/the-variant).
* Buying it through Kindle itself, either the device or the iPhone app. ((I have a hunch that a lot of readers tried out the iPhone app for the first time buying the book.))
* Finding the book on the Kindle bestseller list. ((I’ll have more to say about the bestseller list in another post.))
Downloads provide a lot more data. I’ve already written about the [international readers](http://johnaugust.com/archives/2009/mapping-the-variant), but the numbers also help show the falloff over time. It sold ten times more on the first day than the seventh.
(Note: I grabbed data at different times, so this total is 12 units shy of the table above.)
Today’s [NY Times article about The Variant](http://www.nytimes.com/2009/06/01/business/media/01august.html?ref=business) may provide a bump, but an esteemed colleague (Ze) cautions that he’s never seen a real spike from stories in traditional media.
At this point, I don’t have any big sense of What It All Means.
It’s a fine number of sales for a short story that would have likely been buried in some specialty magazine. But I’m not sure I can offer any meaningful analysis of the publishing model, partly because I started with a higher profile than many fiction writers might.
Could an established novelist duplicate (or exceed) these results? Probably. Could a talented but unknown upstart? Not as likely.
This kind of self-publishing certainly reduces the barriers, but literary brand recognition is still a huge asset. It is reading, after all. People would rather do almost anything than risk reading something bad. Free previews and a 99-cent price tag help, but the reputation of the author is likely a major factor in deciding to buy.
I’ll release sales info for [The Variant](http://johnaugust.com/variant) on Monday, but I wanted to offer up one bit this afternoon in case data-miners were inspired to do something with it over the weekend.
Sales of the [Kindle version](http://www.amazon.com/gp/product/B0029ZAPRW?ie=UTF8&tag=johnaugustcom-20&linkCode=as2&camp=1789&creative=390957&creativeASIN=B0029ZAPRW) are pretty opaque; I only know totals. But the [downloadable versions](http://johnaugust.com/variant) (pdf and ePub) give me names and shipping addresses. I was curious how many readers were international, particularly since Amazon’s Kindle is U.S.-only. (The answer: 37%.)
There may be other interesting things to explore and/or mash up, so I’ve stripped out identifying information and uploaded it in [Excel](http://johnaugust.com/Assets/variant_geo.xls) and [.csv](http://johnaugust.com/Assets/variant_geo.csv) formats. The file shows only transaction date/time, initials, city, state, postal code and country. ((Originally, I was going to include first name, which adds a nice degree of personality. But in a couple of cases, I worried that it was too individualizing, particularly with unusual names in smaller towns.))
If you do something interesting with the data, leave a link.
I flew up to Oakland yesterday for a lunchtime lecture and Q&A at Pixar. And wow. It’s really nice up there.
If I had to work in an office, I’d work there. It combines everything I like about Dreamworks/Amblin (lunch, toys, a noticeable lack of evil) with everything I’ve read about Google (daylight, servers, smart people on scooters). They even showed me the secret cellar where they mine joy.
My presentation was on Expectation as it relates to story. It was brand-new material that I was trying out for the first time, and I was fairly happy with how it went. Once it’s in a bit better shape, I’ll post some of the lecture on the site.
As frequent readers know, my geekery is almost limitless, so it was great to be able to ask questions about 32-bit color spaces and whether model articulation and prop interaction relied on message-passing. Sample inquiries: If an animated character picks up a can of soda, does that can of soda become part of the character’s domain or does it remain a separate object? (Answer: the latter.) If an explosion casts light, is that handled by VFX or the lighting department? (Lighting will probably get the last word.)
Many thanks to the Pixar folks for good questions and better answers. And a special thank you to Stephan Bugaj and Michelle Lindsey for setting it up.