July 30th, 2015


Monthiversary of self-pub: ALL THE DETAILS

RA Sales Graph 2015-07-29

This is the Amazon sales graph for my polyamorous fantasy romance novel, A Rational Arrangement. (Available at a URL near your mouse cursor! Amazon ~ Kobo ~ Nook ~ iBooks ~ Print).

Graph by request of alinsa, who thought the sales graph in particular would be cool.  I left the scale on because I do not care if the whole world knows how many copies it's sold. HERE, HAVE SOME DATA, WORLD.

Incidentally, this is the graph that Kindle Direct Publishing gives authors when checking their reports. There are add-ons available for KDP, but I haven't gotten any because the basic page suits my purposes fine.

The "days" are weird.  I don't know when Amazon rolls its day clock over, because the downloadable spreadsheet they'll give doesn't have the same sales-per-day as the graph.  Like the graph shows I sold 3 copies on 6/23 and 1 copy on 6/24, but the spreadsheet shows no sales on 6/23 and 4 copies on 6/24.  Even stranger: it's not consistently "graph shows some sales with a datestamp a day earlier than the spreadsheet". Yesterday morning, the graph had 8 copies for 7/28 and 3 for 7/29, while the spreadsheet gave 11 sales on 7/28 and 1 on 7/29. Uh. Okay! Whatever, the overall numbers match up and this is still data a billion times better than what authors get from major publishers.

Some landmarks for reference:

6/23: Uploaded book to Amazon at night.  Alinsa bought a copy to make sure it was working. It was not.  Alinsa ironed out an erratic problem with the .mobi's display of italics on some devices. Alinsa got it sorted (Amazon screws with the .mobi after upload and she has to make sure they can't mess it up with what they do). I re-uploaded on 6/24.
6/25: "Soft launch": Amazon copy looks good now! I let my Twitter followers know it's up on Amazon and propagating to other book-selling sites. (Kobo and Barnes & Noble already had it -- and were not affected by Amazon's glitchy handling of italics anyway -- but iBooks was dragging its feet). Several people retweeted the announcement. 16 copies sold! I was a little excited, and also a little "this is probably half of what I am ever going to sell, oh well."
6/29: Official launch! I pimped my book everywhere I have a web presence. I mentioned it was on sale ($2 off! Buy now!) I got a fair number of retweets and tweet-quotes on the lines of "buy it! It is good!" plus a tremendous boost from haikujaguar, who included a recommendation and link to it from her blog.
7/7: The last day of the $4.99 sale.  I originally intended to have the sale run for 7 days, but that would've ended it on a Sunday, which is a bad social-media day. So then I planned to end it on Monday 7/6, but forgot my phone and so couldn't make a "last day of the sale!" announcement. So I ended it on Tuesday instead.

My best guess is that most of the books during this period sold to people who follow haikujaguar.  Not that people following me weren't buying it (I love you all!) but she has a much bigger audience than I do, and frankly, a third-party endorsement of a book is much more appealing than the author's own endorsement.

Also, Micah released her Blood Ladders trilogy, which begins with An Heir to Thorns and Steel, just a few days after RA came out. So my book showed up on the "customers also bought" section of her Amazon page (and vice versa), meaning people who go off to buy her book were reminded that mine existed and might check it out.

7/8: Predictably, RA's sales dropped after the price increased to $6.99.
7/11: Unpredictably, RA's sales went back up.

My best guess at why my book started selling again:

Lois McMaster Bujold released a new novella, Penric's Demon, on 7/6. Micah's new series is in the top results of Bujold's "also bought", and RA is in Micah's. RA started out buried at around #12 of Bujold's "also bought" list, but it eventually climbed to #5 (and is currently #6).  So it looks like people are finding it either via Bujold's novella, and/or because Amazon is putting it on customer's home pages under "Recommended because you liked [X]" page. In any case, I started getting reviews from people who've never reviewed Micah's work but have reviewed Bujold's. One reviewer mentions that it was recommended to her because she liked Diana Wynne Jones's The Lives of Christopher Chant, which, wow, Amazon, that is SO SWEET OF YOU TO SAY.  And also whacked because that is a YA/middle-grade novel and um RA really is not.  But I still appreciate the thought!  And she did like RA, so I guess it works?

7/19: I have no idea what happened that Sunday.  Literally, I got nothin'. I hope it happens again someday.  That was nice.

7/27: My average up through this day has been about 15-16 copies per day.  Overall, it's been selling a little better per-day at $6.99 than it did at $4.99 (even discounting the slower sales prior to official launch).  I think the early boost from Micah was critical to getting RA up in the Amazon rankings and into Amazon's recommendations algorithm. Without that, I am pretty sure the book would've stopped selling once the sale period ended. Sales through iBooks/Barnes & Noble/Kobo dried up at that point. So it looks like people who hear about me directly (via my social media or someone else recommending me) might go to those stores to buy it. But only Amazon is selling A Rational Arrangement without anyone else directly pointing to it.

7/28: In honor of my monthiversary, sales have decided to halve. Ah well.

I don't know if the last couple of days are a fluke or a new trend towards lower sales.  I'd be quite happy to have sales hanging around 10 per day, in truth. Time will tell.

Total books sold at the $4.99 price point: 170*
Total books sold at the $6.99 price point: 380

Days at $4.99: "Available" period was 15 days. "Official launch" period was 9.
Days at $6.99: 21

Total sales to date: 550.

For a big publisher, 550 sales in the first month is "fire this author" territory.  For me: I am floored.  I am absolutely amazed that 550 people bought my book.  People I don't know bought my book.  People who don't even know anyone in common with me bought my book. WHAT HOW IS THIS I DON'T EVEN.

I keep trying to picture it: a random stranger browsing Amazon saw the cover image that I created and Alinsa laid out and thought "that looks interesting". And then read the blurb I wrote and went "sure, I'll try it."  This does not seem like it could happen even once, much less happen dozens of times.


Seriously, my expectations were set at "eh, maybe 50-100 copies for the year". This does not exceed my wildest dreams (I am a fantasy author: my dreams get a whole lot wilder than this, believe me) but it still feels unreal.

Great! But unreal.My expectations are that 7/28 heralds the beginning of the end, and that sales will peter off to 0-1 a day from here. But hey, (a) my expectations were wrong last time, so whatever, and (b) 550! Woo! That's not gonna change.**

In conclusion:


* Total sales is accurate, but the split between price points is +/- 5 copies, because the breakdown from D2D is less accessible than Amazon's, and I don't feel like untangling it.
** Okay, yes, it can change. Amazon lets you return ebooks. But it's not a huge factor the way "shipped" vs "sold" is wildly different with big publishers.

Edited to add purchase links at top & bottom, because I fail marketing 101. *^_^*

If you'd like to be sale #551, A Rational Arrangement is available at any of these fine websites: Amazon ~ Kobo ~ Nook ~ iBooks ~ Print
Me 2012

Excel VBA wheeeee x_x

My Google-Fu has failed me. Help me, Social Media, you're my only hope.

I'm working on a ridiculously overcomplicated Excel workbook, with a bajillion triggered events based on the cell contents.

I want the user to be able to hit "delete" on one cell, and have the spreadsheet trigger an event that clears a varied assortment of other cells.

This workbook has existing code that already does this "clear assorted cells" operation whenever the cell is changed to one of the six values on its picklist. The picklist values are assigned from a sequence of cells elsewhere in the workbook. One of those six picklist cells is empty. If you choose the empty value from the picklist, it will trigger the "cell is empty" event. It does this with code that reads either:
* ElseIf IsEmpty(Target) 
* ElseIf Target.Text = ""

But if you hit the delete key, it doesn't trigger the "cell is empty" event. It looks like Excel doesn't consider the deleted cell to be a "Target" the way it does when the user selects an item from the picklist, because deleting the cell won't trigger an event phrased as EITHER "ElseIf IsEmpty(Target)" OR ElseIf IsEmpty(Target) = False, and one of those should be hit if it were triggering the "change happened" code at all. There's code that prevents the user from entering an option that's not on the picklist, but this won't stop the user from using the delete key.

So. Any idea how I get "delete cell contents" to trigger an event?