Over on Lindsay Buroker’s excellent blog there is a post which very lucidly points out all the reasons Amazon may be making it less and less attractive to release free ebooks. Rather than rehashing Lindsay’s arguments here I recommend you simply go read the post and the comments. I’ll still be here when you get back.
I say may be making rather than is making because, as with everything related to the Zon, the only people who can be sure what Amazon is actually up to are the people who work there. It’s a black box as far as the rest of us are concerned. Certainly though Lindsay makes a cogent case as to what is going on and why and I have to say my recent experiences tend to support her theory.
As I’ve mentioned before I always thought that some of the powerful effects of the post-free sales bounce might have been entirely accidental, a simple consequence of the system being used in a way that no one at Amazon anticipated. In any case, the glory days of Kindle Select Free days do seem to be well and truly over.
As far as some people are concerned that’s not a bad thing. There are many people who think that giving away books for free devalues all writers work etc. I tend to be neutral on the issue. I think that Kindle Select Free used to be a powerful tool for boosting sales whether by accident or not, and I still think free may still be a useful strategy for many people in some situations. In terms of results, I can’t say mine have been all that impressive.
Making the first book in a series free has worked very well for increasing the sales of subsequent books in that series for a lot of people. I have been giving away the first books in both of my main indie series (The Terrarch Chronicles and Kormak’s Saga) for some time now and the results have not really been any better than when I used to use a 99 cent loss-leader. In fact, in one case, in the UK, the results have actually been significantly worse, although I suspect there were other factors, such as Amazon changing its algorithms, at work there.
You’ll notice that I have extensively qualified my statement above and that is deliberate. The internet is full of people fighting to the death about the correct price for ebooks, all of whom can produce data to support their arguments. The problem with this is that the data they are using is their own and may be directly contradicted by someone else’s. It is, unfortunately, the nature of the beast. Two books in the same genre with similar covers written to roughly the same level of competence can be released on the same day and one can go on to do very well and the other sink without trace. No one knows why. My own guess is that one might have been marketing slightly better or got some better reviews on the day of release and the algorithmic nature of the system amplified that advantage. But that’s just a guess. Luck is just as good an explanation as anything else.
Anyway, I digress. At the end of her article, Lindsay says that if she had two series, she would maintain the free book loss-leader on one and charge for the first book in the other and see what the results were. Unfortunately she does not, so she can’t.
I do and I am going to. Yesterday I began the somewhat convoluted process of making Stealer of Flesh, the first book in the Kormak series, non-free again. Because of the nature of the beast it is difficult to predict when this will actually happen, but it will eventually, so if you want to pick up a free copy of the book I recommend you do so in the not-too-distant future.
Of course this is not a perfect experiment by any means. The Terrarch books have been out far longer, have many more reviews and are much more deeply embedded in Amazon’s recommendation engines. All things being equal they tend to sell slightly better than the Kormak books anyway, unless it’s in a month when I release a new Kormak book.
And, as is always the case, many other factors can affect sales. This is more just a finger in the wind experiment to see whether having a free loss-leader for a series makes much difference now that Amazon has once more tinkered with its algorithms. I’ll report back after three or six months if I feel there is anything worth reporting. Watch this space!
Any data to add or theories of your own? Feel free to mention them in the comments below!