So, you’ve written your book, had it read and done the rewrites. At long last it’s time to push the ‘Publish’ button on the Amazon KDP site. So, what happens now? A huge booming silence usually. You next have to kick start your book sales by doing some marketing. I won’t go too deeply into this as what might be effective for your book might not be effective for mine and vice versa. All I’ll say is that you will need to advertise in some form so it’s important that you do your research and choose the options that best suit your book and the demographic that you’re aiming at.

Okay so your marketing is now working and you’re looking at your KDP book sales and ‘pages read’ to see what happens. Your book sales go up – Yippie! Then they go down again. Then up, down and all over the place. What’s happening? If you expected to be able to fine tune your marketing strategies by looking at your book sales, I’d forget it.

Let me explain.

I’ve worked for may decades in an industry where I used statistics every day as a core part of the job. We used statistical sampling to control our processes and we did this by identifying all of the inputs. Some inputs could be classified as ‘variables’ which might adversely affect the outputs so we then needed to either eliminate or control these. So what are some of the input variables if the output is your ‘book sales’ page?

I’m just going to mention a few obvious ones here as some variables may differ depending on the genre you’re working in.

Your ad quality and placement – This is something that you may think that you might be able to control using the feedback from your book sales. Not really. It may be impossible to say if any minor tweaks to your ads or marketing strategies have worked due to the noise from all the other factors. The only useful feedback I managed to identify is when I turned all of my ads off for six weeks. With that much data I could be reasonable sure that my ads do indeed help my sales but by how much I’m still not too sure. I think that you can only do your best with your ads and then try to compare the ‘before’ and ‘after’ using at least three month’s worth of data. Even then, due to the other factors mentioned below, the data may not be too trustworthy.

Seasonality – If your book is a Christmas cookbook then you might reasonable expect higher sales in December. However, if your book is deemed to a be a good beach read then you might get more sales in the summer months. I suspect that there might be a seasonal pattern to book sales in general but that’s something I can’t prove. So it just puts a little more noise in the data.

The competition – I’m a crime writer and I might be wondering why sales are sluggish. Unknown to me, several world famous crime writers have all published their books within a short space of time. Are the two facts linked? They could be but, in truth, you can do all the research you like on your competitors but you will never know for sure.

Social media – Has someone given your book a really good review on social media or has someone said that it stinks? It might be worthwhile doing a search from time to time for posts about your book but it will be like looking for a needle in a haystack. Not only that but there are quite a lot of large groups that keep their posts private so you’ll never be able to access them.

Events – Holidays, the weather, elections, major sporting events, disasters etc etc may all affect book sales in some way. Exactly how these events affect your sales is anybody’s guess.

Differing markets – If, like me, you sell in different countries then this just complicates matters even more. What sells in the US may not sell as well in the UK or Europe. If you want to do some analysis then you must separate the oranges from the apples as each market will have its own variable factors the obvious one being differing holidays. So, do Americans buy less books around Thanksgiving for instance? I’ve no idea and I doubt that I ever will.

The ‘black box’ – This is represented by the ‘Process’ block above and it covers all of Amazon’s algorithms, data and the internal processing that goes into producing your book sales page. I’ve called it a ‘black box’ because we have no idea of what happens in there. For instance, it seems to provide an author with real time data but we don’t know if this really is the case. I get the odd large spike in sales that could be down to something I’ve done but it also might be just a time lag in sales data from one of the markets. We don’t know, and will never know, exactly what’s going on under the hood.

So, we may never be able to pin down the true affect on book sales of the examples above and don’t forget that there will be many other variables out there, all adding noise to the system, that we will never be aware of. If you want to attempt some analysis then use the averages of large chunks of data and compare like with like. For instance, three months of data averaged out might tell you something when compared to another three months of data. You may find that, on average, your sales have gone up but don’t assume that it’s totally down to whatever marketing strategies you’ve adopted. Even more importantly, don’t get depressed if your sales go down as this might have nothing to do with the quality of your book or ads. It could just be a variation due to any of the above and the hundreds of other factors out there tossing your book around like a little boat on a stormy sea.

However, don’t despair. If you are getting any sales at all then you’re probably doing okay.

(Thanks to Geralt and Tho-Ge for the lovely images)