Recently I was looking at Brian Meeks’ book on Amazon Ads again – perusing his data analysis chapters, when I noticed a way of approaching the data I hadn’t tried before, or had at best only guesstimated.
Actually tracking conversions.
For weeks I’ve had the sense that the Wren thrillers weren’t converting. I was losing money, so they couldn’t be. Readthrough had to be bad. I was throwing clicks and ad money and nothing was going on.
Data is your friend, Brian Meeks said, so I decided to dig in.
First up, I needed a consistent idea of sales. Basic sales are easy – it’s just the number. KU sales can be calculated by dividing the total number of page reads by the KENP page count for the book. So my Last Mayor box set is 2600 pages long. If I only had 2600 pages read, that’d be one sale.
You add those numbers together – that’s your sales for a given period. I decied to go monthly. Then you get your total clicks from all your ads, and divide your sales by your clicks. It’ll come out as a decimal, but it’s easy to make that a ratio or fraction in your head.
Eg – Sold 100 books, had 1000 clicks, 100/1000 = 0.1 = 1:10. For every 10 clicks I paid for, 1 sold.
Next we can take the whole revenue and the whole cost for that same period, and figure out not only my CPC (Cost Per Click) but also my profit per click. I can end up with a number like, for every single click, I’m earning 50 cents, or something like that. Or losing 10 cents.
On conversion, Brian Meeks says 1 sale in 5-10 clicks is fantastic. 1 sale in 10-20 is OK but needs improvement. Above 1:30 is poor and likely to lose money.
So, I started gathering data. It took hours. Here’s my findings:
Last Mayor Box Set 1-9
I’m well aware this box set is making money. I’ve also become aware it’s been making less. A month or two ago it was ranked 3000 in the whole store and bringing in $150 a day. Now it brings in $80 or so a day. What gives?
It’s pretty straight forward. In December I made a big investment in Facebook ads, which I guerss I’d forgotten about. 1500 clicks at 22c a click, alongside another 1500 AMS clicks at 60c a click, for 3000 at 40c average. Phew. $1200 outlay.
So did it convert?
Well, there were 620 sales at 99c, not worth much, and 162 KU sales, which was worth a lot. That’s some 800 sales for 3000 clicks. Is that a good conversion rate?
Yes. It is slightly better than 1 in 4.
When I saw that I was floored. It couldn’t be right. Brian Meeks said any conversion better than 1:6 probably meant your data was off. So I checked again, ran it for Jan and Feb, but the conversion remained consistent. If anything, it got better, rising to 1:3. Each click is costing me 40c, and I’m getting anything from 70c to $1.20 back for every click.
So – why has my rank dropped? Why is income going down, if conversion is still high?
This is easy to see now. I spent $1200 in December, freaked out at the big Facebook ads bill, and turned them off. I spent nothing in Jan or Feb relying solely on Amazon ads, which came in about the same level – 1500 clicks each month. So I put in half the ad spend, got half the clicks, and made half the conversions.
The next step is to crank ad spend back up, see what happens. If conversions hold steady, this should be sweet.
The zombie news was nice, but the real reason I broke out the data was to figure out what’s going on with Wren. Why is he losing me money?
So I gathered the data – counting not only sales of book 1, but also 2 and 3. They are the readthrough generated by the ad, so it’s all good. First thing I see for December – is a pretty amazing conversion rate. What? 1:3?? I spent $700, got 1100 clicks evenly spread across Facebook and AMS, and sold 400 in total.
Wow, right? But wait, I’m still losing money? And as I mve through Jan to Feb, the conversion rate is falling? I end up in Feb with a conversion of 1:20.
What happened? That’s really bad.
Then I remember, right at the end of November I had my Bookbub. Readthrough and page reads from that would be coming in all December and even Jan, skewing all my ad results.
But then – even with the Bookbub tail, I’m STILL losing money? And once Bookbub is done, I’m really at 1:20???
I get depressed about that half the day, then just now at home realize I linked the Excel sheet wrong, and my Feb conversion is actually 1:10.
This cheers me up no end. 1:10 is good! It gives me the courage to go look at my readthrough, which I’ve also suspected is awful. I’ve had a number of reviews saying book 1 is so dark that people quit and read no more. I believed it.
But – my readthrough is pretty good! For KU reads 60% of people who read 1 go on to 2, and 40% go on to 3. I think that’s pretty standard. For units sold it drops off a bit more, 40% from 1 to 2, 30% to 3, but still – there’s maybe some influence of running a 99c price point.
Then it hits me – why I’m losing money. It’s not for lack of conversions or readthrough (although both could certainly be better) – it’s the pricing!
So obvious. At 99c, I’m tossing away $2 from the off.
I re-ran the analysis with $2.99 plugged in instead of 99c, and instantly I’m making a small profit.
Well. This is why they say you can’t make money on a single book – you need a series – preferably a long one. It’s the reason the zombie books work for me. I need more Wren books, basically. I also need, for now, to raise the price of book 1. It’ll depress sales, I know that, but it can’t just keep bleeding money like this.
So – $2.99. Hopefully, if conversions hang at 1:10, I can then spend more on ads. And of course, I need to get book 4 out, and a box set of books 1-3, and the audiobooks, which could add some fringe income.
I’ve analyzed data before – but perhaps never with such clarity, or provokng such revelations. I’m doing good. The books are doing good. It’s my loss-leader pricing and ad-spend aversion that have caused drop-offs and losses. In other words – business side, not creative side, and that’s both a lot easier to stomach and to fix. In fact – it’s done!
Wren 1 up to $2.99. Zombie FB ads cranked up. Now I’ll sit back and see what happens.