FB & AMS ad mysteries – 2020 Writing Week 12

Mike Grist Weekly Writing Update, Writing 1 Comment

First off, the grand news – Wren book, Reparation, is finished in audiobook format. More on this when it releases. 3 months ago I had just 1 book in audio – The Last – pretty soon I’ll have 4, with 3 in a series. All the backlist is tight, produced, and polished.


Can I get the zombie box set back into profit? This is the mystery. If I can, then it will seem worthwhile to buy back the Last Mayor rights and make the whole series in audio. That’ll be about $10k – real money. It’s also a whole new audience – audio listeners love the zombie apocalypse, it seems.

Which brings me to the major challenge of today – figuring out what the heck has been gping on with my ads.

The FB & AMS ad mysteries

Earlier this week I posted in some detail about recent ad losses. The key takeaways were:

  • From Sept to Fed the Last Mayor box set was making strong profits with an ROI around 100% – doubling ad spend. Every dollar spent earned a dollar. CTR (clickthrough rate) was high. CVR (conversion) was between 1 in 3 and 1 in 5. This was on a varying mixture of FB & AMS ads.
  • In March I noticed the returns were reducing, and tried to juice them with ramped up ad spend on FB, AMS and Bookbub too. In return, I got punched. I started making losses.
  • Why?

At that point, I didn’t know. So I started recording daily stats, CVR, clicks/sales, clicks/KU, sales/KU and tried to spot the pattern. I tried FB only, then tried adding AMS back in. It’s only been a week, and it’s not statistically significant, but still I want to be guided by the stats as much as possible.

Facebook ads:

  • FB ads for the zombie set in both US and UK convert to sales pretty reliably at 1 in 4/5. At 15c a click or so, that’s 75c to make a sale. I only get 33c from each direct sale, so on this level, as expected, I’m losing money. So we turn to KU page reads.
  • The same FB ads, where I can separate their influence from AMS ads, convert to full KU readthroughs at anywhere between 1 in 30 to 1 in 100. That’s low. Compare it to the best KU conversion I got, in Jan when I was running only AMS ads, of 1 in 10, and it’s a long way off. Some figures:
  • For a full readthrough I earn about $12. 100 clicks at 15c costs me $15. Add the $12 to the money from sales (100/5=20×0.33=$7) and I get $19. Actually, that’s a profit. But that’s only if the clicks convert 1 in 5. In Feb I was getting 1 in 10. That works out to a $3 on 100 clicks. $12 plus $3 = $15, just what it cost me for those clicks.
  • Except, in Feb my average CPC (cost per click) was 43c. Not 15c. Because AMS and Bookbub are both much more expensive. So, major losses.

So – we can clearly see that if I can get 1 in 5 CVR on sales and even 1 in 1000 CVR on KU, I can make money as long as clicks are cheap at 15c. The only place I can get such cheap clicks though are on FB, and here’s what I found about FB ads:

  • Throughout this week I increased FB ad spend in both UK and US. Over the 5 days, US CVR sales peaked at 1 in 5 and CVR KU at 1 in 30, for 100 clicks total at 17c.
  • The cost of that was 0.17×100=$17.
  • The sales revenue was 100/5=20×0.33= $7
  • The KU revenue was 100/30=3x$12= $36. Total $43 – $17 = $26. An ROI above 100%. Good.

But that dropped as I increased US FB ad spend. I went for 150 clicks and sales CVR dropped to 1 in 7 and KU CVR dropped to 1 in 100, for 20c clicks! Loss-making territory.

What’s the conclusion here? Well, these are just single days during an unusual time, so I can’t be concrete – but it may be that trying to spend too much in a single day reduces the quality of FB clicks, while increasing the cost. So I should find the sweet spot of max clicks for cheapest CPC and best CVR, and not try to scale up any further.

How then do I get more KU page reads?

Amazon Ads:

I said above my best KU CVR number, full readthroughs to number of clicks, was 1 in 10 in Jan when I was running only AMS ads. So maybe AMS ads are the key to getting page reads. The only trouble is, AMS ads are so expensive, averaging around 60c a click in the US.

But if they sell at 1 in 10 levels?

  • Full readthrough = $12
  • 10 clicks at 60c/click = $6
  • 1 in 10 KU CVR means $12-$6 = $6
  • Great. But 1 in 10 was my max. Average seems to be about 1 in 20. Run it again, and we see that’s $12 cost. Breakeven.
  • Ugh. Crazy. How did I ever make money?

Hardly once this week has KU CVR risen above 1 in 50. If 1 in 20 is breakeven, 1 in 50 is clearly loss-making. Now, this week I’ve primarily been doing FB ads. Perhaps it makes sense that FB readers are less likely to buy through KU.

If I look back to Jan and Feb, when I ran only AMS and no FB, I was doing well with KU reads. Somewhere between 1 in 10 and 1 in 20 is profitable. Add to all this that KU reads in the US totally dropped off in the last few days, when I was only doing FB ads. Like, dropped in half, when it’s never done that before.

So, hmmm…


  • What is the use of FB ads? I can sell a lot of books at 99c, but I basically lose money. It can push my book into bestseller charts, but I’m not convinced such visibility is worth what it once was. There are too many charts, and people don’t peruse them like they used to.
  • I need to get into Amazon’s automatic emails, pushed to whale readers, and the only way to do that is to generate a lot of KU reads. The best way to generate KU reads, it seems, is to do a lot of AMS ads.
  • Which I was doing. For Jan/Feb I was running no FB ads at all. It was only after I saw the data and the power of FB clicks that I doubled down on running them for March.
  • I guess I overlearned the lesson.

My current thinking is that a little FB ad trickle, juicing straight sales and keeping the ranking respectable, benefits the Amazon algorithm on whether it shows my ads or not. That little trickle lets KU reads bloom.


My next step is throttle back the FB ads to a trickle, say $5 or $10 a day, then turn up the AMS ads again. I’ve got Machete to help me stay on top of things. Yes, OK. This is tomorrow’s experiment. I may lose my rank and bestseller tag on Amazon UK, but whatever. That’s vanity, not profit.

Another interesting step I could take would be to drop the crazy 99c price point, then double down on FB ads. Currently I get 1 in 5, say that drops to 1 in 10, clicks 20c, that’s $2 to make a sale, but I make $7 at $9.99, so $5 is mine. 100 clicks a day, 10 sales, that’s $50 profit. If I could also make AMS ads work at that price point, I’d be killing it. You’d think KU readers would see a $9.99 box set as better value for their limited borrows than a 99c one.

Hmm. Maybe this is next week’s experiment. My only concern is that I spend money on ads, send people to my page, then they get pulled down a rabbit hole of my competitors whose box sets are still at 99c. We shall see.

Comments 1

Leave a Reply

Your email address will not be published. Required fields are marked *