Here is a screenshot of statistics for three of my AMS ads:
One is a non-fiction title, one is fantasy, one is romance. If you look at the far right-hand column, you can see that the non-fiction title has an ACoS of just 33%, the romance title has an ACoS of 255%, and the fantasy novel is at 123%.
If you were just judging the ads by ACoS you’d probably think “shut down the fantasy and romance ads”, right? I mean, if you assume a 70% payout then anything above that is losing money.
But not so fast…
There are a few things at play here that make the non-fiction ad perhaps not as impressive as it looks and the romance ad actually profitable. (In fact, all three are profitable.)
First, with respect to the non-fiction ad, AMS ads report both paperback and ebook sales. And for this particular title, paperback sales are a large portion of the sales reported. As nice as it would be to get a 70% payout on print sales, that just doesn’t happen. Which means for a title that sells predominantly in print you can’t use 70% as a benchmark. (Same goes for an ebook priced below $2.99.)
Instead, for print, you need to look at the list price of the book compared to your payout and calculate a percentage from that.
Of course, most titles aren’t clean in terms of print versus ebook sales, so ideally you’d then calculate a weighted average that takes into account approximately how much of your sales are print versus how much are ebook. (I give some examples of just how different that ratio can be in CreateSpace for Beginners. For non-fiction, I see a decent amount of print sales. For romance, I see almost none.)
So the first thing to realize about those three ads above is that the non-fiction ad is potentially not the most profitable ad on that list even though it has the lowest ACoS.
Now let’s look at the romance. With an ACoS of 250%+ it looks horrible. It looks like I’m taking a bath on that ad and just handing Amazon my money.
But another thing you have to account for with AMS ads is that they don’t include KU page reads. This particular title had about 500K page reads when it was in KU. When I crunched the numbers I found that I had about 3.3 full reads from KU for every sale. (For those of you who picked up Excel for Self-Publishers, I walk through how to do that calculation in there.)
This title was also part of a series. So when I sold book 1 with an AMS ad, a certain percentage of the time that also lead to a sale of book 2. Between KU reads and sellthrough to book 2 that ad was profitable even though it doesn’t look it.
The third ad, the fantasy ad also benefited from KU reads and sellthrough. In that case, for most of the time I was running this ad the book was the first in a three-book series with each book priced at $6.99. Now, in terms of borrows to buys, it didn’t reach the level of the romance novel. I was closer to 50:50 borrows to buys. If I’d had an ACoS of 250% on this ad I would’ve been losing substantial money. But 125% was still profitable.
Which is all an argument for judging your ads based upon the performance of the individual titles not some arbitrary number that Amazon chooses to display on its AMS dashboard.
Something else that’s not addressed in the AMS ACoS is that I am positive the non-fiction title has had sales due to its increased visibility from my running AMS that are not reflected in those numbers. (AMS are the only ads I’ve run on it on Amazon.)
And, also, I’ve noticed with the fantasy series that readers will often buy all three books at once but AMS will only count the purchase of book 1.
I will add, too, that you don’t need a lot of keywords to have a productive ad, you just need good ones. That non-fiction ad only has 66 keywords and I’ve paused some of those.
So anyway. There you have it. A real-life comparison of three very different titles and how they were each successful with AMS even though they look like they had very different outcomes.
M.L. Humphrey 
Glad to hear that you don’t need a lot of keywords. I bought KDProcket when it first came out and while it gives you lots of keywords, most aren’t a good fit for the book compared to good ol’ knowing what your book is about and maybe the authors best suited to your genre and not the whole gang of ghostwritten ones. So I may get lots of impressions but not many clicks. I also need to bid higher apparently LOL
LikeLiked by 1 person
I find that even keywords that eventually generate a sale but don’t have a good click ratio will slowly kill an entire ad, so I try to focus in on those keywords that have a good click ratio and also generate sales.
In terms of bidding, it’s really a question of strategy. On Kboards there are a number of authors who use AMS and bid low and have profitable ads, but for me I want enough sales to get those follow-on effects of natural visibility, too, and in my genres that does require higher bids. If you bid 15 cents in romance you may get a couple sales a month, but that’s probably it.
Also, keep in mind that an average click cost of say 35 cents is representative of a bid about twice that.
LikeLiked by 1 person
As far as the 35 cents click cost, that’s so true. I’ve gone as high as 50o cents (but now will go higher) and once, I put the decimal in the wrong place and didn’t realize it until a week later when my costs was sky high lol
LikeLiked by 1 person