Amazon Ads keywords profitability standard

[LilyBLily]

I just had a thought about Amazon Ads and wanted to share it. All along, we've acknowledged that since Amazon does not show KU page reads matching specific keywords in our ads, we've simply been guessing about their effectiveness. We settled on the concept that any keyword showing clicks that result in purchases is probably also resulting in an equal number of KU reads. We have zero proof of this, of course. For those of us who have titles that sell very few copies, this is easy enough to parse as credible even so. We look at the KDP dashboard and see that the advertised title is getting KU reads equivalent to a read-through of the complete book.

But wait. If the advertised book is part of a series, then the KU royalty per keyword click is Book 1 page reads + Book 2 page reads +Book 3 page reads, and so on, ad infinitum, ad nauseam.

And most important, it means that the profitability of the keyword should be recalculated to reflect the KU pages read. An example: Once ACoS hits 70%, we are told, the keyword is no longer profitable. But considering the KU reads, probably the keyword could go to 100% before becoming unprofitable. Or even higher if you have a very long series and/or high sell-through.   

I think this means I should un-pause some keywords that went over 70%.

Someone tell me where/if this logic breaks down.

[Maggie Ann]

No, you are right. I believe Mark Dawson also subscribes to this theory.

Page reads should be counted as well as sell-thru. What weight they should be given, I have no idea, but I usually let my ads go well over 70% if I'm seeing sell-thru and/or page reads.

I'm not in KU anymore so I don't run AA ads. I feel the new iteration of ads favors KU books.

[DrewMcGunn]

I take it one step further. I disregard the ACoS altogether so long as my royalties exceed my AMS spend. Case in point, I had a sale on the AMS dashboard that correlates to my paperback price on one ad. I can't use the 70% rule as it is meaningless when someone buys a paperback. So, for those of us in KU, in addition to trying to figure out how to reconcile KU page reads, we also have paperback sales that can throw things off, in addition to read-through in a series. Also, throw in the probability that the data AMS is providing us may be incomplete into the equation, and the AMS report is IMO, pretty much crap.

Even though that's my take on AMS, I still want to make sure that I'm not wasting my marketing dollars. Mark Dawson and Brian Meeks have their own systems to calculate RIO. I've looked at Meeks' system and it's very complex. I do some basic math in a spreadsheet to see how my ads are performing. What I do is tally the impressions and clicks for a specific period as well as the spend for that time period. Then I take my royalties for the same period and measure them against the clicks to arrive at what I think of as my true ACoS.

It looks something like this:
Time period: Last 7 days
Impressions: 82,511
Clicks: 128
AMS Spend: $39.66
Royalty for the period: $268.87
advertised book sale count (included KU): 33 sales
Conversion: 33/128 or around 1 in 4.
RIO: 268.87/39.66 or just under 700%

There are some obvious weaknesses to this. First, I don't know how many people organically find my books. I could be throwing away money and not even realize it. Second, while this works for me (or rather, I think this works for me), it may not work for others.  It's just a quick and easy way for me to calculate return on investment for AMS.
 

[David VanDyke]

You can get 80% of the way with rules of thumb. One rule of thumb is to turn off or reduce the bid on your worst keywords, especially if they are above 100% (and you are in KU, above 70% if you are not). This will increase efficiency, but it also scales down your discoverability.

Read-thru is a huge factor in calculating overall ROI. As a rule of thumb, if you have at least 3 books in the series, and the read-through is above 70%, you can double your ACOS limits and still break even, especially if you're in KU.

If you have data from before you started using AA and after, you might be able to compare the befores and afters and get a rough idea of the ROI of the change. In other words, if (all else being equal) you spent $150 to gain $300 in ebook sales at 70% earning (overall increase from your baseline, that's the key, not what AA says), which nets you $210, you made about $60. Even if AA says your ACOS was (say) 110%, if your overall ebook sales and net profit increased, it's worth it. Some of it will be from read-through, and some will be echo effects of greater visibility.