So, for the last month, I’ve been experimenting with Amazons Marketing Services. Or, for those of you who aren’t familiar with the term, what amounts to paying Amazon in exchange for Amazon running ads for your product based on keywords and the like. So if someone searches for, say, The Expanse on Amazon right now in the books section, Colony comes up, because the two are similar Science-Fiction.

After a month, I’m starting to see a few of the things I’ve been told about AMS confirmed. One of the reasons I’d avoided it until now was because my research into other authors trying it out came to the conclusion that it was basically a way to get advertising for your books … but to in turn make almost no money off of them, if not none. This because of the strange way Amazon runs its ads, and the system by which they do it.

See, how it works is you set a book to be advertised, followed by a per-keyword ad cost and a daily limit to how much you want to spend. So the keyword may be “action adventure.” You set a cost of 25 cents, and then a daily limit (say, a dollar).

Now what happens is that whenever someone searches for books with the keyword “action adventure” Amazon performs a “bid” for the highest paying ads for that keyword. The ones paying the most go up, and then if the viewer clicks them, it pays one cent more than what that bid beat—so, for instance if the 25 cent bid beat out a 22 cent bid, then it would pay Amazon 23 cents—and the viewer looks at the book, and that 23 cents is counted towards the daily limit.

A little convoluted, but not bad, right? Well … there’s a catch. There’s obviously a catch. See, as was pointed out to me long before I ever tried Amazon Ads, and one of the big criticisms leveled against them is that Amazon has more data on who buys what than the Ad service uses. It simply acts off of keywords, rather than Amazon’s own “We know you’ll like this” system. And so you may end up with clicks that lead to nothing at all quite frequently, because the person who search “Science Fiction” reads Foundation and Hyperion, not Colony. Amazon knows this, but they let the clicks go through anyway.

Maybe? Because here’s were things get weird. Of the accounts I’ve read, and now adding my month-long experience into the mix, for some strange reason Amazon Ads services almost always end up just slightly above “cost neutral” before the 30% Amazon take anyway. Right now, sales of Colony from the ads? The cost of the ads equals 94% of the take I’ve made … and that take doesn’t include Amazon’s 30% cut, so I’ve actually lost a little money selling more books. Definitely going to count that as a business loss on my taxes.

Now, I didn’t go into this blind. One of the bits of advice I had about using AMS is that you only spend money on it you can afford to lose. I’m just amazed at how  on the nose everyone’s analysis of AMS is. You will make earn pretty much exactly what you spend. And to be fair, AMS doesn’t seem to count people who come back to your book later and pick it up. Overall I’ve made more than I spent through this run.

But it is interesting. I’m not saying it’s a huge conspiracy by design of Amazon. Sure, they could be tailoring who sees those ads on the backend so that things keep at a near 100% margin the whole time, intentionally or unintentionally. But it could also just be the weird way their system works out, and Amazon’s fine with it reaching the balance. And by no means do I feel the money has been wasted. After all, more readers mean more reviews, which means more exposure, which means more eyeballs on the book, which probably accounts for a few of the sales that aren’t counted toward the AMS system.

That said, AMS doesn’t make it … hard, shall we say, to suspect some sort of weird back-end oddity going on. It’s amazing, for instance, how closely it sticks to 100% cost neutrality. The moment it starts getting either too high (ie, too much spent on advertising), there’s a sale that brings it back to the 90-100% range. And if it’s too low, sales will suddenly go dry despite there being clicks for a nice long period. But is it anything deliberate? Well, I’d say no, if it’s even intentional at all. If anything, it’s probably just one of those modern consequences of the way computer systems try to level things out and reach conclusions on their own, and in this case it levels everything out at 100%.

Ultimately, it is a bit odd that I have to pay Amazon to advertise something that, by nature of their business, you think they’d be advertising anyway, except that you have to pay for it. And I must add my experience to those who have also tried AMS in that if you’re thinking about it, go into it with money you can afford to lose, and expect that you’re going to come out of it with new sales … but little to no actual profit. Which is funky.

Is it worth it to you? Well … you get eyeballs and possible future readers you won’t have to advertise for.

But it’s still a strange system.

In other news, Shadow of an Empire editing is going well, and I’m negotiating a cover with a talented artist. More to come! Oh, and thanks to a work shift, BaBW will be Tuesday again.