I advertise, but not with AMS. I don't trust it. There is no way to know if what they are telling you about clicks and buys is true. I tried a few ads at the beginning of their targeting it to us, and they claimed I had sales, but I turned off the ads and there was no fall in sales. So color me suspicious. For years, people have been handwringing about "what if Amazon lowers our royalty?" With AMS, now they don't need to. In effect, they have lowered royalty by handing authors a $250 or $2500/month AMS charge. That good for stockholders (which they mention to stockholders in their reports). Not so good for me.
FB was caught exaggerating its ads' clicks and return a few years ago, though I know of too many people earning six figures who use them as their main advertising to entirely discount their efficacy. Still, the effort to get the image and wording just right to maximize returns will cost you plenty--so there had better be the potential for a big return before you do that. (The big return means you need more than a few books for sale other than the advertised one.)
I don't advertise often, and I stick with newsletter ads. That way, I don't turn on some automatic ad system that will drain my account without my stopping to analyze the expenditure every week. With one-shot ads, I can test and see which sites give me what ROI, and I know the break-even sites (which may win me fans for my back list, so is still worth trying now and again) and those who have given my a positive ROI. However, every time you advertise a book, the return on an ad at any given site will drop. (I know marketing gurus say that increased runs of an ad help, but I certainly haven't seen that to be true in the book biz, and as a customer, I've turned off newsletters for advertising the same books too often.) So I only run an ad for any first in series one week each year. That seems to be most effective and saves me a lot of time and hassle in focusing on promotions rather than on creating more product to sell. I suppose one could use AMS like that, running only a three-day ad twice a year, but you'd still have no idea if their reporting clicks/buys to you was fact or fantasy, and so you cannot compute an accurate ROI...and to me, that seems like poor business practice on my part. (If a rather clever one on their part. We're their customers for AMS, and they are making a good deal of money off us.)
The best "advertising" is always to write more books, in series, and buy good covers for them, indistinguishable from trade publishing covers. Periodically run an ad for the first in every series--never for stand-alone books. The more books you have, the more series, when you find a new fan through an ad, the more good that ad will do you as your new reader buys her way through your back list. Develop your own mailing list so that every new book gets you a guaranteed chunk of sales and increased free visibility with a high ranking for its first few weeks of release.
I didn't run ads until I had several books out, and I've made my living as a self-published novelist for several years now. Most people who do earn a living this way have 30+ books out. So that's where to focus the vast majority of my time, attention, and money, I've concluded. More books, not more ads.
Good luck to you.