Amazon Advice

[Demesnedenoir]

30%+ spike in impressions in an ad set yesterday, and that over its previous high, and I didn't touch the ad. Only thing I can think of is a relevance kick on the part of the algorithm... or I have a psychic connection that alerted Amazon I was working on more ad sets to shoot for 100k impressions in a day, LMAO. I've attempted to contact an Amazon guru for a quick opinion on this, we'll see if I get an answer.

I think these threads are cathartic, I can spew the statistical mysteries from my head so they don't drive me crazy.

 

[Prince of Spires]

Keep the thread going. I enjoy the ramblings. It gives me some pointers for when I'll start up ads again

I'm putting my money on the psychic connection. Everyone knows the Evil Amazon empire works on mind control. Perhaps it works both ways.

It could also just be as simple as other people lowering bids or stopping ads, which raises the position of your ad without doing anything. Or indeed Amazon noticing your ads convert well which makes them more attractive to show.

How many sales are you aiming for with the 100k impressions?

 

[Demesnedenoir]

I have no expectations outside of 50-100k impressions in a day. My sales have absolutely tanked from FB ads after the latest changes so I'm going into this with a "click" mentality more so than sales. I've been seeing between 15-20 clicks per 10k impressions for a rough guesstimate, so I'd expect 75-100 for a 50k splash. Sales? I'm not even going to wager a guess on this shotgun approach, I'm looking more to find successes and failures faster and see what can be done from there.

I expect FB to sort itself out as it has done in the past, but if not, I figure it's time to learn Amazon, heh heh.

Still a bunch of keyword mining to go before I fire this up.

Keep the thread going. I enjoy the ramblings. It gives me some pointers for when I'll start up ads again

I'm putting my money on the psychic connection. Everyone knows the Evil Amazon empire works on mind control. Perhaps it works both ways.

It could also just be as simple as other people lowering bids or stopping ads, which raises the position of your ad without doing anything. Or indeed Amazon noticing your ads convert well which makes them more attractive to show.

How many sales are you aiming for with the 100k impressions?

 

[Ned Marcus]

I think these threads are cathartic, I can spew the statistical mysteries from my head so they don't drive me crazy.

A good thread. Helps me understand ads a bit more. I'm still working on Amazon ads. I've given up on Book Bub and FB for now.

 

[Demesnedenoir]

BB in my opinion is specific use. Bargain shoppers can be useful. Heck, if I had 10 books ina. series out I might do a freebie on BB for book 1, but or regular sales? No. FB used to be good after figuring out what works for a particular book, but my pessimism grows for its current and future states as things keep changing. Amazon's opacity, which also held back FB's and others' functionality, and its basic advertising strategy, make it difficult as heck to make work for you. And yet, people clearly do. From what I can figure it's a matter of having all the ducks in a row and then surpassing a particular threshold in getting the algorithm to work for you and then spending to keep that mutual relationship alive. One big question I have is whether for indie KU is necessary to strike and maintain that relationship, or if a wide book is still capable of maintaining high rankings without being a major trad release or cultural phenom. Almost all indie books that stick in the Algo are KU, but they can stick for years, which is impressive.

One "Guru" who was flashing impressive stats on a free marketing video (trying to sell me, of course) was much less impressive once looking at the "now" of their books. I flipped over and their rankings were basically 1 sale per day on a $0.99 book, despite the impressive #13 in an obscure category, heh heh. Bupkiss, in other words. That said! They did have books with 3000+ reviews that were at one time a big deal, so they aren't without the ability to get books into people's hands, but their staying power was exaggerated in their promotion.

 

[Demesnedenoir]

Thinking further, I had a conversation with a gent a couple of years back—right before Covid— when he worked for a major PR firm in NY, and he headed up their book division. This will be a paraphrase, but some discussion on what I would like to achieve with my next release, and I kind of just put the generic question to him: What do you think it takes to succeed and that's what I want to do... and his answer was nonspecific, as you might suspect. He wouldn't even give me 1 single thing, it was more a matter of just surpassing a mysterious threshold to enter the public consciousness, from ads to blogs, to radio interviews, to, etc. etc. That was the goal of a PR campaign, and by all means necessary.

So, Amazon probably functions in much the same way, reaching both the public consciousness and the algo's 1's and 0's. That said, the Algo functions not just by Amazon Ads, all other ad streams feed the Algo, but that feeding is getting trickier as FB gets further limited, which reduces the ability to get Algo's attention and hence the public consciousness.

Most of my attempts have been assaults on the Algo from the outside, now it's time to see what happens from the inside, heh heh. Then, build strategies for the next release.

 

[Prince of Spires]

I don't think it's necessary to be in KU to become sticky in the algo's, but it certainly helps. And the reason is simple. Many KU readers only (or mainly) read KU books. And many KU authors have 2 KU reads for each sale (it varies, but that seems fairly average). And that does 2 things:
1. it improves your ranking, since borrows count the same as sales in ranking. So you get 3 "sales" instead of one
2. it makes your ads more profitable, since you have an easier time converting people. More profitable ads means you can bid higher and run more of them, which means even more sales, which means even more algo love.

As for what works to sell books, in the end it all comes down to getting people to see your books. More eyes on your books, more sales. Some eyes convert better than others, which gives different ratios for how many eyes you need for a sale, but that doesn't change the idea. For instance, your newsletter might get you a sale for every 10 people reading it. While for an Amazon add you might be happy with 0.1% clicks and 10% conversion, which means 1-in-10.000 eyes converts. And maybe 1-in-every-100.000 TikTok views gets you a sale, while friend recommendations convert at 1-in4, and so on.

The trick then is to "simply" figure out what is profitable and sustainable for you, and do that.

 

[Demesnedenoir]

And hope that source doesn't change, heh heh. My ink went red with FB whatever the heck is going on. There's still some discussion as to what is actually going on, as people are getting some wild fluctuations in CPM rates.

I don't think it's necessary to be in KU to become sticky in the algo's, but it certainly helps. And the reason is simple. Many KU readers only (or mainly) read KU books. And many KU authors have 2 KU reads for each sale (it varies, but that seems fairly average). And that does 2 things:
1. it improves your ranking, since borrows count the same as sales in ranking. So you get 3 "sales" instead of one
2. it makes your ads more profitable, since you have an easier time converting people. More profitable ads means you can bid higher and run more of them, which means even more sales, which means even more algo love.

As for what works to sell books, in the end it all comes down to getting people to see your books. More eyes on your books, more sales. Some eyes convert better than others, which gives different ratios for how many eyes you need for a sale, but that doesn't change the idea. For instance, your newsletter might get you a sale for every 10 people reading it. While for an Amazon add you might be happy with 0.1% clicks and 10% conversion, which means 1-in-10.000 eyes converts. And maybe 1-in-every-100.000 TikTok views gets you a sale, while friend recommendations convert at 1-in4, and so on.

The trick then is to "simply" figure out what is profitable and sustainable for you, and do that.

 

[Demesnedenoir]

Still only 2 ad sets, 22k+ impressions and 40+ clicks today, so something is beginning to click. Sales aren't following like I want, but... baby steps toward people seeing the book enough times to buy it, heh heh. Plus, Amazon tends to contact people over things they've clicked on. So... we'll see.

Still building the keyword sets for my 100k push, but 100k appears more likely than before, having hit 22k+ with almost 3 hours left in the Amazon day.

 

[Demesnedenoir]

Right now, I am going against the advice that started this thread. I am slamming piles of keywords and letting the writing gods sort them out.

I haven't had time to fully flesh out keywords in 8 ads, still on two ads, but click rates have improved a good deal while sales are beginning to move a little better. Outside of 1 day where I had insane impressions and clicks (Saturday) BUT no known sales* the impressions have remained steady, the clicks are trending up, and sales are beginning to look steadier than I've had before, and I've gotten a couple of sales from author names I have never heard of. If sales go steady, efficiency will be the next key. Okay, after building my other keyword sets and testing for about 2-3 months to gather data, then I'll work at efficiency.

Promising so far.

* — Nothing shows in Amazon, but a few audion sales likely dated to Saturday.

 

[Demesnedenoir]

I have not done the Amazon big push yet, time has been short, and I've decided to try a little FB switcheroo to see if it functions better than what's been going on before said big push. That said, I've decided that Amazon ads make for strange shelf fellows, instead of bedfellows, for books. I would not have guessed that Stephen King's Dark Tower series would be a click magnet or sales maker for me.

Color me dumbfounded. Stephen King and Daniel Abraham, nope, I wouldn't have guessed that at all.