Author Topic: AMS Ads Learning  (Read 88682 times)  

Offline LilyBLily

  • Status: Scheherazade
  • *****
  • Posts: 1321
    • View Profile
Re: AMS Ads Learning
« Reply #1400 on: July 15, 2017, 01:57:34 PM »
Somewhere here on KBoards there's a comment naming the Amazon algorithm. Sorry I don't know how to find it.

Offline Colin

  • Status: Scheherazade
  • *****
  • Posts: 1444
  • Gender: Male
  • UK
    • View Profile
Re: AMS Ads Learning
« Reply #1401 on: July 15, 2017, 02:02:37 PM »
Somewhere here on KBoards there's a comment naming the Amazon algorithm. Sorry I don't know how to find it.

I believe it could be Algorithm #666...

Offline weigle1234

  • Status: Lewis Carroll
  • **
  • Posts: 125
    • View Profile
Re: AMS Ads Learning
« Reply #1402 on: July 15, 2017, 02:09:32 PM »
Somewhere here on KBoards there's a comment naming the Amazon algorithm. Sorry I don't know how to find it.

Probably obscene - not printable on most forum without moderator censorship.

Online khotisarque

  • Status: Dr. Seuss
  • *
  • Posts: 30
  • Gender: Male
  • California
    • View Profile
    • Khoti Sarque
Re: AMS Ads Learning
« Reply #1403 on: July 15, 2017, 09:12:14 PM »
I believe it could be Algorithm #666...

Or love potion #9?  But certainly not the First Law of Robotics.   :D

Fine and Nifty: 80%

Barebones: 5%
Moderation in all things, especially in moderation.
khoti sarque | blog

Offline Decon

  • Status: A A Milne
  • ******
  • Posts: 4104
    • View Profile
    • Declan Conner, Blog
Re: AMS Ads Learning
« Reply #1404 on: July 16, 2017, 11:37:18 AM »
Hmm, early days, but since increasing my prices to $4.99 and the bids on two of my book's keywords. I've had 2 AMS sales,  and from no page reads to 1000 page reads.

The cost was $1.69 for the clicks

$4.20 for the page reads

$6,80 for the royalties.

So that's $11 earned - cost $1.69 =  Profit $9.31

Not much faith it will continue, but promising so far.


Scorpion ebooks: Full length  thrillers with a sting in the tale. All enrolled in KU & Prime.
Declan Conner | blog

Offline bberntson

  • Status: Jane Austen
  • ***
  • Posts: 375
  • Gender: Male
  • Boulder, Colorado
    • View Profile
    • brandonberntson
Re: AMS Ads Learning
« Reply #1405 on: July 16, 2017, 06:20:46 PM »
Hmm, early days, but since increasing my prices to $4.99 and the bids on two of my book's keywords. I've had 2 AMS sales,  and from no page reads to 1000 page reads.

The cost was $1.69 for the clicks

$4.20 for the page reads

$6,80 for the royalties.

So that's $11 earned - cost $1.69 =  Profit $9.31

Not much faith it will continue, but promising so far.


That's encouraging. I was doing some weird goofy stuff like pausing my ads until about 5pm then letting them run until Midnight, and I noticed I got some major page reads in the evening, along with some sales.  That has since dwindled of course, but it was an interesting experiment. 

Brandon Berntson | Blog | website | Facebook | twitter

Online Jena H

  • Status: Edgar Allan Poe
  • *******
  • Posts: 5551
  • North Carolina
  • Desperate character
    • View Profile
Re: AMS Ads Learning
« Reply #1406 on: July 16, 2017, 07:22:01 PM »
I had paused an ad about July 8, then when I realized Amazon Prime Day was July 11, I unpaused it the day before.  Lo and behold my AMS dashboard is crediting the ad with a sale on July 12.   :P  I think I paused the ad again (re-paused) on that same day, the 12th.  It's a hardcover sale, which was my goal with the ad.

I can't say that I've learned anything that will be of use, but hey, a result is a result.  (In fact, as Mythbusters will teach us, even the lack of result is a result.)

Jena

Offline Author A.C. Salter

  • Status: Madeleine L'Engle
  • **
  • Posts: 51
  • Gender: Male
  • Cotswolds UK
  • Author of Eversong, Shadojak and Ethea
    • View Profile
    • acsalter.com
Re: AMS Ads Learning
« Reply #1407 on: July 17, 2017, 10:49:12 AM »
Even though my books are urban fantasy, there is a strong romantic thread through the story. So 1 of my keywords is 'romance'. Since beginning AMS in May, I've had plenty of impressions but no clicks on that word and was about to pause it. However, several days ago I had a sale and when I checked, I had a click and a sale for romance and I'm chuffed to buggery. It's a low bid yet I'm reluctant to put it up incase it alters the algorithm.

Maybe if you wait long enough, the algo eventually sweeps you up for a bit.


A.C. Salter | blog | website

Offline Harald

  • Status: Jane Austen
  • ***
  • Posts: 277
  • me gusta nadar
    • View Profile
    • HaraldJohnson.com
Re: AMS Ads Learning
« Reply #1408 on: July 17, 2017, 03:08:44 PM »
Even though my books are urban fantasy, there is a strong romantic thread through the story. So 1 of my keywords is 'romance'. Since beginning AMS in May, I've had plenty of impressions but no clicks on that word and was about to pause it. However, several days ago I had a sale and when I checked, I had a click and a sale for romance and I'm chuffed to buggery. It's a low bid yet I'm reluctant to put it up incase it alters the algorithm.
Maybe if you wait long enough, the algo eventually sweeps you up for a bit.

"chuffed to buggery" eh? Ah, you Brits! :)

I've had sales on super-general keywords like "fiction" so if your bids are low and you're not getting many clicks on 'romance' then not much of a damp squib, is it?
   
Fiction starting with The Manhattan Series novellas (1609+). Older work is non-fiction.
Website | FacebookBook Cover Reviews discussion group

Online khotisarque

  • Status: Dr. Seuss
  • *
  • Posts: 30
  • Gender: Male
  • California
    • View Profile
    • Khoti Sarque
Re: AMS Ads Learning
« Reply #1409 on: July 17, 2017, 04:44:49 PM »
"chuffed to buggery" eh? Ah, you Brits! :)

I've had sales on super-general keywords like "fiction" so if your bids are low and you're not getting many clicks on 'romance' then not much of a damp squib, is it?

Hey, keywords are free.  Why not try "chuffed" and "buggery" and "damp" and "squib".  Maybe not too many impressions unless a cult forms ;-) but those that do match might give 100% sales-to-clicks-to-impressions ratios.   8)
Fine and Nifty: 80%

Barebones: 5%
Moderation in all things, especially in moderation.
khoti sarque | blog

Offline LilyBLily

  • Status: Scheherazade
  • *****
  • Posts: 1321
    • View Profile
Re: AMS Ads Learning
« Reply #1410 on: July 17, 2017, 05:18:54 PM »
Had to look both idioms up. Learned something.

I just paused an ad that had gone over the 70% mark, and changed keyword bids on another to maximize its efficiency by dropping the bids on nonperforming keywords. I'll be interested to see if that reduces sales at all, or brings my ad cost per month down, or whatever. It's all a game, but at least by doing these ads I'm in the game instead of watching on the sidelines.
 

Offline Harald

  • Status: Jane Austen
  • ***
  • Posts: 277
  • me gusta nadar
    • View Profile
    • HaraldJohnson.com
Re: AMS Ads Learning
« Reply #1411 on: July 17, 2017, 05:29:35 PM »
Had to look both idioms up. Learned something.
...
It's all a game, but at least by doing these ads I'm in the game instead of watching on the sidelines.

Right on! (and another idiom  ;)
   
Fiction starting with The Manhattan Series novellas (1609+). Older work is non-fiction.
Website | FacebookBook Cover Reviews discussion group

Offline Author A.C. Salter

  • Status: Madeleine L'Engle
  • **
  • Posts: 51
  • Gender: Male
  • Cotswolds UK
  • Author of Eversong, Shadojak and Ethea
    • View Profile
    • acsalter.com
Re: AMS Ads Learning
« Reply #1412 on: July 17, 2017, 07:13:15 PM »
"chuffed to buggery" eh? Ah, you Brits! :)
Yeah sorry - at least I didn't use military terminology.


A.C. Salter | blog | website

Offline LilyBLily

  • Status: Scheherazade
  • *****
  • Posts: 1321
    • View Profile
Re: AMS Ads Learning
« Reply #1413 on: July 18, 2017, 09:42:10 AM »
Another thread got started asking about romance keywords, and this is what I posted on it:

"My romance keywords that deliver the most sales are bid at $0.46 and $0.51 and they win their spots at a low of $0.23 and a high of $0.33. That's to keep them on page 1 of any carousel and on page 1 of the in-line ads.

Bids at $0.25 often win the auction at $0.15, but those ones usually sell only one or two copies. While they may possibly generate page reads (not reported on AMS), I consider those keywords losers and after an arbitrary period of time I reduce my bids on those words to under ten cents each. I don't kill them if they've earned me any money. Ad keywords that deliver neither impressions nor clicks nor sales are ripe for killing. Ads with impressions and clicks but no sales are iffy. Did they generate page reads? No sure way of knowing."

My cumulative cost for these ads this month since the last invoicing averages $14.59 a day, which is the most I can tolerate, but also may include a little bit of extra from my $45 dollar two-day budget experiment. Now I simply have to be patient, and let these ads do their job. The problem with that is fiddling with the ads produces improved results, so I'm constantly fiddling. And that leads to paying too much attention to these ads.  :-\



Offline Laura Rae Amos

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 858
  • Gender: Female
  • Metro DC
    • View Profile
    • LauraRaeAmos.com
Re: AMS Ads Learning
« Reply #1414 on: July 19, 2017, 05:19:34 PM »
I just can't get Amazon to take any of my money. I have my best keywords set to $0.25-$0.45. My daily spend is $10, but the ad only uses up pocket change, at most. My sales to clicks is good, ACoS 33%, but I'm just not getting many impressions most days. (~1000 a day over the whole set of keywords) It's kind of hard to get clicks if they're not showing my ad to anyone. I've tried pausing and unpausing, but the ad is just crawling. Most of the previous ads I've tried would at least get an initial burst of impressions, but this one has been slow crawling since it was born. Using the same keywords that performed well for previous ads.

How many impressions a day are you guys getting?

I'm not in KU, by the way, if that makes a difference.
« Last Edit: July 19, 2017, 05:22:16 PM by Laura Rae Amos »

Laura Rae Amos writes kooky and emotional relationship dramas, stories that are usually witty, sometimes poignant, and a little bit crass. Visit me at LauraRaeAmos.com // Tweets @LauraRaeAmos // Or chat on Facebook

Offline LilyBLily

  • Status: Scheherazade
  • *****
  • Posts: 1321
    • View Profile
Re: AMS Ads Learning
« Reply #1415 on: July 19, 2017, 06:01:11 PM »
I just can't get Amazon to take any of my money. I have my best keywords set to $0.25-$0.45. My daily spend is $10, but the ad only uses up pocket change, at most. My sales to clicks is good, ACoS 33%, but I'm just not getting many impressions most days. (~1000 a day over the whole set of keywords) It's kind of hard to get clicks if they're not showing my ad to anyone. I've tried pausing and unpausing, but the ad is just crawling. Most of the previous ads I've tried would at least get an initial burst of impressions, but this one has been slow crawling since it was born. Using the same keywords that performed well for previous ads.

How many impressions a day are you guys getting?

I'm not in KU, by the way, if that makes a difference.

On my lead ad, in the last two days, nearly 35k impressions. On the other, 20k.

I have had the same trouble with new keywords: Sometimes they don't result in my ad being shown at all, and I don't know why. Typically this is when I add a new author name I've gotten from an ad. You'd think that author name would be popular within a short period after the ad runs, and would attract readers because it's probably discounted. But I get no traction.

As to being in KU making a difference in impressions, doubtful. One of my books is in KU and the other isn't; one is in a popular romance subgenre and the other is nonfiction.

As my recent experiment proved yet again, AMS isn't interested in more money thrown at the daily spend. A catchy new keyword might help boost impressions, but so far, I haven't found the secret to what those keywords might be.

Online Anma Natsu

  • Status: Arthur C Clarke
  • *****
  • Posts: 2442
  • Gender: Female
  • Texas
    • View Profile
    • Anma Natsu
Re: AMS Ads Learning
« Reply #1416 on: July 19, 2017, 06:36:10 PM »
My first ad, averaged 8,214 impressions a day, and got about 1 click per 1,065/impressions or so.

Second ad, run a few days later and with more keywords, the average dropped to 3,272/day, but the click rate went up to about 1 click for every 954 impressions.

Still working on another round of keyword expansion to try round three.

I had all mine set to 25 cent rates for the keywords.

Online BillyDeCarlo

  • Status: Lewis Carroll
  • **
  • Posts: 121
  • Gender: Male
    • View Profile
    • Billy DeCarlo, Author
Re: AMS Ads Learning
« Reply #1417 on: July 20, 2017, 04:46:08 AM »
I just can't get Amazon to take any of my money.

They have no shortage of people throwing money at them. That's one problem. I think their algorithm factors in how good the books are (probably based on number and average review score, and how many people finish reading it), and how 'saleable' they feel the book is by other factors (rank in genre, etc) , how popular you are as an author (number of backlist or series titles, author page following on amazon.com) before they place the ad. Remember, your ad money at so many cents per click *if* someone clicks it, is a pittance compared to the revenue they get if someone buys a book that they are placing.

So that's the primary goal of the algorithm, and what I'd be looking at if I was writing the code.
Vigilante Angels Book I: The Priest: 100%

Vigilante Angels Book II: The Cop: 95%

Vigilante Angels Book III: The Candidate: 1%
Debut Novel Available: Vigilante Angels Book I: The Priest. Book 2 available for pre-order.
Billy DeCarlo | Website/blog | Facebook | Twitter | Google+ | Amazon | BookBub | Goodreads | LibraryThing

Offline Cassie Leigh

  • Status: Scheherazade
  • *****
  • Posts: 1082
    • View Profile
Re: AMS Ads Learning
« Reply #1418 on: July 20, 2017, 05:51:06 AM »
They have no shortage of people throwing money at them. That's one problem. I think their algorithm factors in how good the books are (probably based on number and average review score, and how many people finish reading it), and how 'saleable' they feel the book is by other factors (rank in genre, etc) , how popular you are as an author (number of backlist or series titles, author page following on amazon.com) before they place the ad. Remember, your ad money at so many cents per click *if* someone clicks it, is a pittance compared to the revenue they get if someone buys a book that they are placing.

So that's the primary goal of the algorithm, and what I'd be looking at if I was writing the code.

I think they do look at momentum, but I don't think the factors you've listed matter to them.  My most successful ad the book only has 11 reviews and that ad would spend $40+ a day every day if I let it.  That pen name only has two titles out and I doubt it has any Amazon followers.  It's in contemporary romance which is highly competitive so I don't think it's on any of the top lists anywhere.  What got that ad moving was a strong free run that led to a number of borrows in a short period of time and then AMS have allowed me to keep it at the rank it achieved during that free run.
8 Pen Names.


Genres: Non-fiction, Speculative Fiction, Romance.


Only NF titles under Cassie Leigh and M.L. Humphrey.

Offline Laura Rae Amos

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 858
  • Gender: Female
  • Metro DC
    • View Profile
    • LauraRaeAmos.com
Re: AMS Ads Learning
« Reply #1419 on: July 20, 2017, 08:56:00 AM »
What got that ad moving was a strong free run that led to a number of borrows in a short period of time and then AMS have allowed me to keep it at the rank it achieved during that free run.

I think you might be onto something with this theory. The only time I ever had an ad spend close to its limit was when I did a 99c promo and the book was selling better than normal. That ad still gets more impressions than my other (the slow ad, for a different book) even after the fact. Maybe I'll run a promo on the the book with the slow ad and see if that makes a difference. Kind of dumb that you have to get your book selling on your own first before the ad will work to sell your book, lol!

Laura Rae Amos writes kooky and emotional relationship dramas, stories that are usually witty, sometimes poignant, and a little bit crass. Visit me at LauraRaeAmos.com // Tweets @LauraRaeAmos // Or chat on Facebook

Offline Steven Kelliher

  • Status: Lewis Carroll
  • **
  • Posts: 134
  • Boston
    • View Profile
    • StevenKelliher.com
Re: AMS Ads Learning
« Reply #1420 on: July 20, 2017, 09:32:50 AM »
My random experience with PD ads taking off like rockets. Some here may have advice or find it useful.

So ... I've recently been experimenting with PD ads more than SP. I find that when PD ads take off, they're WAY more effective than SP for me personally, even though everyone has told me not to bother with them.

That said, AMS's reporting delays have made it really difficult to scale ads and avoid overspending.

What I did: Started 16 PD ads on July 1 with low bids (11c). Those ads sat for two weeks and generated zero reported impressions across the board. Setting: "Run as quickly as possible."

I then went into the ads and adjusted the bids up to 26c across the board. Three days later, impressions ROCKETED. Thousands and thousands, and by the time I saw those reported numbers, I had spent $300 across 16 ads. I jumped in in a panic and adjusted them down to 16c bids, but with clicks still coming in furiously at the new bid (I'm now over $450 in spend this week alone,) I adjusted them down again to 12c on Tuesday. The data that I'm getting in today is still displaying 16c spending, so I'm not sure yet if the 12c bids are spending or not.

Book sales did shoot up, but not enough to account for the spend at the 26c rate.

That said, this whole experience has given me a weird theory that I sort of tricked the system into kicking off my two-week old ads by adjusting/increasing bids, and then when I scaled back down, they kept going. It's almost like once the ads have kicked off, the bids don't matter as much, but the bids may have to be higher to actually catch the attention of the system.

Maybe the system would have eventually kicked the ads off at the 11c bids, but according to my reporting, not a single impression ever came from those low bids. Only when I scaled up did I get impressions/clicks.

Either way, the reporting delays (3 days on average) is really pretty ridiculous.

Steven Kelliher | Web site | Twitter

Offline Cassie Leigh

  • Status: Scheherazade
  • *****
  • Posts: 1082
    • View Profile
Re: AMS Ads Learning
« Reply #1421 on: July 20, 2017, 12:09:00 PM »
I decided to set up a new ad for my fantasy novel that has about 5 million impressions across all of its ads at this point and for the first time ever I saw that they were actually suggesting some author names as keywords for that ad.  Usually their suggestions are such helpful things as "edition" and "trilogy".  Nicely enough their list of author names coincided pretty closely with my list of top authors who've generated sales for me with one exception which was an author who'd been in my also-boughts at one point but had never shown sales on my AMS reports.

I checked a few other books and my romance novel that has about 2 million impressions across its ads was also showing some author names, but another one of my books that has far fewer impressions across its ads didn't have any author names suggested.

For those who have ads with a lot of impressions built up who are in KU it may be worth looking at to see which authors are listed in case any of your no sales but lots of clicks authors are on there.

Also, the latest notice email I received said it could take as long as 3 business days to review my ad.
8 Pen Names.


Genres: Non-fiction, Speculative Fiction, Romance.


Only NF titles under Cassie Leigh and M.L. Humphrey.

Offline Cassie Leigh

  • Status: Scheherazade
  • *****
  • Posts: 1082
    • View Profile
Re: AMS Ads Learning
« Reply #1422 on: July 20, 2017, 12:14:49 PM »
My random experience with PD ads taking off like rockets. Some here may have advice or find it useful.

So ... I've recently been experimenting with PD ads more than SP. I find that when PD ads take off, they're WAY more effective than SP for me personally, even though everyone has told me not to bother with them....

When I was working on the AMS book I started a PD ad for my beginning writer book and set it to spend as fast as possible and it took off and spent something like $50 on its first day, but I didn't see an equivalent level of sales to match the spend so shut it down.  It took some time to shut down, too, so cost me a pretty penny by the time all was said and done.  I think PD ads are a way to get a lot of impressions fast, and one of my most successful ads on a romance novel (over a year ago) was using them, but I think they can be much more costly to use because they spend so quickly and don't shut down immediately.

I've also noticed with the handful I've run recently that they won't run at my original bid and I have to keep pushing it higher and higher until they kick in which sounds like what you found as well.  It's sort of feast or famine with them.  I either have no impressions or tons of them.
8 Pen Names.


Genres: Non-fiction, Speculative Fiction, Romance.


Only NF titles under Cassie Leigh and M.L. Humphrey.

Online khotisarque

  • Status: Dr. Seuss
  • *
  • Posts: 30
  • Gender: Male
  • California
    • View Profile
    • Khoti Sarque
Re: AMS Ads Learning
« Reply #1423 on: July 20, 2017, 11:47:58 PM »
They have no shortage of people throwing money at them. That's one problem. I think their algorithm factors in how good the books are (probably based on number and average review score, and how many people finish reading it), and how 'saleable' they feel the book is by other factors (rank in genre, etc) , how popular you are as an author (number of backlist or series titles, author page following on amazon.com) before they place the ad. Remember, your ad money at so many cents per click *if* someone clicks it, is a pittance compared to the revenue they get if someone buys a book that they are placing.

So that's the primary goal of the algorithm, and what I'd be looking at if I was writing the code.

Putting aside the philosophical question as to whether the ad-carrier's staff should be making value judgments on which products to push, I queston whether the ad revenue stream is really a pittance compared to sales margin.  On a $2.99 book, Amazon takes $0.90 off the top as its share.  At a supposedly typical 10 clicks per sale, and for example an acpc of $0.10, the AMS revenue exceeds the primary margin.  At the suggested bid level of around 25 cents, the acpc is likely to be in the 12 to 15 cents range, giving AMS perhaps $1.20 to $1.50 per sale.  More than a pittance.  At higher book prices, the situation may be reversed, but the ad stream is still significant.

It is entirely possible that the most efficient way to use AMS, from an author's point of view, is to bid low and seek exposure without clicks; background passive brand awareness at zero cost.  Then run non-AMS promotions in front of a potential readership who have at least seen your name somewhere before (like on page 77 of a sponsored product carousel).   :D
Fine and Nifty: 80%

Barebones: 5%
Moderation in all things, especially in moderation.
khoti sarque | blog

Offline LilyBLily

  • Status: Scheherazade
  • *****
  • Posts: 1321
    • View Profile
Re: AMS Ads Learning
« Reply #1424 on: Yesterday at 05:45:16 AM »
It is entirely possible that the most efficient way to use AMS, from an author's point of view, is to bid low and seek exposure without clicks; background passive brand awareness at zero cost.  Then run non-AMS promotions in front of a potential readership who have at least seen your name somewhere before (like on page 77 of a sponsored product carousel).   :D

When I started my ads, I was thrilled just to get impressions. It's huge when you can't afford the second tier ads, can't get the holy grail ad, and get minimal results from the little ads. Impressions gave me hope.

Then my sales took off exponentially--they had nearly flatlined, so it's not a fortune we're talking about--and the engine has been running smoothly ever since. Since then I have run a couple of ads elsewhere, but they never did a lot. It was the usual: discount my book to almost no profit per sale, and barely make enough sales to cover the ad cost.

AMS has been consistently profitable, although not for every ad I've done, which is why I shut several of them rather quickly. No need to advertise the second or third book in a series, I learned. My one nonfiction ad (ACPC $0.07) funded everything else until I released the third book in my romance trilogy. Now I can confidently expect lots of sell through on my lead ad for the series, which without counting the massive KU reads would look very unprofitable. I wouldn't get those reads, or the sales, without the top placement my AMS ad gives me day in and day out, because my rank/sales don't justify Page 1 placement in a category. Right now I'm testing whether that ad alone sells my women's fiction and other genre books, too. If it does, then it's the only AMS ad for fiction I need. 

Now that I have a complete series, do I even need the AMS ad? Yes, I believe so, since it's not a best-seller. All the anecdotal evidence this year has been that continuous advertising is now necessary to keep a book selling, whether that advertising takes the form of frequent new releases or of ads. I don't release as often as I'd like, so I intend to keep the ads going.

 

Buy Scrivener for Windows or Mac