Author Topic: AMS Bafflement  (Read 1187 times)  

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
AMS Bafflement
« on: June 11, 2018, 04:13:26 PM »
Just a couple of questions about Amazon Marketing Services...

I'm pretty stymied by the whole thing. I've got a LOT of ads right now and the vast majority (unscientific survey - 90%) never turned on at all. I've used all the different ad types, my covers are (I think) halfway decent, and I don't think my copywriting skills are too bad. Of the ads I've placed that have turned on, very few convert. Is this par for the course or am I doing something terribly wrong? Anybody out there who's rocking the AMS care to share their wisdom?

Paul Neuhaus

Online Dpock

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 738
  • Gender: Male
  • North Idaho
    • View Profile
Re: AMS Bafflement
« Reply #1 on: June 11, 2018, 04:20:35 PM »
How long have they been running?


Online Jena H

  • Status: Edgar Allan Poe
  • *******
  • Posts: 6739
  • North Carolina
  • Desperate character
    • View Profile
Re: AMS Bafflement
« Reply #2 on: June 11, 2018, 04:24:42 PM »
And what are your bids?
Jena

Offline Shawn Inmon

  • Status: Scheherazade
  • *****
  • Posts: 1514
  • Gender: Male
    • View Profile
Re: AMS Bafflement
« Reply #3 on: June 11, 2018, 04:33:19 PM »
It takes time for ads to produce meaningful results that can be analyzed. For me, I don't really pay much attention to an ad until it has 100,000 impressions and 100 clicks, and even that's pretty bare bones. Anything below that can be statistically irrelevant.

If I'm not getting enough impressions, I think it's down to two things: my bid, or the algo. If my bid is too low, then it's game over. I might as well start another ad with a higher bid. If I think it might be the algo (and what makes me think this is if I get a burst of impressions in the first few days, then it slows to a trickle) then I have to dig into where the disconnect is. Typically, it's because my keywords are wrong for the book.

If I'm getting impressions but not enough clicks (at least 1 click per thousand impressions) then I pin it on the ad copy. I scrap it, try another approach with the same keywords and try again. Wash, rinse, repeat, until I find an ad that is getting that click ratio. If revising the ad copy doesn't help, then I look with a critical eye at my cover.

If I'm getting clicks, but not enough sales (my standard is one sale for every ten clicks, but again, not until I have at least 100 clicks) then I think there's either a disconnect between my ad copy and what they are finding on my page, or my blurb needs work. One way to combat the first problem is to use the same hook for the blurb and the ad copy. That lessens the disconnect. The truth is, most people's blurbs aren't great, and are more likely to convert at one sale for every 20 clicks, or even every 30. It's tough to be profitable like that.

One other thought is that, for me at least, AMS doesn't work well with every book. I have two travel memoirs that AMS absolutely slays. My ROI over the last year on these two books is over 500%. At the same time, my two best selling books of all time, have a lifetime ROI of about 65%. Yes, I've lost money on AMS ads on my best selling books.

More than anything, it takes patience and a willingness to experiment and pay attention. Good luck.

Shawn Inmon | Website | Facebook

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #4 on: June 11, 2018, 04:44:45 PM »
How long have they been running?

My earliest still-running ad went up on April 27th. (And, to be fair, it's one of the better performers.) Since then, I've thrown in batches from time to time. (The ad from the 27th is on page 3 of the list if that gives you any idea.)

Paul Neuhaus

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #5 on: June 11, 2018, 04:46:28 PM »
And what are your bids?

Thanks to a book I read on the subject, I've taken a very experimental approach. I've got bids all over the spectrum. From, say, 12 cents a click to 40 cents a click. What do you recommend?

Paul Neuhaus

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #6 on: June 11, 2018, 04:49:16 PM »
It takes time for ads to produce meaningful results that can be analyzed. For me, I don't really pay much attention to an ad until it has 100,000 impressions and 100 clicks, and even that's pretty bare bones. Anything below that can be statistically irrelevant.

If I'm not getting enough impressions, I think it's down to two things: my bid, or the algo. If my bid is too low, then it's game over. I might as well start another ad with a higher bid. If I think it might be the algo (and what makes me think this is if I get a burst of impressions in the first few days, then it slows to a trickle) then I have to dig into where the disconnect is. Typically, it's because my keywords are wrong for the book.

If I'm getting impressions but not enough clicks (at least 1 click per thousand impressions) then I pin it on the ad copy. I scrap it, try another approach with the same keywords and try again. Wash, rinse, repeat, until I find an ad that is getting that click ratio. If revising the ad copy doesn't help, then I look with a critical eye at my cover.

If I'm getting clicks, but not enough sales (my standard is one sale for every ten clicks, but again, not until I have at least 100 clicks) then I think there's either a disconnect between my ad copy and what they are finding on my page, or my blurb needs work. One way to combat the first problem is to use the same hook for the blurb and the ad copy. That lessens the disconnect. The truth is, most people's blurbs aren't great, and are more likely to convert at one sale for every 20 clicks, or even every 30. It's tough to be profitable like that.

One other thought is that, for me at least, AMS doesn't work well with every book. I have two travel memoirs that AMS absolutely slays. My ROI over the last year on these two books is over 500%. At the same time, my two best selling books of all time, have a lifetime ROI of about 65%. Yes, I've lost money on AMS ads on my best selling books.

More than anything, it takes patience and a willingness to experiment and pay attention. Good luck.

I appreciate the thoughtful response. Do you have a sweet spot in terms of bids? Do you prefer one ad type over another?

Paul Neuhaus

Online Jena H

  • Status: Edgar Allan Poe
  • *******
  • Posts: 6739
  • North Carolina
  • Desperate character
    • View Profile
Re: AMS Bafflement
« Reply #7 on: June 11, 2018, 05:08:51 PM »
Thanks to a book I read on the subject, I've taken a very experimental approach. I've got bids all over the spectrum. From, say, 12 cents a click to 40 cents a click. What do you recommend?

It's not for me, or anyone else, to recommend a specific bid.  But what I do is start out with a flat bid rate, usually about 20 or 25 cents across the board.  Then I raise or lower each keyword's bid based on how it performs.

Keep in mind, I'm not an AMS expert, and my experience with ads has been limited compared to other writers' usage of them.
Jena

Online LilyBLily

  • Status: Arthur C Clarke
  • *****
  • Posts: 2245
    • View Profile
Re: AMS Bafflement
« Reply #8 on: June 11, 2018, 05:17:41 PM »
There's a long-running thread about AMS ads. Make yourself read it. Lots of experiences detailed.

Offline Shawn Inmon

  • Status: Scheherazade
  • *****
  • Posts: 1514
  • Gender: Male
    • View Profile
Re: AMS Bafflement
« Reply #9 on: June 11, 2018, 08:58:40 PM »
I appreciate the thoughtful response. Do you have a sweet spot in terms of bids? Do you prefer one ad type over another?

Typically, I bid around .31, end up paying around .19 per click. If I can get a sale every eight clicks, which is my overall average, I'm nicely profitable all the time.

Shawn Inmon | Website | Facebook

Offline TimothyEllis

  • Status: A A Milne
  • ******
  • Posts: 4340
  • Gender: Male
  • Somewhere in space.
  • Aussie in the 27th century.
    • View Profile
    • Timothy Ellis Author
Re: AMS Bafflement
« Reply #10 on: June 11, 2018, 09:59:03 PM »
How many keywords do you have?

How many ads are using the same keywords?


I run 1 single ad, on my latest book. It has 950+ keywords. The bids start above where Shawn's end. And I spike bid key books and authors.

Where ever you see a keyword with no impressions, there is a bid which is too low. The question is, do you want your book on that one or not? If you do, up the bid.

I approach AMS 2 ways on the same ad. I want a very broad impression range across as many books as possible. And I want to be seen on the main performers in my sub-cats. I bid enough to get the broad impressions, and I bid specifically to get the high performer impressions. You dont need a lot on any one book to get exposure.

Too many people look at AMS in terms of how many clicks they get. The stats are rubbish, the collection mechanism obviously doesn't work. I look at impressions, and daily sales and reads totals.


I advertise only 1 book at a time. It brings people into my catalogue. I gauge the effectiveness of the ad, by how well the whole catalog is selling right now. And I budget the ad based on 10% of my gross earnings from day to day. With a max of $80 a day, as I've never had Amazon use it.

This approach worked when I was seriously in trouble after going wide, and I was $10 a day on my book 1. The ad put me back on track until I released the next book, which took off, largely because I stopped the first ad, and did a new one with the same keywords and better bidding for the new one. By back on track, I mean it restored a level of sales and reads which paid the mortgage and bills, and kept me fed.

One of the things I think people do wrong is use the same keywords on many ads, and they all compete with each other. The other things people do wrong is not bid enough, and use way too few keywords. You can have 1000 keywords, so use them! Keywords should include every single author who writes what you do, and every single book currently visible which yours fits alongside. It also should include all the top authors and books in your main sub-cats. People are often afraid of bidding high. Don't be. You can bid $1 (you dont need to) and still not spend more than your daily limit. But 100 impressions and 1 click for that $, may actually bring you $100 across your catalogue.


People dont do what is expected of them. They see impressions, but dont necessarily click then and there. But the impression sticks, and at some point they find your book another way. Maybe off someone-else's also-boughts, which wont show up in the ad stats. Maybe they see it 10 times on ten different books, and then click it off an also-bought somewhere else. How they get there doesn't matter. Maximize your impressions for what you can afford to pay each day, and pay attention to your daily revenue, not what the clicks cost.



Offline Patty Jansen

  • Status: Harvey Chute
  • *********
  • Posts: 12319
  • Gender: Female
  • Sydney, Australia
  • Destroyer of Science Fiction
    • View Profile
    • Patty Jansen Author of SF and fantasy
Re: AMS Bafflement
« Reply #11 on: June 11, 2018, 10:18:16 PM »
AMS is weird, and in the past months, I've thrown several hissyfits at it.

I've been running ads since they opened it to non-KU books. I have a number of lists with hundreds of keywords. I ran those ads mainly for one series, then distilled the keywords that got me sales into one file of a few hundred keywords. Then I put a really high bid on it (after all, the keywords had already proven themselves).

Zip. Nada.

Then I copied the same ad, and put a 5c bid on it. I got some sales.

I mean FFS!

The whole thing is random and ridiculous. The impressions show up after a day or two, but the clicks and sales may not show up at all. I know this because in the past few months there were several time that I grew angry with it, and quit all ads... and even though they showed ZERO sales for weeks, my sales dropped.

So. I have arrived at a low-care system, because IMO any time you spend on trying to finagle these ads is time you could spend writing and with the reporting as atrocious as it is, you may even be basing your decision on wrong information.

I find that the ads die after a week or two.
I want to advertise series starters.
I have 30+ lists of keywords in a spreadsheet.

I copy one ad per series per day, attach a list of keywords and a bid of 21c. Then I kill an equal number of ads at the other end, unless an ad is going well, but as I said, the ads usually die after two weeks.

In this way, the ads are always fresh and have a big carousel of ads for each first in series.

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #12 on: June 12, 2018, 12:57:51 AM »
How many keywords do you have?

How many ads are using the same keywords?


I run 1 single ad, on my latest book. It has 950+ keywords. The bids start above where Shawn's end. And I spike bid key books and authors.

Where ever you see a keyword with no impressions, there is a bid which is too low. The question is, do you want your book on that one or not? If you do, up the bid.

I approach AMS 2 ways on the same ad. I want a very broad impression range across as many books as possible. And I want to be seen on the main performers in my sub-cats. I bid enough to get the broad impressions, and I bid specifically to get the high performer impressions. You dont need a lot on any one book to get exposure.

Too many people look at AMS in terms of how many clicks they get. The stats are rubbish, the collection mechanism obviously doesn't work. I look at impressions, and daily sales and reads totals.


I advertise only 1 book at a time. It brings people into my catalogue. I gauge the effectiveness of the ad, by how well the whole catalog is selling right now. And I budget the ad based on 10% of my gross earnings from day to day. With a max of $80 a day, as I've never had Amazon use it.

This approach worked when I was seriously in trouble after going wide, and I was $10 a day on my book 1. The ad put me back on track until I released the next book, which took off, largely because I stopped the first ad, and did a new one with the same keywords and better bidding for the new one. By back on track, I mean it restored a level of sales and reads which paid the mortgage and bills, and kept me fed.

One of the things I think people do wrong is use the same keywords on many ads, and they all compete with each other. The other things people do wrong is not bid enough, and use way too few keywords. You can have 1000 keywords, so use them! Keywords should include every single author who writes what you do, and every single book currently visible which yours fits alongside. It also should include all the top authors and books in your main sub-cats. People are often afraid of bidding high. Don't be. You can bid $1 (you dont need to) and still not spend more than your daily limit. But 100 impressions and 1 click for that $, may actually bring you $100 across your catalogue.


People dont do what is expected of them. They see impressions, but dont necessarily click then and there. But the impression sticks, and at some point they find your book another way. Maybe off someone-else's also-boughts, which wont show up in the ad stats. Maybe they see it 10 times on ten different books, and then click it off an also-bought somewhere else. How they get there doesn't matter. Maximize your impressions for what you can afford to pay each day, and pay attention to your daily revenue, not what the clicks cost.

There's a lot to chew on here. I took the liberty of copying and pasting your suggestions into a word doc, and I'm gonna give your methodology a try. Thanks!

Paul Neuhaus

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #13 on: June 12, 2018, 01:02:08 AM »
AMS is weird, and in the past months, I've thrown several hissyfits at it.

I've been running ads since they opened it to non-KU books. I have a number of lists with hundreds of keywords. I ran those ads mainly for one series, then distilled the keywords that got me sales into one file of a few hundred keywords. Then I put a really high bid on it (after all, the keywords had already proven themselves).

Zip. Nada.

Then I copied the same ad, and put a 5c bid on it. I got some sales.

I mean FFS!

The whole thing is random and ridiculous. The impressions show up after a day or two, but the clicks and sales may not show up at all. I know this because in the past few months there were several time that I grew angry with it, and quit all ads... and even though they showed ZERO sales for weeks, my sales dropped.

So. I have arrived at a low-care system, because IMO any time you spend on trying to finagle these ads is time you could spend writing and with the reporting as atrocious as it is, you may even be basing your decision on wrong information.

I find that the ads die after a week or two.
I want to advertise series starters.
I have 30+ lists of keywords in a spreadsheet.

I copy one ad per series per day, attach a list of keywords and a bid of 21c. Then I kill an equal number of ads at the other end, unless an ad is going well, but as I said, the ads usually die after two weeks.

In this way, the ads are always fresh and have a big carousel of ads for each first in series.

This is another good response and it lines up with some of my own experience. Every bit of feedback is greatly appreciated.

Paul Neuhaus

Online MissingAlaska

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 725
    • View Profile
Re: AMS Bafflement
« Reply #14 on: June 12, 2018, 06:20:13 AM »
A couple of theories on why AMS is so unpredictable and wonky:
- There's a lot of big money going into AMS now with authors spending at a deficit, or pushing stuffed novels that are 2000 pages for KU reads. Trade publications selling books at $9.99 are also bidding against you. A bid that works this week might not be high enough next week. AMS ads are a moving target that changes constantly depending upon who is participating. This is only going to get worse.
- Amazon probably tests ads to see which works. For an equivalent bid, an ad that gets clicks for a keyword might be shown more often than an ad that doesn't. Given that Amazon wants to sell books, I would also venture to guess that any ad that doesn't see high-enough sales gets a penalty even if its bids are higher. That would explain why some ads show results for a few days and then are rarely shown again.
- Amazon probably prefers not to show the same ad to the same customer if they saw the ad five times and never clicked buy. Again, Amazon wants to sell books. Two ads with the same bid price might go to the ad that hasn't been seen multiple times before. New ads would get an advantage here.
- When two ads have the same bid price, Amazon might also choose the ad more likely to appeal to a given customer and their preferences. The customer's buying history coupled with also-boughts would help guide this algorithm.

All of this seems to be a way to push book prices higher (filtering out low-priced books that can't afford expensive ads) and increase the amount of profit for Amazon (by effectively decreasing the royalty).  It's a smart move by Amazon -- but not so good for indies.

Author of historic and futuristic fiction
Michael S. Nuckols | Webpage

Offline danpadavona

  • Status: Lewis Carroll
  • **
  • Posts: 169
    • View Profile
    • Dan Padavona Author Page
Re: AMS Bafflement
« Reply #15 on: June 12, 2018, 07:20:37 AM »
Sponsored (keyword) ads should begin to run immediately if you have a sizable list of good keywords. However it is common for some ads to generate a modest # of impressions for a few days and suddenly catch fire later in the week.

Product Display (PD) ads targeting interests don't tend to turn on for a few weeks, many times not until the last several days of the active period. PD ads targeting other products turn on faster. For the record, I've never had good success with PD ads. The vast majority of mine never turn on at all, so I usually concentrate on sponsored ads, which are very difficult to scale up.

Though I consistently generate positive ROI with AMS ads, scaling is always an issue.


Dan Padavona | Website | Facebook

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #16 on: June 12, 2018, 10:13:51 AM »
A couple of theories on why AMS is so unpredictable and wonky:
- There's a lot of big money going into AMS now with authors spending at a deficit, or pushing stuffed novels that are 2000 pages for KU reads. Trade publications selling books at $9.99 are also bidding against you. A bid that works this week might not be high enough next week. AMS ads are a moving target that changes constantly depending upon who is participating. This is only going to get worse.
- Amazon probably tests ads to see which works. For an equivalent bid, an ad that gets clicks for a keyword might be shown more often than an ad that doesn't. Given that Amazon wants to sell books, I would also venture to guess that any ad that doesn't see high-enough sales gets a penalty even if its bids are higher. That would explain why some ads show results for a few days and then are rarely shown again.
- Amazon probably prefers not to show the same ad to the same customer if they saw the ad five times and never clicked buy. Again, Amazon wants to sell books. Two ads with the same bid price might go to the ad that hasn't been seen multiple times before. New ads would get an advantage here.
- When two ads have the same bid price, Amazon might also choose the ad more likely to appeal to a given customer and their preferences. The customer's buying history coupled with also-boughts would help guide this algorithm.

All of this seems to be a way to push book prices higher (filtering out low-priced books that can't afford expensive ads) and increase the amount of profit for Amazon (by effectively decreasing the royalty).  It's a smart move by Amazon -- but not so good for indies.

Makes perfect sense.

Paul Neuhaus

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #17 on: June 12, 2018, 10:17:29 AM »
Sponsored (keyword) ads should begin to run immediately if you have a sizable list of good keywords. However it is common for some ads to generate a modest # of impressions for a few days and suddenly catch fire later in the week.

Product Display (PD) ads targeting interests don't tend to turn on for a few weeks, many times not until the last several days of the active period. PD ads targeting other products turn on faster. For the record, I've never had good success with PD ads. The vast majority of mine never turn on at all, so I usually concentrate on sponsored ads, which are very difficult to scale up.

Though I consistently generate positive ROI with AMS ads, scaling is always an issue.

Brian Meeks' book swears by Product Display ads, but my experience is similar to yours. My list is littered with PDAs that never turned on (or turned on for under a hundred impressions and died).

Paul Neuhaus

Offline danpadavona

  • Status: Lewis Carroll
  • **
  • Posts: 169
    • View Profile
    • Dan Padavona Author Page
Re: AMS Bafflement
« Reply #18 on: June 13, 2018, 06:39:26 AM »
Brian Meeks' book swears by Product Display ads, but my experience is similar to yours. My list is littered with PDAs that never turned on (or turned on for under a hundred impressions and died).

If forums / Facebook groups are any indicator of reality, the vast majority of us find PD-I ads never turn on. It's crazy that I'll bid 20-25 cents on Horror, create 10 ads over a few weeks, and none turn on. Then someone else on the Facebook group will bid 10-15 cents and achieve success.


Dan Padavona | Website | Facebook

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #19 on: June 13, 2018, 10:21:14 AM »
If forums / Facebook groups are any indicator of reality, the vast majority of us find PD-I ads never turn on. It's crazy that I'll bid 20-25 cents on Horror, create 10 ads over a few weeks, and none turn on. Then someone else on the Facebook group will bid 10-15 cents and achieve success.

There's always gotta be one smart aleck making the rest of us look bad.

Paul Neuhaus

Offline Atlantisatheart

  • Status: Arthur Conan Doyle
  • ****
  • Posts: 563
    • View Profile
Re: AMS Bafflement
« Reply #20 on: June 13, 2018, 10:44:27 AM »
My best performing ads are the ones I start for .15. Once they have produced a click, I'll raise the price on it to .25. Cost per click doesn't usually raise too far above what it started at 0.9- .12 but those keep producing clicks and sales.

I have ads that have been running for 6+ months that still produce the low cost clicks and sales.

Now the ads I start at .25 or higher for fast advertising to get the book in front of eyeballs are always my worst performing ads. They hit fast and hard and blow themselves out quickly.

Advice; set up a few ads at .15 with as many keywords as you can and forget about them for a few weeks - at first they won't cost you anything, but over time and with some TLC they are worth their weight in gold and quietly sir there clicking up sales for 0.10-0.15 a click.

Offline paulneuhaus

  • Status: Dr. Seuss
  • *
  • Posts: 29
  • Gender: Male
  • Southern California
    • View Profile
    • paul neuhaus dot com
Re: AMS Bafflement
« Reply #21 on: June 13, 2018, 12:38:55 PM »
My best performing ads are the ones I start for .15. Once they have produced a click, I'll raise the price on it to .25. Cost per click doesn't usually raise too far above what it started at 0.9- .12 but those keep producing clicks and sales.

I have ads that have been running for 6+ months that still produce the low cost clicks and sales.

Now the ads I start at .25 or higher for fast advertising to get the book in front of eyeballs are always my worst performing ads. They hit fast and hard and blow themselves out quickly.

Advice; set up a few ads at .15 with as many keywords as you can and forget about them for a few weeks - at first they won't cost you anything, but over time and with some TLC they are worth their weight in gold and quietly sir there clicking up sales for 0.10-0.15 a click.

Here's another one I'm filing away for further experimentation. You guys are a wealth of info!

Paul Neuhaus

Offline Elizabeth Barone

  • Status: Scheherazade
  • *****
  • Posts: 1417
  • Gender: Female
  • Watertown, CT
  • Contemporary Romance & Suspense
    • View Profile
    • Elizabeth Barone: Contemporary Romance & Suspense
Re: AMS Bafflement
« Reply #22 on: June 13, 2018, 01:44:37 PM »
I feel your pain, Paul!

AMS is weird, and in the past months, I've thrown several hissyfits at it.

I've been running ads since they opened it to non-KU books. I have a number of lists with hundreds of keywords. I ran those ads mainly for one series, then distilled the keywords that got me sales into one file of a few hundred keywords. Then I put a really high bid on it (after all, the keywords had already proven themselves).

Zip. Nada.

Then I copied the same ad, and put a 5c bid on it. I got some sales.

I mean FFS!

The whole thing is random and ridiculous. The impressions show up after a day or two, but the clicks and sales may not show up at all. I know this because in the past few months there were several time that I grew angry with it, and quit all ads... and even though they showed ZERO sales for weeks, my sales dropped.

So. I have arrived at a low-care system, because IMO any time you spend on trying to finagle these ads is time you could spend writing and with the reporting as atrocious as it is, you may even be basing your decision on wrong information.

I find that the ads die after a week or two.
I want to advertise series starters.
I have 30+ lists of keywords in a spreadsheet.

I copy one ad per series per day, attach a list of keywords and a bid of 21c. Then I kill an equal number of ads at the other end, unless an ad is going well, but as I said, the ads usually die after two weeks.

In this way, the ads are always fresh and have a big carousel of ads for each first in series.

AMS is so weird. I set up ads for my new release and nothing is happening. My other books' ads are moving slow as molasses, too; in early May they were doing great.  (This is the first time I've run continuous ads, rather than short spurts.) I guess it's time to pause them and copy them.

I'm going to tweak my keywords, too.

My best performing ads are the ones I start for .15. Once they have produced a click, I'll raise the price on it to .25. Cost per click doesn't usually raise too far above what it started at 0.9- .12 but those keep producing clicks and sales.

I have ads that have been running for 6+ months that still produce the low cost clicks and sales.

Now the ads I start at .25 or higher for fast advertising to get the book in front of eyeballs are always my worst performing ads. They hit fast and hard and blow themselves out quickly.

Advice; set up a few ads at .15 with as many keywords as you can and forget about them for a few weeks - at first they won't cost you anything, but over time and with some TLC they are worth their weight in gold and quietly sir there clicking up sales for 0.10-0.15 a click.

I found this to be true, too... for a little while. AMS really is weird!

Maybe I raised the bid too soon.

Thankfully, I'm also running BookBub self-serve ads, and even though they don't perform as well as AMS did for me in early May, they are more consistent.

Connect With Me: BlogFacebook | Instagram | Amazon | Goodreads | Book+Main