Author Topic: A New AMS Thread  (Read 101164 times)  

Offline CassieL

  • Status: Arthur C Clarke
  • *****
  • Posts: 2428
    • View Profile
Re: A New AMS Thread
« Reply #50 on: October 14, 2017, 04:10:16 pm »
Edited to say: Sorry guys. Based on new ownership at this site and their response to legitimate author concerns I've chosen to withdraw from participation here.
« Last Edit: September 20, 2018, 11:12:07 am by Cassie Leigh »

8 Pen Names. Genres: Non-fiction, Speculative Fiction, Romance.

KBoards.com

  • Advertisement
  • ***

    icarusxx

    • Guest
    Re: A New AMS Thread
    « Reply #51 on: October 14, 2017, 04:42:31 pm »
    Amazon is supposed to track sales related to clicks, but how good is the tracking? I'm getting sales that don't show up in AMS even after 4 or 5 days have passed, and I am fairly certain the sales came from the advertising. So how accurate is the tracking of sales related to cliicks? Much thanks for your help.

    Offline Jeff DeGordick

    • Status: Madeleine L'Engle
    • **
    • Posts: 78
    • Gender: Male
    • Ontario
      • View Profile
    Re: A New AMS Thread
    « Reply #52 on: October 14, 2017, 06:49:35 pm »
    Also, Jeff, yes $2 is a very high bid. Surprised you're not getting impressions with that high a bid. (I can usually get a book running pretty well on PD display ads in the 60 cent range.) Do you by chance have the ad set to only target those specific books and not do a broad match? When I've tried that, those ads have not moved for me no matter how high I upped the bid. (But I also haven't let an ad sit for weeks like some do with PD ads, so factor that into what I just said.)

    I see under my ad that "Targeted Expansion" is set to "on", if that's what you mean. I'm also surprised I'm not getting more impressions. The books I chose to advertise on are ranked in the #1,000-20,000 range, so I figured maybe I was being outbid. My two clicks show that the average CPC is $1.11, which is still quite high in my opinion, but still much lower than $2.00. It's weird, because I ran a PD ad on an old book of mine for a month about a year ago, and I only selected 12 books with an $0.80 bid, and I got 210,000 impressions.

    Jeff DeGordick

    Offline JDMatheny

    • Status: Lewis Carroll
    • **
    • Posts: 137
    • Gender: Male
      • View Profile
      • Website
    Re: A New AMS Thread
    « Reply #53 on: October 14, 2017, 07:01:34 pm »
    Yeah, that's a click ratio of 1 to 5,000 which is not good and not likely to improve. I would look at which keywords are generating impressions and shut down the one that's doing that to your ad because my personal theory is this:

    Any new ad gets a certain amount of time to run without a lot of judgement from Amazon. It's probably 5-7 days or so. Maybe less, maybe more. During that time you're establishing a track record for that ad. When that free period is up, Amazon's secondary algos kick in. If the ad isn't performing well, the ad gets shut down hard. Or it becomes much more expensive to keep it going. Establish a long enough track record of poorly targeted keywords (which I suspect is what's causing this, a popular keyword that doesn't work with your book) and it's an uphill battle to gets ads going on that book. You have to pay a lot to get movement and then it takes time to show that it isn't your book that was the problem but some badly-targeted words.

    That's just based upon observation and gut feel, it could be wrong. But it does fit with the fairly widely-reported fact that SP ads run well initially but die off for a lot of people after a week or so.

    Also, Jeff, yes $2 is a very high bid. Surprised you're not getting impressions with that high a bid. (I can usually get a book running pretty well on PD display ads in the 60 cent range.) Do you by chance have the ad set to only target those specific books and not do a broad match? When I've tried that, those ads have not moved for me no matter how high I upped the bid. (But I also haven't let an ad sit for weeks like some do with PD ads, so factor that into what I just said.)

    Thanks for the detailed response!  It's interesting, the first ad I did for the book I left into the fourth day and ended up with about 4300 impressions and 2 clicks.  At that time I had been getting burnt out and settled for a description.  After those numbers I re-did the ad to current and created what I consider to be a stronger description, but apparently that hasn't helped.  Might have to re-work it again.

    Another interesting thing, I took the advice to use all author names in the best-sellers and hot and trending in my genres.  I also included automatic keywords, one of which was "book book".  Eh, whatever.  That keyword has gotten over 2500 impressions, a quarter of my total.  It also got one click.

    J.D.Matheny

    Offline Jeff DeGordick

    • Status: Madeleine L'Engle
    • **
    • Posts: 78
    • Gender: Male
    • Ontario
      • View Profile
    Re: A New AMS Thread
    « Reply #54 on: October 15, 2017, 01:05:56 pm »
    I don't think "book book" is going to help you much. Even if you got one of your clicks from it. You're going to have tons of impressions from people who aren't looking for anything close to your book.

    I wrote a haunted house book, and I mistakenly thought it would be a good idea to use the keyword "house". Well, I got tons of impressions, but no clicks. Not surprising; I don't think people looking for home decor books want a haunted house novel.

    Also, if you're having trouble getting clicks, maybe you should post the ad blurb and your cover here for people to critique. Also your keywords as well would be helpful.

    Jeff DeGordick

    Offline JDMatheny

    • Status: Lewis Carroll
    • **
    • Posts: 137
    • Gender: Male
      • View Profile
      • Website
    Re: A New AMS Thread
    « Reply #55 on: October 15, 2017, 06:47:14 pm »
    I don't think "book book" is going to help you much. Even if you got one of your clicks from it. You're going to have tons of impressions from people who aren't looking for anything close to your book.

    I wrote a haunted house book, and I mistakenly thought it would be a good idea to use the keyword "house". Well, I got tons of impressions, but no clicks. Not surprising; I don't think people looking for home decor books want a haunted house novel.

    Also, if you're having trouble getting clicks, maybe you should post the ad blurb and your cover here for people to critique. Also your keywords as well would be helpful.

    Yeah, it seemed weird.  I managed to get automatic keywords in with my manual ones.  I noticed it but then figured, why not.  Guess that's the wrong attitude.  The cover is the one from my signature.  The description was hard because it's so limited on characters, here it is-

    "A deadly confrontation on a remote island will lead Jacob Daniels on the run from dark spirits - but are they real or imagined?  Enter the Torchbearer."

    I have about 140 keywords, without the auto ones included, and they're all authors that I got from the bestseller/hnt lists from my categories.  I've really struggled to find an author/book that are like mine.  I think I'm too picky with that and need to think more generally, but it's been a pain.

    J.D.Matheny

    Offline Philip Gibson

    • Status: Arthur C Clarke
    • *****
    • Posts: 2542
    • Gender: Male
    • Laos (Moved from Yorkshire)
    • Philip and Cookie the Dog
      • View Profile
      • Hashtag Histories
    Re: A New AMS Thread
    « Reply #56 on: October 16, 2017, 06:40:18 am »
    Yeah, I'd turn that "book book" one off. I had a sale off of "edition" once, but its click to impression ratio was horrible so I had to kill it.

    I also tend to kill keywords that have horrible click to impression ratios (worse than 1 click per 2,000 impressions).

    But then, many people point out that if a keyword is getting thousands of clicks with no or tiny numbers of clicks, where's the harm since we only pay for clicks and all those impressions are free eyeballs on our book ad even if they're not optimum eyeballs?

    What is your view on that?

    Philip

    What if there had been social media during important historical events?
    Philip Gibson | Hashtag Histories | Word By Word Graded Readers for Children

    Online Anarchist

    • Status: Arthur C Clarke
    • *****
    • Posts: 2945
    • Methodological individualist
      • View Profile
    Re: A New AMS Thread
    « Reply #57 on: October 16, 2017, 06:54:19 am »
    I also tend to kill keywords that have horrible click to impression ratios (worse than 1 click per 2,000 impressions).

    But then, many people point out that if a keyword is getting thousands of clicks with no or tiny numbers of clicks, where's the harm since we only pay for clicks and all those impressions are free eyeballs on our book ad even if they're not optimum eyeballs?

    What is your view on that?

    I know you asked Cassie, but I'll chime in to reiterate what I posted in January.

    Purging low-performing keywords improves your ad's historical CTR and account's historical CTR. I suspect both are signals to AMS regarding your ads' relevance to site visitors, and will affect how frequently and prominently your ads are displayed over the long run.



    "Opportunity is missed by most people because it is dressed in overalls and looks like work." - Thomas Edison

    "Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat." - Sun Tzu

    Offline Philip Gibson

    • Status: Arthur C Clarke
    • *****
    • Posts: 2542
    • Gender: Male
    • Laos (Moved from Yorkshire)
    • Philip and Cookie the Dog
      • View Profile
      • Hashtag Histories
    Re: A New AMS Thread
    « Reply #58 on: October 16, 2017, 07:39:39 am »
    Purging low-performing keywords improves your ad's historical CTR and account's historical CTR. I suspect both are signals to AMS regarding your ads' relevance to site visitors, and will affect how frequently and prominently your ads are displayed over the long run.

    For that to be true, it would mean the algorithm evaluates not only performance of individual keywords within a campaign, but the campaign itself as a whole.  And then makes decisions/takes action based on the campaign's performance, not just the performance of individual keywords.

    I don't know that we have evidence that that is the case, but meanwhile I will continue culling under-performing keywords in hopes of improving the overall performance of each campaign - something very much against the advice of Brian Meeks and many others.

    Love to hear Cassie's view on how she thinks it works.

    Philip
    « Last Edit: October 16, 2017, 07:42:44 am by Philip Gibson »

    What if there had been social media during important historical events?
    Philip Gibson | Hashtag Histories | Word By Word Graded Readers for Children

    Offline JDMatheny

    • Status: Lewis Carroll
    • **
    • Posts: 137
    • Gender: Male
      • View Profile
      • Website
    Re: A New AMS Thread
    « Reply #59 on: October 16, 2017, 08:11:39 am »
    I make no claim to being great at ad copy, but I'd tweak that to something like: "After a deadly confrontation on a remote island, Jacob Daniels is on the run from dark spirits. But are they real?"

    Depending on how many books you've sold you can also try yasiv.com to see what books connect to yours or use your also-boughts. (Although some of my strongest also-boughts are ones that the ads won't run on. I put them in as keywords and get no impressions.) Or look for books with covers like yours.

    Thanks, I appreciate the advice!

    J.D.Matheny

    Offline BillyDeCarlo

    • Status: Jane Austen
    • ***
    • Posts: 385
    • Gender: Male
      • View Profile
      • Billy DeCarlo, Author
    Re: A New AMS Thread
    « Reply #60 on: October 16, 2017, 08:47:57 am »
    For that to be true, it would mean the algorithm evaluates not only performance of individual keywords within a campaign, but the campaign itself as a whole.  And then makes decisions/takes action based on the campaign's performance, not just the performance of individual keywords.

    I also think it factors in your book, how 'saleable' it is at the moment, i.e. trending/ranking. I see Amazon ads spending more money when my book is on promo and climbing in rank than when it's more stagnant. I think they are very complicated algorithms that Amazon is using.


    All available WIDE!
    Billy DeCarlo | Website/blog | Facebook | Twitter | Google+ | Amazon | BookBub

    Offline NoLongerHere

    • Status: Dr. Seuss
    • *
    • Posts: 49
      • View Profile
    Re: A New AMS Thread
    « Reply #61 on: October 16, 2017, 10:13:49 am »
    Bye
    « Last Edit: September 21, 2018, 04:28:28 pm by NoLongerHere »

    Offline Philip Gibson

    • Status: Arthur C Clarke
    • *****
    • Posts: 2542
    • Gender: Male
    • Laos (Moved from Yorkshire)
    • Philip and Cookie the Dog
      • View Profile
      • Hashtag Histories
    Re: A New AMS Thread
    « Reply #62 on: October 16, 2017, 11:03:49 am »
    Just so I can be clear...

    If a keyword has, say, 1,587 impressions, that means it has won 1,587 auctions for that particular keyword.

    Is that correct?

    Philip
    « Last Edit: October 16, 2017, 11:08:19 am by Philip Gibson »

    What if there had been social media during important historical events?
    Philip Gibson | Hashtag Histories | Word By Word Graded Readers for Children

    Offline BillyDeCarlo

    • Status: Jane Austen
    • ***
    • Posts: 385
    • Gender: Male
      • View Profile
      • Billy DeCarlo, Author
    Re: A New AMS Thread
    « Reply #63 on: October 16, 2017, 11:22:41 am »
    Just so I can be clear...

    If a keyword has, say, 1,587 impressions, that means it has won 1,587 auctions for that particular keyword.

    Is that correct?

    Philip

    It's not an auction. The algorithms have placed it that many times, along with other ads of course. And who knows "where" on the carousel it was placed. Could be page 99.


    All available WIDE!
    Billy DeCarlo | Website/blog | Facebook | Twitter | Google+ | Amazon | BookBub

    Offline EmparentingMom

    • Status: Lewis Carroll
    • **
    • Posts: 236
      • View Profile
    Re: A New AMS Thread
    « Reply #64 on: October 16, 2017, 11:23:07 am »
    Just curious: what's the most people have been able to spend per day on a single sponsored product ad?

    Offline Sunbird Books

    • Status: Dr. Seuss
    • *
    • Posts: 44
    • Gender: Male
      • View Profile
      • Sunbird Books
    Re: A New AMS Thread
    « Reply #65 on: October 16, 2017, 12:04:56 pm »
    Thank you for this thread and all the information within. I started using AMS ads around August playing low-ball with $2 and $3 per day ads. I think that's a great way for those just starting to learn to get their feet wet. I've been maintaining the list of keywords, adding to it every day or two, pausing the keys that rack up major impressions but don't preform in terms of clicks, and upping the bid on those that do preform and lead to sales. My highest bid is $0.26, but I've made some sales and maintained an ACoS below 50% with one click per 688 impressions on average, so you don't have to break the bank to figure it out. I actually believe it's best to play low-ball at first and really learn how the system operates before throwing down $10 or $20 a day with $2 bids. Now that I have a better grasp on things, I'm going to experiment with higher bids and daily budgets when my next book rolls out later this year.

    Offline BillyDeCarlo

    • Status: Jane Austen
    • ***
    • Posts: 385
    • Gender: Male
      • View Profile
      • Billy DeCarlo, Author
    Re: A New AMS Thread
    « Reply #66 on: October 16, 2017, 04:00:06 pm »
    On some of my keywords, under Impressions and Clicks there is an "" for some words, and "0" for others. What is the quantitative difference between the two nothings?   ???   This has been a puzzler for me.

    I think Cassie just answered that two answers ago.


    All available WIDE!
    Billy DeCarlo | Website/blog | Facebook | Twitter | Google+ | Amazon | BookBub

    Offline Philip Gibson

    • Status: Arthur C Clarke
    • *****
    • Posts: 2542
    • Gender: Male
    • Laos (Moved from Yorkshire)
    • Philip and Cookie the Dog
      • View Profile
      • Hashtag Histories
    Re: A New AMS Thread
    « Reply #67 on: October 16, 2017, 05:49:47 pm »
    It's not an auction. The algorithms have placed it that many times, along with other ads of course. And who knows "where" on the carousel it was placed. Could be page 99.

    So what is the "auction" component Amazon talks about?

    What if there had been social media during important historical events?
    Philip Gibson | Hashtag Histories | Word By Word Graded Readers for Children

    Offline Philip Gibson

    • Status: Arthur C Clarke
    • *****
    • Posts: 2542
    • Gender: Male
    • Laos (Moved from Yorkshire)
    • Philip and Cookie the Dog
      • View Profile
      • Hashtag Histories
    Re: A New AMS Thread
    « Reply #68 on: October 16, 2017, 06:03:25 pm »
    This is how Amazon explains how the auction component works:

    "Every time a shopper conducts a search using a competing keyword, advertisers engage in an auction using the different bids assigned to the keyword. The advertiser who has the highest bid and expected CTR wins the auction and gets the right of impression for that search."

    Philip

    What if there had been social media during important historical events?
    Philip Gibson | Hashtag Histories | Word By Word Graded Readers for Children

    Offline Accord64

    • Status: Arthur Conan Doyle
    • ****
    • Posts: 630
    • Gender: Male
      • View Profile
    Re: A New AMS Thread
    « Reply #69 on: October 19, 2017, 02:10:30 pm »
    I know it's been mentioned before, but AMS is really getting strict on their AD policy guidelines.

    My latest ad was rejected for using all caps for a review attribution. I had done this on prior ads without an issue. Not so anymore.

    No complaints because it was an easy fix, but now my ad is stuck in the pending review dungeon. While they say it can take up to three days to be reviewed, none of my prior ads took longer than 12 hours. I'm well into day 2.  ::)

    So, the lesson here is don't get your ad rejected or they'll make you wait after you make corrections.     
     

    Offline weigle1234

    • Status: Lewis Carroll
    • **
    • Posts: 226
      • View Profile
    Re: A New AMS Thread
    « Reply #70 on: October 20, 2017, 09:25:20 pm »

    So, the lesson here is don't get your ad rejected or they'll make you wait after you make corrections.
         

    I have been submitting tons of ads lately (like 5 - 10 each day) - testing various Keyword and bid combinations.  For whatever reason, within the last few days most of my ads are placed into Running within a few hours!

    I never make corrections or edits of any kind.  I simply delete any rejected ads, and start anew.  Doing otherwise usually leads to problems and/or confusion for myself and/or AMS.

    Offline weigle1234

    • Status: Lewis Carroll
    • **
    • Posts: 226
      • View Profile
    Re: A New AMS Thread
    « Reply #71 on: October 20, 2017, 11:03:32 pm »
    Ever since getting involved with AMS about 6 months ago, I have done extensive ad testing (literally hundreds of unique ads).  It is a habit from my 30+ years in the mail order business, where testing of everything (from ads to promotional copy) is the name of the game for making serious money.

    Early on, I discovered that using the AMS default bid level of 25-Cents led to disaster.  Usually within a month or two all 25-Cents ads lost money.

    After that, my first test ads (about 200+) where set at a bid level of 2-Cents (the minimum).  With very few exceptions, all such ads have been profitable, although piddling profits.  Almost without exception I am charged a flat 2-Cents per Click.

    However, I am now testing the identical ads at a bid level of 5-Cents.  The average charge is now only 3-Cents per Click; sometimes 2-Cents, and rarely 4-Cents.  However, the number of Impressions and CTR have increased significantly (sometimes dramatically) with 5-Cents bids.  It is still too early to make any permanent ad changes since I intend to continue such testing with gradually increasing bid levels.

    Since my books are all in a relatively small genre (DIY), my ads seem to be effective in even loosely related sub-genres; which allows much opportunity for testing.  How everything works out long-term remains to be seen.

    To garner accurate stats, each unique test ad has to remain identical, with the exception of increased bids.  Also, each 2-Cents ad must be Terminated before the corresponding 5-Cents ad is placed into Running.  Otherwise, the new 5-Cents ad immediately destroys the effectiveness of the corresponding 2-Cents ad; thus rendering all testing stats useless.

    Offline Bob Stewart

    • Status: Arthur Conan Doyle
    • ****
    • Posts: 613
      • View Profile
      • Street Car Mysteries
    Re: A New AMS Thread
    « Reply #72 on: October 23, 2017, 12:13:39 pm »
    One point I can contribute: AMS does not even track all sales of the book advertised.

    I have some novellas that sell nothing unless I advertise. I can put up an AMS ad that tells me I sold two books a day. In fact, I sold 4-5 and they weren't advertised anywhere else. And they certainly didn't rise high enough in the bestseller lists to get noticed that way.

    I also have one question: I usually do a daily bid of $5-10 for Sponsored Products ads and have never had it pause due to reaching my daily max. Can I assume, raising my daily spend would make no difference whatsoever?

    Offline BillyDeCarlo

    • Status: Jane Austen
    • ***
    • Posts: 385
    • Gender: Male
      • View Profile
      • Billy DeCarlo, Author
    Re: A New AMS Thread
    « Reply #73 on: October 23, 2017, 12:59:37 pm »
    Raising your budget probably won't change things at all. You might get increased spend if you up your bids on individual keywords, but even that's not always guaranteed. It seems that getting the book to move outside of AMS can sometimes then lead to your ads delivering more. For example, I saw that happen when I did a free run on a book that put the book in the top 50 free on the Amazon store.

    I can vouch for that. My AMS ads weren't doing much until I did a Kindle Countdown with promos on it every day, which zoomed it up the charts, and then my AMS ads started spending as well (and I had to up the daily cap).


    All available WIDE!
    Billy DeCarlo | Website/blog | Facebook | Twitter | Google+ | Amazon | BookBub

    Offline ThrillerWriter

    • Status: Jane Austen
    • ***
    • Posts: 476
    • Gender: Male
    • Athens, GA
    • Killing things one word at a time
      • View Profile
    Re: A New AMS Thread
    « Reply #74 on: October 24, 2017, 01:17:55 pm »
    For those of you that are shutting down keywords which aren't performing, how do you decide when to shut them down?

    Is there a certain views that you wait for per keyword? A certain total views with the overall ad? Or a certain amount of time?

    KBoards.com

    • Advertisement
    • ***