Yes, this is the long-promised post where I share my experience advertising with Amazon and the things I’ve learned — things that could make or break your campaign.
What I’ve Learned Advertising With Amazon
You may remember my past experiences with Amazon Marketing Services (AMS) and the recent plan I’d set for myself. I started promoting in April with 3 kinds of ads:
Ad #1: Sponsored Products, Manual Keywords
This was the bulk of my promos. I chose up to 1,000 keywords for each book and used them to target potential readers.
How does one come up with so many keywords, you ask? Well, there are two easy ways:
1. Choose The Best-Selling Books In Your Genre
The first strategy requires that you find the genres in which your books sells. Amazon does some automatic choosing for you, and you can sometimes see these listed under your book details. If you wish to influence Amazon’s placement of your book, here is a helpful page that details which keywords will place you in which genre, courtesy of Amazon itself.
When you have the category you’re listed under, click on it and you will see what books are doing well that day. My advice would be to choose those with many reviews, as these are more likely to be on the charts for longer. Use both the book’s title and the author name as keywords, and you can easily scoop up a few hundred keywords from this alone.
However, there’s also a better way:
2. Choose Books Also Bought Alongside Yours
Under your title, you will notice the “Also bought” books. Amazon is already linking these to your book, making for a highly targetted group. Just choose the most relevant of these, and you will now have reached almost 1,000 keywords.
Choosing your keywords is probably the hardest party of this kind of ad. However, it can be over in as little as a couple of hours if you do it like I suggested above. After that, all you have to do is come up with a catchy tagline (VERY important!) and set your daily budget and Cost Per Click (CPC) bid.
I chose a daily budget of $5 and an average bid of $.025.
Ad #2: Sponsored Products, Auto Keywords
If you’d rather skip the hard work of coming up with all those keywords, you can trust Amazon to come up with its own keywords. These are automatically selected from your blurb, reviews etc, and aren’t too bad, actually. Set up your budget and CPC bid as above, and you’re set to go in a dozen or so easy clicks.
I did this for some of my ads and got mixed results. For Emotional Beats, auto keyword campaigns outperformed my manual keywords ones. In most other cases, it was the other way around. So, my advice? Try both and see what works for you.
Ad #3: Product Display Upsell
Except for a Sponsored Product campaign, you can also run a Product Display one. The nominal difference is where these appear on the page (see image above), but the real difference lies in targeting: with Sponsored Products, you can target specific keywords, whereas with Product Display you can target specific products.
There are two ways to use this:
1. Target Your Competition
The most obvious way to use a Product Display campaign is by targeting books similar to yours and advertising there. Be sure to only choose books priced lower than yours, as it makes little sense to upsell your product against a (usually more successful) competitor.
2. Protect Your Products
A sneakier way to use Product Display is to protect your own products from your competitors. So, set up one campaign to target your own products and offer either similar products or more expensive ones. To give you an example, I have set up campaigns that promote my series bundles. These ads appear next to each of my individual books. So, if you visit the Pearseus: Rise of the Prince page, you will see that you can either buy this for $2.99 or the entire 5-book bundle for $4.99.
My Results
After the first month of promoting, I realized I was almost $200 in the red. Why? First of all, because I was stupid enough to enter a bid of $35 instead of $0.35 (try to bulk-update these and you’ll see how easy it is to make such a mistake). After a long weekend, this had raised my costs by almost $100.
And second, because I trusted AMS metrics.
Don’t trust the metrics
After two months of promoting, I realize I hate the AMS interface with a passion normally reserved for those who have run over your pet goldfish with their car. You see, unlike Facebook, AMS has no easy way of telling you what you’re spending. The only way you can do this is if you download a spreadsheet with all of your campaigns and using Excel to add the numbers. Not exactly intuitive, right? Even worse, if you want to see how a campaign is performing on a day-to-day basis, well, you can’t. Not even through that spreadsheet. There’s absolutely no time data; just when your campaign started and when it’s scheduled to end.
So, I downloaded a plugin called Machete. This is the only way to know how your campaign has performed through time, plus it gives you an improved interface to see which keywords are under/over performing and allows you to bulk-update them, minimizing the risk of blowing your budget right out of the water. This becomes crucial very fast, as you need to keep an eye on these babies on a daily basis. If you don’t, you can lose your money in no time.
Don’t Trust The Numbers
Which brings me to the main problem with AMS’s metrics: to put it nicely, they’re unreliable. Like, you know that girlfriend you had, the one who stood you up because she forgot she had a hair appointment? The one who was 2 hours late because she had stopped to play with some stray kittens and had lost track of time (true story)?
Well, AMS is even worse than that. They have a metric called ACoS (Advertising Cost of Sales). This is your most crucial piece of information. Basically, this is a percentage telling you whether you’re spending more money than you’re making on an ad. Theoretically, as this percentage approaches 100% you start losing money. When it grows over 100%, you’re spending more money than you’re making and you’re in the red.
So, after downloading the dreaded spreadsheet each day and summing up cost and revenue, I was thrilled to discover I had spent some $500 and made $540. Not bad for my first try, I thought.
And it wouldn’t have been, were it true. Only, it wasn’t. When I added up my actual revenue using KDP/Book Report and CreateSpace, I found out I had only made some $300. After contacting Amazon and asking which estimate was true, I got the following response:
“The estimated total sales in your advertising report will often not match the sales details in your KDP sales report by varying amounts. Your advertising report sales data doesn’t account for declined payments or refunds and may also include sales of associated books, such as if a customer clicks your ad and decides to purchase another book by you or a co-author. Your KDP sales reports show only the final sales numbers and royalties you earned from books published in your KDP account.
The KDP sales reports are exact sales on your book and cannot be under or overstated. The AMS estimated sales is normally overstated as it doesn’t take into account the customer canceling the order (not deciding to make the purchase), Orders declined for payment issues or associated purchases as explained above.”
To me, that’s nuts. How can Amazon not know whether someone proceeded to buy an item or not? Also, how can the AMS estimate be lower than the actual, when it doesn’t even take into account KENP revenue?
Regardless, I have now found out that the actual break-even point as reported by ACoS is closer to 60% instead of 100%. In other words, if a campaign is over 60%, it’s losing money. This is so hugely important that it bears repeating:
If a campaign has an ACoS of over 60%, it’s losing money.
And if you thought this was the last way in which AMS metrics will trip you over, think again: I recently started a Product Display campaign for my Pearseus bundle. AMS reports zero sales, but I actually had 6 sales and some 660 KENP within 2 days of starting that campaign. As I wasn’t advertising it elsewhere at the time, I’m fairly certain these were the result of my AMS campaign.
So, my advice is this: only trust AMS metrics as far as advertising cost is concerned. Don’t hurry to stop a campaign that reports no sales, and don’t trust a campaign that has a low ACoS, unless these agree with your actual sales as reported by KDP/CreateSpace.
The way I’ve stopped losing money (and actually started making a small profit) is to check my billing history every day to see what I spent the previous day. I then compare that to which books I’ve actually sold, and how much I’ve made. Any of these don’t match up, I check the respective campaigns to see what’s going on.
So, in the above example, I know that I have spent $103.13 since May 16th. When I add up my sales from KDP/Book Report and CreateSpace, I find that my revenue is $131,19:
Yes, it looks like a lot of work and requires a calculator. But it actually takes some 15-20′ of my time each morning and it helps me sell books. Which is more than I can say for most other means of promotion.
Scaling Up Is Hard
Another thing I’ve found out the hard way is that scaling up is hard. This is because of a number of factors, such as:
1. Keyword Voodoo
The same keyword that has given you all your sales in one campaign may tank in another. Heck, the same keyword may perform brilliantly on Monday and sink all your profits on Tuesday. I can’t begin to explain this but a sure-fire way to lose a lot of money is to think that by increasing your CPC bid for a high-performing keyword will result in higher sales. My advice? Make all CPC changes incremental, then scale back if you see they backfire.
2. Campaign Voodoo
Like above, only campaign-wide: just because a $5 campaign has a 10% ACoS, don’t think that you can throw another $50 at it and it will generate ten times the sales. Indeed, it may well spring to 1000% within a couple of days. Again, make all budget changes incremental, then scale back if you see they backfire.
3. Be Ruthless
Don’t think that by copying ten times a campaign that gives you $5 a day, you will end up with $50 a day. Most of the copied campaigns will tank. Be ready to pull the plug on them and be ruthless about it. Indeed, be ruthless about any non-performing campaign. Even if it only loses you cents each day, expenses scale up pretty fast. A penny here and a penny there, and you end up spending ten times more than you make.
4. Measure Twice, Cut Once
I know I’m repeating myself, but it’s worth repeating: double-check all of your numbers. Don’t trust any of AMS metrics,except for the billing one: that’s the money Amazon will take from your credit card at the end of the month, so always keep an eye on that. Given the lack of any meaningful time metric, it’s easier to start your campaigns at the first of each month as it’s clearer how many copies you’re selling that month.
If you keep one eye on the billing tab, the other should be on KDP/Book Report and CreateSpace (or whatever you use to sell your print copies). These are the only trustworthy metrics out there.
Further Reading
While writing this post, I assumed you are somewhat familiar with AMS campaigns; that’s why I didn’t write it as a detailed how-to. If it’s detailed instructions you’re looking for, I have the perfect post for you: Amazon Marketing Services by Amazowl. The article provides an overview of Amazon Marketing Services (AMS) and the various advantages it offers such as prominent display ads, the targeting of competitors and the ability to tap into Amazon’s data on individual buyers. Amazowl’s article also explains how does the auction model work, the different ad formats, basic ad creation, brand pages and how to use Amazon analytics. Subscribe to Amazowl for a free guide with the Top 5 Most Costly Mistakes.
Also, check out Five Mistakes Brands Make When Using Amazon Marketing Services (AMS) by Preston Keaton of content26.com for some advanced tips on using AMS.
Finally, Dave Chesson, aka Kindlepreneur, has some great tips on how to choose your keywords for maximum effect, plus KDP Rocket; a nifty piece of software that can help you with this. You can also take his free course on Amazon Marketing Services that covers everything from setting up your first campaign to optimizing your keywords. I’ve taken it and highly recommend it, as it’s one of the best courses on the subject I’ve seen whether you’re a beginner or a seasoned pro. Plus, free. ‘Nuff said.
If you enjoyed this post, please consider voting for me in the 2017 Annual Bloggers’ Bash Awards. Some kind soul nominated me in the Most Informative Blog category. Whoever you are, thank you!
Tweet This
Regarding Optimizing Amazon ads
Thanks for all this information! Do you happen to have any blog post or handout that would address specific keyword optimizing for Amazon ads?
I want to know how to use the data in each of the below keyword instances to optimize my Amazon ads:
1. Keyword is getting no impressions. (Should I just eliminate it? Or is it better to leave it there just in case it starts working? Or should I increase the CPC and see what happens?)
2. Keyword is getting impressions but no clicks. (Leave it for the simple sake of exposure?)
3. Keyword is getting impressions and clicks but few/no sales. (Lower CPC? Raise CPC? etc.)
4. Keyword is getting impressions, clicks and sales. (Do I increase CPC to try to get that keyword out to more people?)
I’ve posted this on a couple of Facebook author/advertising pages, and I found that many others are wanting this same information. If you haven’t already written on this but have mastered your own system, it would make a tremendous blog post!
Hi Nicholas, I’m finally giving this a go and working through all this great info. A question on number 2… is your recommendation to use the “also bought” book titles as keywords, or to search their descriptions for keywords. Just needed clarification on that. Thanks!
You can do both, but it was the former I had in mind. Use their titles and authors as keywords. That way, if someone is looking for those books, they’ll also see yours.
Thanks for clarifying! That’s what I thought you meant. 🙂
I read all your AMS posts back to 2015, started adding categories to my books, watched the kindlepreneur video/class, spent hours gathering 400 keywords, and submitted my first campaign this afternoon. Phew! Fingers crossed! I can’t tell you how helpful your posts were! Thank you!
Woot! I’m thrilled my posts helped you! Best of luck with the promo 😀
AMS is befuddling but you explained it all very clearly. My spend is very small as I’m just dipping a toe in the water, but I’m getting 33% on one ad and virtually nothing on the others. Looks okay so far…
Well done, Tessa, and thank you for the kind words 🙂
ACoS is cost of sales, not cost of profits. Amazon’r reply is off the point. If your average profit margin is 60%, then you are losing money when your cost of sales is over 60%, not 100% as you assumed.
True, ACos is cost of sales. From that, you should be able to deduce your profit. As you say, the point where your campaign is no longer profitable is around 60%, not 100% as one might assume (and several Internet sources suggest). My main gripe with Amazon is that they wildly overestimate sales. Specifically, for every $100 of reported sales, KDP reports around $60 of actual ones. This is misleading and led me to lose a lot of money at first.
Hi Nick,
I want to thank you for making some sense out of a complex KDP marketing program. I’m in my second week and if my sales continue as they have, I should have a successful program. My ACoS is at 26.4%. I’m assuming this is good since it is below the 60% you said was breakeven.
When a have a keyword that is not producing anything, (i.e. no impressions and no clicks) should you go ahead pause the keyword? I have approximately 130 keywords, but only about 20 are actually generating any activity and of those only 5 are generating sales.
I think my book (genre Vietnam War) has a limited number of keywords that are effective. However, should I add more keyword (similar books in my genre)?
Thanks for your help and providing a great post.
Hi Chuck,
Your ACoS is brilliant! I wouldn’t fiddle with anything if I were you, so as not to accidentally break it 🙂
I suggest you copy the campaign to a new one and remove/add keywords there. Run a series of A/B tests to see which one performs better, watch like a hawk any campaign that goes over 50% and kill on the spot any campaign that goes over 60%.
Thanks Nicholas. Last night I added about 30 more keywords that were books of my most active keyword books. I watch every day and download stats every 3 days. Again thank for your help.
Sweet! I hope the book rockets off 🙂
Great post, Nick. I’ve loaded Machete and I hope it keeps working with the new Amazon interface. One thing I’ve wondered about that might hurt is if competitors start clicking your add to use up your spend that day and they can move up the sponsored list. That means you waste money without sales and then you go add more for the day and end up with far fewer returns. It might seem far-fetched but there are unscrupulous people out there willing to spend their time on such a tactic. This could prove costly in and of itself. It seems that there maybe be safer ways to promote without someone intentionally wasting your money. I’ll monitor mine for a while longer and see what happens but I don’t know that AMS will prove to be that helpful. That being said. The tools will help me better determine the usefulness moving forward.
Thank you, Paul. You’re right; unscrupulous people abound! I’ve been hearing some really amazing stories. Stay tuned; I’m sharing later this week.
Looking forward to it!
It’s scheduled for the 25th 🙂
I’ll keep an eye out for it – thx!
This is GREAT information, Nicholas! And extremely generous of you to share. But it also is more proof how much work is involved in being a self-published writer! All the different hats one has to wear.
Oh, tell me about it 😀
Scary how misleading the information is. I also took The Kindlepreneur’s (Dave Chesson) free course recently and will put it to good use at first opportunity. It showed me exactly what I’ve been doing wrong, including setting up multiple campaigns with the same set of keywords. I now see how pointless that is. I blogged about the course and AMS campaigns too only yesterday – the timing is spooky!! Best of luck, my friend! And I had to laugh about the ex girlfriend and the kittens. But kittens are soooo cute! Surely it makes sense a little 😛
Lol – absolutely. Kittens rule 😀
Thanks, Nicholas, for sharing your experience. I’ve thought about trying this, but have held back. You’ve pointed out some glitches I should be aware of if I try this.
Thanks! I hope you find it useful 🙂
Good info here, Nicholas. I’ve bookmarked to re-read often. Thank you.
Thanks! I hope you find it useful 🙂
wow, I’ve had to bookmark this generous share of yours Nicholas. There’s much to dissect here. Thank you! 🙂
Thanks! I hope you find it useful 🙂
I know I will. 🙂
Nicholas, thank you for your thorough investigation and reporting. I also got a headache trying to figure it out on my own, and reading your process gave me another headache! All is fine and I will carry on without advertising on Amazon. What are we Indies to do? A rhetorical question.
Again, thanks Nicholas.
Lol – I apologize for the headache 😀
Thanks for the info, Nicholas!
A pleasure! I hope you find it useful 🙂
Thanks for sharing your knowledge – still muddling through
Best of luck with it. I hope my post helps a little 🙂
I was glad to see your comprehensive report, but I think click ads are too complicated and time consuming for me. If your numbers are spot on then according to Amazon there is around at a 40% Return rate for ebooks and bad credit cards. I have sold on the internet for years and have never seen or heard of that many returns or bad cards no matter what you sell. Ten percent would be a high rate on any kind of book sales. Bad cards are supposed to be no more than 1% across the nation. In accounting when you have one set of figures showing X amount of sales, and another showing actual sales with a 40% lower amount, somebody has altered something somewhere along the line. I don’t think the answer that Amazon sent you begins to address the substantial (close to half the amount ) differences in the two reports.
My feelings exactly. It feels almost like they are deliberately “fattening” returns in order to entice people to use their marketing services.
Thank you. My thoughts exactly.
Wow. How timely. After several campaigns and some bucks down the drain, it’s clear there’s much to learn. And while it’s somewhat comforting Nicholas to know I’m certainly not alone, at some point we all want to feel like Progress Road is within our grasp. Lots of good stuff here to help with that. Thanks a mil! ‘Nuff said here too.
Thank you! I’m so glad you found it useful 😀
Complicated stuff, Nick… way over my head! But good to know. You’re a whizzkid at this stuff, and its really great that you share it with us. Thank you. Love and hugs to the little one. ?
Aw, you, thank you 🙂
Wow, thank you for the detailed analysis. My brain tends to explode from too many numbers but I think I got the gist. 🙂
Lol – sorry about the exploding head 😀
Perro looks too tired to chase anything! 🙂
These are obviously great tips, Nicholas, and I mean no disrespect by saying that just reading them gave me a headache. I am not that good at figures, but my conclusion is that this was a great deal of work for insufficient return. Feel free to correct me, If I got that wrong.
Best wishes, Pete.
The return is getting better now, but yes, it is a lot of work. Software like Dave’s Rocket promises to make it easier, but even so it’s hard work. As for Perro, he’s doing the chasing in his dream 😀
I have saved this to my One Note for when the time comes. Many have advised me not to advertise unless you have at least 3 books. And I’m only on #2! But this post is so excellent to have on hand. Thank you so much for all the time you put in and work you do to help us!
Thank you so much, Luna 🙂
Saving this in a special file for my next book Nicholas. Thanks for being the canary in the coal mine for us!
Lol – this canary almost perished down there, but it was worth it 😀