Amazon Marketing Services (AMS) - a case study from author Mark Brownless
I should just say at the start I’m not professing to be some kind of expert in the use of AMS, I’m merely passing along the experience of using these ads for some six months or so now. I feel like I’ve done a degree in self-publishing in the last year or so, and since I self-published my debut novel in December 2017, I’ve spent considerable amounts of time learning the ways of Amazon including AMS advertising. I’ve also done quite a lot testing of what does and doesn’t work for me. I'm hoping there is a benefit in me explaining what I've learned along the way.
What are AMS ads?
Amazon Marketing Services (AMS) ads are Amazon’s in house advertising platform for any product that they sell. Previously only available in the US at Amazon.com, AMS ads in the UK have now been available for a couple of months (and similarly so in Canada, so they are slowly rolling out worldwide). This new expansion into the UK market is relevant because AMS UK are quite keen to get folks on board and are offering quite a lot of support at the moment. AMS UK / US are slightly different to each other, but not so much as to make it confusing. It’s worth running ads for both markets.
Registering with the US web-site is easy, you put in an email address to get in and add credit card details.
UK-wise, you need to do a few extra steps in the form of setting up an Amazon Advantage Central account. Don’t worry, you don’t need to do anything with this account, you just need one for AMS UK. The full instructions can be found here, and it’s very straight-forward.
Once you’ve set up your account you can create a campaign. You have three options in the UK, two in the US. The first is the Sponsored Products ads – these use keywords. The next is Product Display ads that use specific genres or even specific book titles to target. The third option in the UK is the Header ad, which is a banner that appears across the top of a search page, but which requires you to have three products or more displayed, and as I have only got two, I haven’t done one of those!
Sponsored product ads are the easiest. You choose which of your products to advertise. Next is the name of your campaign, and it is really important to have a naming convention for your ads so you know which is which. I’ve run twenty eight ads so far and if I hadn’t got a naming convention I’d have to go through each ad to know the difference. You select the daily budget (more of that later), whether you want to run it as an open ended campaign or have a time limit (I always have mine open-ended, because why would you want to stop an ad that just might be getting somewhere?), and then have the option to have automatic or manual targeting. An ‘auto’ ad, where Amazon chooses where to present your ad based on the search terms and relevance, is quite useful to do – it takes seconds to set up and it lets the Amazon algorithm decide where to place your ad. On that basis they have the potential to work very well and I would always have one running. If you choose manual targeting, you can choose a list of keywords Amazon suggests or enter your own list – which you can copy from your own keyword research (see separate blog post on this). You also need to select your ‘bid’ for each keyword. The idea is that adverts are seen by people (called an impression), that they might click on it (called a click!) and that is what you’ll pay for with your bid.
When someone searches on a keyword you are using, all those using that keyword are put into a blind auction. If your ‘bid’ is 12p a click, and someone else’s is 8p, not only will you win and your ad will be presented to the customer instead of theirs, but you’ll only be charged 9p if they click on it (1p more than the next highest bid) – Amazon won’t rip you off. Of course if someone else is bidding 20p, then they will beat your 12p and be charged 13p if the customer clicks on the ad. Remember, you aren’t charged for impressions, just clicks.
When you’ve selected your keywords you can preview your ad – which will use the main thumbnail from your books page. In the UK you don’t add your own copy, in the US, you do, adding a short line of copy to further entice people to buy – very much worth working on, and testing several ads with different copy against each other to see which works.
So that’s it, you’ve set up a Sponsored Product (Keyword ad) and submitted it for Amazon’s approval. As long as you don’t use banned keywords (things like Amazon, Kindle and other brand name terms are a no no), and provided you haven’t got typos in there you should be fine.
When will my ad start running?
Amazon will take a day or two to approve your ad and have it live on your dashboard. After that it’s running. Sponsored Product ads can take up to a week in the US to fully roll out, as if it takes time for your ad to spread out through the whole country. Or something. UK ads don’t have that lag – when I ran my first ones, I was very pleasantly surprised to see thousands of impression in the first day.
A word about daily spend
So, yes, you can start a campaign on a minimum of £1 or $1 per day. Nice and safe, right? Means you aren’t going to spend more than £7 a week. Yes, very true. But think about how that ad will be served to people. Amazon assumes when someone sees an advert that they will click on it.
So, say you’ve got an ad with a £1 daily spend and a bid of 10p. That means the maximum number of clicks you can have before you reach your budget is 10. Amazon assumes everyone will click on your ad – just because they have to, because, however unlikely, it is possible that they will. So the greatest number of impressions you can get on your ad at any one time is 10. If someone then clicks away from your ad without clicking on it, it means that someone else could then be served you’re ad. It’s anyone’s guess how quickly Amazon recognises that you haven’t had all 10 clicks and allows someone else to see the ad, however. Does it re-set straight away, or does it take maybe an hour to re-set? If it takes an hour, you can serve your ad to a maximum of 10 people an hour, making 240 impressions a day.
Brian Meeks, in his excellent book on AMS advertising, says that any ad gaining less than 1000 impressions a day, isn’t really running, and his stats get him 1 click per 800 or so impressions.
So, going back to £1 a day and using Brian’s data, it will take you around 4 days to get one click. Brian also says he expects a sale for every 8-12 clicks. So with 4 days for a click, to get 8-12 clicks and therefore a sale, expect it to take 4-6 weeks. One sale in 4-6 weeks from Amazon ads will lead you to think that AMS ads don’t work.
Okay, so maybe Amazon does recognise someone has left your book’s page without clicking on it and re-sets it more quickly, but you see my point. A daily budget of £1 per day isn’t going to get you more sales.
So how much?
Why not double your daily spend? Why not take it from £1 to £10? Oh but hang on, that sounds expensive, right? So here is where it gets personal, and I’ll give you some of my data. I’ve been running AMS ads in the US for six months and in the UK for a month. I’ve run a total of twenty eight ads – some are still running, some stopped ages ago. I started out with daily spends of between $2 and $5. More recently I’ve upscaled these ads so most of the ads run in the last month have had a minimum daily spend of $50 and some have had a spend of $100. I’ve just totalled up the potential daily spend since I started running ads, and it comes in at an excess of $8000, and most of that has been over the last six weeks or so. That’s an eye-watering sum, and about $7900 more than I can afford until I know I’m getting sales. How much have I actually spent? Just under $50 dollars. In six months. How much have I made in sales? $84 dollars. That isn’t profit, that’s amount of sales. Okay, so the royalties from that wouldn’t be great but the whole point is to get the book out there, to create relevance, so to have the ads as a small loss leader, is okay at this stage.
And besides, I’ve used this as a learning experience. The experience of how to run AMS ads. So I’ve used all I’ve learned about keywords that work, daily spends and bids, and pumped that into AMS UK over the last couple of weeks, such as this example, here:
So, my ad has a daily budget of £100, and in 2 weeks has had 28k impressions – or 2k a day. It’s had 86 clicks. Brian Meeks reckons to have a click per 800-1000 impressions. For this ad, I’ve had clicks per 320 impressions, which is, therefore, pretty good going. Each click has cost me 17p and the overall campaign has spent £14.20. The sales produced has been just under £42 at an average cost per sale of just under 34%. AMS doesn’t directly link to KDP or Createspace, so you just have to guess whether these are ebook or paperback, but I reckon I’m about breaking even with this ad.
Is breaking even worth it? Well if nearly 30 thousand folks have seen my ad, 80-odd folks have been interested enough to click on it and around 10 people have bought it, I would say yes, that is worth it to get it out there and, again, build relevance (more of that later).
I’m now toying with the idea of ramping up the daily spend even more to see if that increases impressions and clicks.
Finally, Brian Meeks says that he has never gone above 48% of his daily spend in any of the thousands of ads he’s run, to such an extent that he thinks Amazon doesn’t allow you to – another failsafe to stop people going over budget if there are too many clicked impressions at the same time, perhaps.
Product Display ads
These are different to keyword ads as they allow you to choose a genre in which to place your ads – thrillers for example or sci-fi – or allow you to choose specific books on which to piggyback your ads. So you could choose all of Sarah Pinborough and Rachel Caine’s books for example. In these ads, whenever someone searches on that genre or that title / author, you book will have the opportunity to be presented to the customer, with the same kind of idea of impressions / clicks / daily budget etc etc. In the US store, you have to specify a total budget beforehand for a campaign – minimum being $100. Again, don’t worry, it’s unlikely you’ll spend anywhere near that. If you do, the chances are you’ll be selling a lot too.
Brian Meeks believes that Product Display ads in the US can take an age to get going and build some relevance – he suggests as much as 4-6 weeks before they will get going. My more limited experience of these would back that up and I haven’t enough data from my UK ads to comment as yet. He prefers Product Display to Sponsored Product ads because your ad is more relevant to those people you’ve targeted and so your impression / click / sales ratios will be better.
Okay, so how the heck do I avoid spending a huge chunk of money I haven’t got?
Simple. Do the math as they say in the US. You need to check your data every day, and I would recommend doing so at the same time every day for accuracy, I do mine in the afternoon. If you don’t have time, just take a screenshot of your summary page, or download a spreadsheet of each ad and save it somewhere to analyse later – you’ll see at a glance how it’s doing. The data is cumulative though, you can’t easily tell if the add is slowing down, growing, or how many impressions / clicks you can expect each day.
What you need is a spreadsheet. Make some columns for the raw data, then identical columns next to these for the daily data. For those without a maths brain, take yesterdays data away from todays for your daily stats. If you miss a day, you’ll need to divide the number by two to average it out etc. From this you can make predictions for how many impressions / clicks / sales you can expect tomorrow and so on. You could also even identify trends in the long term for activity in the week versus the weekend, etc.
So I know I’m getting 2400 impressions a day with 7 clicks and 1-2 sales. I can then monitor the future data to see if it is increasing or decreasing. I can also compare other ads to see how effective they are.
Yes this is a bit of a performance, and some critics of AMS say that it should give you this daily data to save you the bother, but maybe that’s for the future.
I’m not very relevant. No no, before you all start emailing me and telling me that I am, and that someone somewhere (probably) loves me, I don’t mean that. I mean in Amazon terms. For the record, my goldfish loves me by the way.
Amazon will give you a helping hand for the first ninety days of a release, but after that, you’re on your own. If your relevance isn’t there, you’ll be like all those other dead books on Amazon. Either that or you’ll need to kick-start your book, either by releasing another title or by running AMS ads.
I talked to Nicola May a while back about running AMS ads. She ran one and immediately had more impressions in a day than I get in a week. But Nicola is much more relevant than me – she’s got a lot of books on Amazon and her new one is riding high in the charts. Amazon likes Nicola because she’s got a lot of happy customers and is making them a lot of money.
I’ve got a few happy customers (violins playing in background) and some sales. Interestingly, Amazon are now starting to show an interest in me. That’s because my AMS campaigns are increasing my relevance, if not my bank balance, and I’ve just released a new title. From showing me higher in search results with the keywords for my book to posting my ads higher up in searches, my relevance is growing. If sales continue to gradually increase, Amazon will take even more notice, and so on.
Relevance is everything on Amazon – you can even have a search filter for relevance, and it’s the default! So if your ads aren’t making huge amounts of money, but they are getting seen, clicks and some sales, you’re getting more relevant. And it’s all about the long game.
Do use AMS ads
Do use a range of sponsored product and product display ads
Do use an auto ad, because why not?
Do test ad copy, keyword variations, products and product types to find what works best. One ad hasn’t got a point of reference to compare with
Do be prepared to spend some money
Do monitor the ads performance and, yes, look at the dreaded stats
Do be patient – some ads take a while to gain traction
Don’t think they don’t work
Don’t just leave them to get on with it and suddenly find you’re spending a fortune
Don’t keep dead ads going
Don’t fiddle around too much with one ad when it’s running – if you want to make big changes to daily spend, clicks and keywords, make another ad, otherwise your data for that ad will be hard to read.
For true expert input, I would recommend the work of Brian Meeks, Dave Chesson and Dale L Roberts. I’ve read Brian Meeks’ AMS book and continue to dip back into it for reference, Dave Chesson’s Kindlepreneur website provides a lot of free resources for indie publishers from beginners upwards, and Dale L Roberts, with his Self-Publishing with Dale You Tube channel has a wealth of free resources including simple videos on keywords and keyword research, choosing categories and browse paths and, of course, AMS advertising.