Another Look at Amazon Advertising (with someone who is having success)

Last year, I did a blog post on Amazon’s then-new advertising program and whether it’s working for me. It wasn’t, and it still isn’t (I tried again recently with a couple of my pen name books, since they’re in KDP Select), but I had a recent comment from another author who is using Amazon ads to good advantage.

Since it was a useful comment, I decided to post it here where more people would see it, and I also asked the author, Yancy Caruthers, some follow-up questions (if you’re not already familiar with Amazon’s advertising opportunities for authors, check out my earlier post for more on the basics of what the program is and how it works).

Yancy writes:

As I read through the comments, the consensus seems to be that this is a waste of time, but I have had a very different experience.

First, connecting your book to specific titles doesn’t work. People go to that title because they care about that book, not yours. You can be creative with your genres – my book is a military memoir, and I advertise in action/adventure as well as military and medical biographies. Cast the net wide! If I was writing Sci-fi romance, I’d advertise in both sci-fi AND romance, and make sure my cover and title reflected my genre accurately. People who weren’t interested simply wouldn’t click. But some would be, and would want to read more.

I started out bidding at 5c, but wasn’t getting more than a few thousand impressions (in a month) and a handful of clicks. After a couple of months, I increased my bid to 8c and started getting more. Impressions don’t sell books, but they do give you data. Out of several thousand impressions, I was getting 0.8%CTR, and my conversion rate was about 5%. I calculated that at that rate, I could pay up to 12c for clicks and still break even, so I increased my bid to 10c. Realize that you won’t get any impressions at all until the Amazon computer connects your ad to a page or keyword.

I went from selling a handful of books per MONTH to selling a handful every day. I’m averaging between 100-200/month. I’m not getting rich, but for every dollar I spend on the ad, I’m making back about $1.50 (about 30c of that is from KU pages read). Those numbers bring reviews now and then for an added bonus.

The ads are flaky, and I think that’s Amazon’s formula – I’ll get sales every day for a week, then I’ll go a few days without any. It averages out – I even had two days when sales were so good, I hit #848 in the entire catalog. Then I didn’t sell any for almost two weeks and was back down into the high five digits with the one-book-a-day-ers.

A couple of things people should understand – The number of impressions you get is a result of your bid and in what market you are bidding into. Your CTR depends on having a great cover, title, and a catch phrase for the ad – and 1% is considered very good. Your conversion rate depends on having a good landing page with a catchy blurb that makes people who land there want to buy. They are already interested or they wouldn’t have clicked. Just reel them in.

So before you pass judgment on Amazon’s PPC, realize that it’s a lot more complicated than “Does it work or not?” The PPC thing now drives 75% of my online sales. Everything else I do drives the other 25% as well as my physical copy sales (about 10% of the total) but that means I’m spending 98% of my time on 28% of my revenue. I’m currently looking at eliminating my time-wasters (like Twitter – ugh – I’m obviously not doing that right) and focusing on things that I have figured out how to make work for me.

After commenting here, Yancy agreed to answer a few more questions:

You mentioned going really wide with your targeting (i.e. all of science fiction and all of romance if you’re writing science fiction romance, even though that’s a pretty small niche). When I used to do Google ads, I’d find that you would be punished (your ad would be shown less) if you had a low click-through-ratio. It seems that when you go really wide like that, there would be a very small percentage of people who would click and that your CTR would suffer. Thoughts?

Amazon doesn’t want ads that don’t sell anything, but those tend to weed themselves out.  It’s possible that somewhere buried in their magic formula is a mention of CTR and Conversions, but that hasn’t slowed me down as far as I know.

Targeting a wider audience makes sense, within reason.  I made the assumption that readers of fiction action/adventure would be potential readers, even though I wrote a piece of narrative military non-fiction.  Even people who read a narrow genre like sci-fi romance also read other genres.  Which other genres are the most common?

Since CTR is simply a function of the quality of an ad and where it is placed, one could certainly run identical ads in two different genres and measure the CTR.  If one is getting impressions but no clicks, you’re in the wrong market.  If neither is getting clicks, maybe the problem is the ad itself.  If it’s getting clicks but no sales, take a look at your landing page, cover, and blurb.

Do you have any advice for authors on how to measure what their earnings per click end up being? Since you can’t use your affiliates links, the way some do with Facebook ads, there’s no way to tell which sales came from the Amazon advertising campaign. If you weren’t selling any books, and suddenly you’re selling some at the same time as you’re getting clicks, I guess it’s pretty doable, but what if you already sell books, and the amount varies quite a bit per day?

Amazon tracks this data for you, independently of your other sales!  They list impressions, clicks, total spent, total sales, and cost per sale.  You can calculate your CTR by dividing the clicks by the impressions.  The number of sales can be calculated by dividing the total sales by the cost of the title ($2.99 in my case).  Cost of sale is also an important number, even though it’s misleading.  Since Amazon already takes 30% of my $2.99 sale, then 70% is the break-even point.

[Lindsay: Hah, I didn’t remember this feature from when I was tinkering last year, but maybe it’s just because I never got clicks! That’s excellent then.]

I assume you’ve played with ad copy quite a bit. Are there any tips or tricks specific to Amazon that you could share?

I have, but I know very little about it.  I think of it like a Tweet – there are a limited number of characters to tell the viewer why the book is interesting.  Play with it, but be patient and give it time.  No ad generates meaningful data until it’s been seen a few thousand times.  The quality of the ad is important, so check out what others have done.  Equally if not more important is the landing page. Back in September when I started my first campaign, I looked at the landing page and thought, “This is boring.  I wouldn’t buy this.”

It sounds like you also haven’t had much luck targeting specific books. Have you tried doing bestsellers or something that’s just gotten a Bookbub ad in your genre? (With sci-fi romance, I think there just wasn’t that much inventory to pick from.)

After I read your blog, I gave the title-linked ads a try. I chose the top 20 sellers in my specific category and added several more.  In the past 3 weeks I have gotten less than 100 impressions.  My theory is that the more popular titles require a higher bid-per-click.  If I could get impressions on those pages, I would probably sell books, but if I bid that high, then I lose money and I won’t pay people to read my title.

I have scaled it for demonstration purposes, but the CTR and Conversion Rates are actual:

Impressions     x    CTR    =   Clicks.    Clicks x Conversion Rate = Sales.

100,000           x     0.8%  =    800.        800 x 6.2%    =    49 sales…

Assuming a $2.99 title on which I make $2.05 in royalty, those 49 sales made me $100.45.  As long as I didn’t pay more than that for those 800 clicks, then I’m making a profit.  I currently bid a maximum of 10c/click, so my 800 clicks would cost $80 at the maximum and I still make $20.45.  Since my title is also enrolled in KU, then I also get paid for KU pages read.  This has varied, but covers the cost of almost 40% of my clicks.

You may also observe that a very slight variation in CTR or Conversion Rate will make a huge difference in sales.  If I could get my CTR to 1%, for instance, I would average 13 more sales per 100k impressions.

Yancy’s plan going forward:

I am going to stop my title-linked campaign at the end of this month.  It doesn’t seem to get me impressions since I am only bidding 10c/click.

I am going to split my current campaign into two parts.  I want to separate the fiction action/adventure from the non-fiction genres, to see what actually produces the most impressions.  I’ll use exactly the same ad, but I want to see if there is any significant difference in CTR and conversions.  It is possible that I am losing money on half of my campaign and making it back on the other half.

I’m also going up to 12c/click for a month.  If nothing else changes, I’ll be giving away a good chuck of my remaining royalty, but I want to see if I can get a substantial increase in impressions by bidding just slightly higher.

Yancy’s recommendation:

Start out at 5c/click in your own and substantially similar genres.  Be patient, give it a month.  See if you get exposures (some genres are more expensive than others).  Once you have some data (and hopefully some sales) then you’ll know more and can make adjustments.  Understand that none of this happens quickly and real data comes with time.  Steering an ad campaign is a lot like piloting a battleship in a crowded harbor.  I’ve been doing this for almost 6 months and am still making adjustments, trying to find that sweet spot for sales.

http://yancycaruthers.com/