Sucking at Amazon ads, round two

The tougher the job, the greater the reward - George Allen Sr.

Where is my river of money, my shower of Benjamins, my riches in royalties? I did, after, write a whole book and send it into the Amazon machine to be served up on Kindle's around the English-speaking world. Don't I deserve my reward!

When I took my first run at an Amazon campaign to promote my novel Succubus, I failed to generate any results (I wish I was being hyperbolic - I mean no results). I made some tweaks to the wording and placement of my ad and sent it out into the world for a second run. The result? More of the nothing. In fact, less of nothing.

The campaign was programmed to run for ten days with a $10 per day budget. Amazon could have taken $100 of my hard-earned dollars, but apparently chose not to. Because the problem was not that people did not click my ad or that when people did click my ad, they failed to convert into readers of my book. The problem is that the ad wasn't even served by Amazon.

Over the ten-day run, Amazon served the ad (impressions) only three times. Three times! Which is actually less than the number of impressions my first ad triggered (the ad I 'fixed' for the second campaign). Of course I generated no clicks. No-one saw the ad. It never had a chance to be judged, to draw the eye, to encourage someone to want to at least look at the book's page and ponder for a fraction of a minute if they wanted to download a sample.

I don't understand how this happens. And God knows there isn't a human being at Amazon I can call and ask (there isn't, is there?). I want to know what to try next. Should I change my pricing? Should I target a different keyword/category? Should I change my title? My book cover? To find some hint at what to try for round three, I turned to the all-powerful and all-knowing multi-colored Artificial Intelligence behind the curtain of my computer screen - Google.

It's about the motherfudging keywords dummy!

The first article I read covered the basics of what the Amazon ad campaign program is and reinforced that I at least set my campaigns up right. But it didn't really have any deep insight drawn from experience of how to make the system wake up and recognize your ad. It did have this helpful suggestion though - As a general rule of thumb, I’ve found that impressions and clicks can be hard to come by, so more keywords is better. You never know which keyword might turn out to be a surprise winner, and you can always pause under-performing keywords from your campaign once you have data.

For my second ad, I targeted the category of which my book is listed (Supernatural Thrillers). This seems to suggest that targeting keywords, and a broad set of keywords, might get the ad served. Those keywords need to be relevant though. Otherwise, the ad will not resonate with the potential reader.  That led me to this Amazon page regarding keywords for Mystery and Thriller authors.

Back to the campaign. I duplicated campaign number two. There's no need to change the ad because in order to test the ad itself, I have to get it served. My focus this time around is on how to get impressions. Then I can pivot to converting impressions to clicks and clicks to sales.

In the new campaign setup, I paid more attention to the keyword section. Honestly, I am slapping my forehead about it this but I thought you could only choose one. But you can, in fact, select multiple keywords.

Amazon presents you with a list of suggested keywords (keywords tied to your category and, I assume, book description). It also allows you to create your own keywords. The theory I'm testing here is that I need to target the ad to a broader potential pool of readers (as identified by the keywords they are searching in), but not too broad as to waste impressions on readers who are not looking at anything even remotely related to my fiction. That led me to develop a keyword list ten words and phrases deep. Most related to the content (Succubus, Supernatural), but I also added some of the top supernatural thriller author names. I figure if someone is a Dean Koontz fan, they might be open to my book.

I really do feel like I'm stumbling to try and figure this stuff out. And I consider myself a reasonably smart person. The fact that I am only figuring out keywords as part of the ad campaign process the third time through is embarrassing. But I'd rather share that I am dumb and just learning than pretend I have all this stuff figured out because I am writing this for the next person like me, someone who is trying to figure things out for themselves and just on the front end of a lifelong journey to create and share their art.

This campaign will run for two weeks. I will post results after I get them.

A fitful first start at Amazon ads

Why soft-launches don't work for books

Since I published Succubus, I have not done much to promote it. If I had to examine my reasons why, I'd say fear ranks at the top of the list. Fear of negative feedback, specifically. The result of that inaction is predictable; my book has sold only a measly handful of copies.

Not promoting your first self-published book is a predictable mistake. If you read any of the advice available online from the successful independent authors who have blazed the path for writers like us, they all say the same thing. Don't just publish your first book and hope for the best. You have to promote!

Beyond the fact that I have no sales and no revenue from this eighteen-month long project, my lack of promotion means I also have no reviews. Good and bad, reviews are the kind of feedback a writer can grow from. As much as my instinct was to launch my book into the world by sticking it quietly in a corner or Amazon and drawing no attention to it, I need feedback. From actual readers.

Reviews also mean a lot to the almighty Amazon algorithm. That algorithm is like a self-perpetuating machine that can launch your book beyond any reach you can develop for yourself. It's a snowball that if you can just get your book rolled into, builds you readers and money and the opportunity to reach the ultimate dream of writing for you complete living. If I ever want the machine to kick in and start promoting my books to likely readers, I have to it to get noticed. Sales and chatter (reviews) are what matter.

All of which is to say that I know I need to promote my story.

There are, of course, a million ways to promote a book. If you Google the subject, you can very quickly go down the rabbit hole of peoples ideas, theories, past successes, and paid services. I am little-more pragmatic than that. I don't have a blog following or a social-media presence (since I'm not wired that way) so I am going to focus first on learning to use the tools available. Since my book is on Amazon exclusively (for sale and as part of the Kindle Unlimited program - a conscious decision until I actually build a readership and need to expand back out to other markets), I am learning to use the Amazon ad campaign options first.

My first Amazon ad campaign

Setting up my first Amazon Ad campaign was a painless process. The options to try Amazon ads come up right in your Kindle Direct Publishing dashboard. There are a few small steps to get started.

• First, you create the campaign (give it name, select the book you want to promote). It auto-generates an ad from your cover and you type in a snappy, short come-on pitch. My initial pitch "Logan Baker Baker enters a mysterious desert compound and falls under the spell of an alluring but frightening entity that calls itself Constance." 
• You program what categories or reader groups you want to target. In this case, I chose the option for Amazon to decide what categories to place your ad.
• You set the duration of your ad. I opted for a one-week run (to limit the test).
• You decide how much you want to risk/invest. I set a budget of $10 per day maximum. The way the ad works is that Amazon serves the ad and when someone clicks on it, you get charged $0.25 for that click. With a $10 per day budget, I was paying for 20 potential clicks on my ad per day.
• Once you complete the process of building a campaign, it goes through an approval process. It took 48-hours the first time (which meant my ad ran only five days and not seven).

The theory is that a certain percentage of people who click on an ad will download the sample, select the book as part or their Kindle Unlimited subscription, or buy the book. That percentage is the golden number. The higher percentage of people who 'convert,' the less each sale costs (in clicks). If I can yield a sale for an average cost of something that does not consume all my profit per book, then each ad campaign should return greater than its investment (if $10 invested gets me $20 in royalties, then I am golden). If I can get that conversion number right, reinvesting profit from all sales should produce a snowballing reader base. And get my the reviews and readers I need to actually get this thing rolling.

And then there were none

My ad ran. But, unfortunately, at the end of the run, I have nothing to show for the campaign. Looking at the stats, I can see that Amazon barely served the ad at all. And that for the 19 times they did serve it, no-one clicked on it. So I never even had a chance to convert someone into a reader ;-(

On the positive side, the experiment cost me nothing. On the negative side, I again have no feedback on which to grow. So I replicated the campaign to try the experiment again but am left to guess what adjustments to make to try and succeed.

My first guess is that the ad text was not strong enough so I have re-written it this way. "Logan Baker enters a mysterious desert compound and finds himself in the fight of his life with an alluring but frightening supernatural enemy."

My second guess is that the targeting was not specific enough so I set the ad to target the Supernatural Thrillers category on Amazon.

The new campaign is running. The results (impressions served, clicks, and any corresponding sales or KU reads, can take days after the fact to show up in the reporting. I'll be watching and see what I can learn from the iteration of the ad. Hopefully, there will be some feedback to measure, some hint at what is working and what is not. So I can adjust, iterate, try again. And share everything I learn as I do it.