Over the past year or so, I’ve reported on renewed efforts to raise awareness of the novel A Perfect Blindness. Experimenting with a variety of ideas. Including dabbling with AI in the background. Yet, I’ve left off revealing results. Of course, a major reason to care about what I did is to know what happened. So, the results, the good, the indifferent, and the clear blunders, will get trotted out for your amusement and edification.
Content warning—talk about AI.
Psychological Thriller, Y or N?
Just over a year ago, in April of 2025, I’d noticed that APB shares some DNA with Psychological Thrillers. I hadn’t been moving books or KNEP reads with the Amazon keywords and categories I’d been using for several years. Further, in March of ’25, Amazon updated its algorithm to what’s being called A10. This seemed a prime moment for a change. So, dove into the Amazon best sellers in Psychological Thrillers and plugged those titles into Publisher Rocket’s “Reverse ASIN Lookup” function and got slews of ideas for high-performing categories and long tail keywords. Sifting through those, I found what seemed best matches for this book and created a new list of keywords, carefully chosen to not duplicate other metadata, and broad enough categories to grab as wide a swath of potential readers as possible for this book. Not any psychological thriller, but one like APB. Just as the cognoscenti advise. Note: one large change in A10—Authors only get to choose 3 categories now. The rest, if any, Amazon assigns. Good or bad doesn’t matter. It’s a brute fact.On KDP, each was carefully updated, and now came the waiting. Patience is always advised after any change, as any new information takes time to propagate and for Amazon to know what to do with it. (I.e., constant tinkering isn’t rewarded. Patient experimentation, purportedly, is.)
A month later.
An absolute flat line in royalties, and a slow decay in sales rank. Frustration grew. I know it’s a good book. Complete strangers have written up reviews telling me that. So, how do I reach people like those?
A+ Round One
For a long time, I didn’t understand A+ content. It’s all the flashy images and blocks of text that appear below a book (or other item). I had long believed it to be reserved for traditional publishers and softcover books. Which was true. Once upon a time. The general quality of the images and quote blocks reinforced this belief. But with nothing I’d done moving the needle on the book, I revisited the rules on A+ material and saw that it’s open to any indie book. And has been for a while. Now, it does have to meet minimum standards and even then, be approved. So I taught myself the rules and regs, collected my materials, and using Canva, created a nice set of moody items:
A balance of psychological moodiness, hints of danger, and a dash of setting blended with a Kirkus Reviews blurb, a pull-quote from the book, and a cover element. The clef with notes and the dominant black, red, and white color scheme was put in place to connect future ad images to the book’s Amazon page—something all advertising mavens harp on. So, this was set up for future experimentation with Meta and Amazon ads. Thumbs up so far.
More waiting.
More null results. No sales or page reads, and slow but continuous decay in sales rank.
Clearly, that shift in genre and keywords, plus the A+ material, wasn’t what the book needed.
Experiment one in genre shifting seemed to have failed.
So, what’s next?
Yes, I need to do more outreach rather than wait on some sort of Amazon miracle. I’ve known that for years, but I’ve floundered at that. Part of this is yes, I don’t much like being turned down, but beyond that, I still had to understand a different, more subtle reluctance. This niggling problem has me always couching my thoughts about how to present the book.
The core of the story
“It’s Waxx Traxx! Fan fiction,” I had said to Julie Nash, daughter of Jim Nash, one of the founders of the record label, when I gave her a copy of the book. I met her at the NYC premiere of Industrial Accident, the documentary of Waxx Traxx! Records.
I didn’t have a solid understanding of the old Waxx Traxx! store or label. The original store moved, and the label was sold to Uncle Sam’s Records not long after I moved to Chicago. Decades later, when Industrial Accident came out, I understood just how wrong I had gotten many things. Perhaps my reluctance wasn’t so off base.
I’ve always tried to put the goth-industrial music scene of the late 1980s front and center, while sneaking in that it’s really about identity as if that were a dirty secret. It has three, rotating first-person point-of-view narrators, showing the I of the subjective first-person narrative confronting the he or she of the objective third-person narrative. Imaging saying it shows how even marginal people can demonstrate important ideas. That the book is not quite what it looks like, but is bigger and deeper.
It has always had a personality issue, being both an attempt to preserve a moment in time and place, as well as asking how we build our identities and accomplish anything when we are only the misunderstood characters in the stories everyone tells themselves.
It would still take me time to get a good handle on this dichotomy. Rather, to have it clearly pointed out to me.
But that would be months after the disheartening reality of a failed experiment, and a not-so thrilling appearance at the 2025 Brooklyn Book Fest.
Before the BBF
The Brooklyn Book Festival has been around for almost two decades, and is full of speakers and various events, and a lot of space for authors and publishers to set up tables. Those ain’t cheap, suggesting that good things happen there. But the New Wai, a literary group I’m associated with, ponied up for a set of tables and offered several of its writers seats in the festival for $200 a pop. I gave it a shot. Got my New York State sales tax number. My Square card reader ready.
Pitching APB
But I needed a good way to pitch APB. A one-line response to “what’s the book about?” Something I’ve long struggled with. I know Claude was trained on my book, and it seems very, very likely Gemini was trained on it as well, yet I’ve seen nothing for what AI has taken, so I decided to take something from AI. I collected several editorial and blogger reviews, fed that to ChatGPT, and gave it the following prompt.
“Based on the reviews and back cover description, I need up to 10 one-sentence replies to “What is A Perfect Blindness about?”
The responses were useful, but then realized I needed actual user reviews included. I collected and fed the GPT chat a sample of Amazon and Goodreads reviews. Giving it this Prompt:
returning to the ten concise, one-sentence replies I could use to answer “What is A Perfect Blindness about?” add the following customer reviews as sources.
The responses felt better. Less “commercial marketing kit” and more alive, approachable. Then it asked if I wanted various shorter versions of the best ones, then elevator pitches, then a whole Messaging Toolkit as a document with taglines, one-sentence replies, elevator pitches, 30-second spoken versions, and 10-second hooks. That kicked ass.
It even offered a quick reference version I printed out to take with me. I passed on the offered index cards. Posters.
Most of this revolved around music, including a band-poster style promo sheet — dark, grungy, and designed like a concert flyer with bold taglines and hooks in PDF. It felt like I had a marketing team behind me.
The Day of the BBF
Then the reality of a book festival hit. Most people walk by, avoiding making eye contact. I talked to fewer than ten people and sold a total of 2 books. The people beside me fared similarly. A couple sold a handful of books. My wife, who strolled around, reported much the same for most book sellers. A bust by almost all measures.
That is not the kind of outreach I needed.
Blindspots, the Invitation
Less than three months later, I read a piece in Medium that would change much of what I am and have been doing since. The title: “I Asked AI to Figure Out My Blindspots Using a New Prompt”.
And, yes, since I gave GPT that same prompt and implemented some of the insights, the book has seen modest improvements. Even as I am only partway through putting into play all I discovered. I’ll cover each move and what happened in the coming weeks.
To be clear, I’m talking about using AI to assist with the marketing and the platform around the book. AI in the background. Not as author.
If you’ve had success with AI in the background, Shoot me a line. I’d be curious to learn.
Thanks for reading,
WLH
