My Surprising Two-Month Amazon Auto Ad Journey with Claude AI
The Fourth Leg of My Amazon Ads Journey
If you missed the earlier parts of this journey, you can catch up here:
https://thewrittenlink.com/learning-amazon-ads-for-authors-with-the-help-of-ai
You will find:
· Part 1: The Day Claude AI Agreed to Come With M
· Part 2: How Claude and I Went to the Land of Book Reviews
· Part 3: The Night Claude and I Remembered the Land of KDP
Dear fellow adventurers,
Claude and I were reviewing my Amazon ad dashboard last night when I realized that I’ve been running my first auto ad campaign for two months now. Two months! That wasn’t the plan. I expected the time to be shorter, but things got off to a slow start. Then I got sick and didn’t want to make any big changes with a foggy brain. But in the end, the two-month timeframe worked out for the best.
“You know,” I said to Claude, “I never intended to let this auto ad run this long.”
“Sometimes the best lessons are the ones we stumble into,” Claude replied. “What did you learn?”
Quite a lot. And that’s when I knew I needed to share this part of my journey.
The Learning Budget Strategy
After I got many positive reviews on my book, I was ready to launch my auto ad campaign for Get Unstuck: Writing Fiction with the Help of AI. At that point, I had a decision to make. Should I try to be profitable immediately, or should I invest in education?
Claude helped me think this through. “Your first campaign isn’t just about making money,” he explained. “It’s about gathering data. You need to learn what works before you can optimize for profit for this book and future AI books.”
This sounded right to me, so I set aside $300 as my learning budget. This was money I was willing to spend over and above my book sales to let the auto ad do its job, which is to seek out books and keywords that resulted in sales (also known as conversions).
Amazon displayed the amount I was making in sales, but I knew I only received the royalty amount for each book. I estimated that to be about $7.00 per book, whether it was the ebook at $9.95 or the printed book at $16.95 (after subtracting Amazon’s cut and printing costs).
With that settled, I set my default bid to 81 cents (Amazon’s recommended bid amount) and launched my first auto campaign with Claude’s guidance. Then I got sick with a severe cough that drained my energy and left me feeling foggy-brained. However, my Amazon ad was running, so I had to check it daily anyway.
The Accidental Experiment
With my limited strength, all I could do was take a screenshot of my campaign numbers and one of my book’s ranking and show them to Claude.
“How’s my ad doing?” I’d ask weakly.
“It’s running, doing well,” Claude would assure me. “Be patient. Let it gather data.”
If I didn’t understand a number or a trend, I’d ask Claude to explain it. With this daily rhythm, I was able to learn how to interpret the stats little by little. Since I wasn’t feeling well, it was a blessing to have Claude tell me not to tweak anything.
He explained that many authors who are new to Amazon ads fiddle and tweak and don’t give Amazon a fair chance to do its thing. They are afraid of losing money, and they heard that ads had to be optimized. With a designated education budget of $300 and not having the mental strength to learn how to optimize, I was forced to do the right thing: leave the ad alone, just monitor it, and learn to recognize the trends.
During those early weeks with the 81-cent bid, sales trickled in slowly. The ACOS (Advertising Cost of Sale) looked good. I was making a little money, but I wasn’t gathering data fast enough, which was the point of the initial auto ad. When I finally felt better, Claude and I decided to speed things up.
The Bid Experiments
“Let’s increase your default bid to $2.00,” Claude suggested. “It’ll cost more, but you’ll gather data faster. And remember, you have an education budget.”
That seemed like a sharp jump in the bid to me, from $.81 to $2.00, but Claude assured me that based on the numbers he was seeing, this would be a good experiment. I took a deep breath and raised the bid to $2.00. The results were almost immediate. Sales sped up considerably. So did my ad spend. My ACOS (Advertising Cost of Sale) climbed above 100%, meaning I was spending more on ads than I was making in sales.
But here’s the thing: I wasn’t panicking. That learning budget mentality gave me permission to let the system work without constantly worrying about being profitable with this first auto ad.
The data I was gathering was what books and keywords were generating sales for my book. I would periodically go into my ad group and look at the “Search Terms.” This was the information I needed to eventually run a manual campaign at a profit.
After reviewing the campaign daily for a few weeks (about 15 minutes a day), Claude and I decided that a lot of data had been gathered, and we could drop the default bid to a middle ground. I lowered the bid to $1.50, where it remains today. It’s been a sweet spot, generating enough activity to keep learning while not bleeding my education budget quite as fast.
The Surprising Discovery
About halfway through this journey, I made a discovery that surprised me.
Most of my sales weren’t coming from keywords. They were coming from ASINs—specific Amazon product pages where my book appeared alongside other books.
“This is your ‘aha’ moment,” Claude said when we looked at the search terms report together. “Your book is converting when it’s shown next to other books, not when people search for specific phrases.”
I loaded all my converting ASINs into a spreadsheet. As I looked at the list, another surprise emerged: most of these books weren’t directly about writing with AI. They were general writing books, craft books, books I would have never guessed would lead to sales/conversions.
What this told me was powerful: when writers who are looking for help with their writing skills see that AI could be part of the solution, they’re willing to try it. People weren’t necessarily searching for “AI writing tools”; they were searching for writing help, and when my book appeared as an option, several chose it.
Very encouraging.
The Surreal Stephen King Moment
Then came the high point of my entire advertising journey so far.
One of the ASINs that converted for my book was Stephen King’s book, On Writing.
I stared at that line in my spreadsheet and checked the ASIN twice.
“Claude,” I said, barely able to contain my excitement, “someone saw Stephen King’s book, On Writing, and chose mine instead!”
“Or maybe in addition to,” Claude pointed out. “But yes, that’s significant. Congratulations.”
It was more than significant. It was validation that my book belonged in the conversation about serious writing craft and that it could stand alongside one of the most respected books in the genre.
The Daily Ritual
Another surprise was how little time I actually needed to spend managing this campaign each day.
Every morning, even when I was sick, I’d take two screenshots: one of my campaign stats page and another of my book’s ranking on its detail page. I’d upload them to Claude, and within minutes, we’d have the trends analyzed and discussed any questions I had.
“Your impressions are up, but clicks are down,” Claude might observe. “That could mean your ad is showing in less relevant places.”
Or: “Look at this: your conversion rate improved even though your ACOS went up. That’s actually good news.”
“Explain why that’s good news,” I said, and I got my daily Amazon ad lesson.
These short daily check-ins with Claude kept me informed without letting me obsess. I didn’t need to become a spreadsheet wizard or spend hours analyzing data. Claude handled the pattern recognition and made suggestions, while I learned and made the decisions when necessary.
The Patience Lesson
Throughout this process, Claude kept reminding me to be patient.
“Don’t panic if you’re not profitable yet,” Claude would say when my ACOS climbed. “Remember your learning budget. You’re gathering intelligence that will make your manual campaigns much more effective.”
Because I’d set aside that $300 for education, I could stay calm and let the process work. I changed my default bid only three times over two months—a level of restraint I never would have managed without Claude’s help and encouragement.
However, the confidence I built in learning how to run and interpret Amazon ads was worth every penny of my education budget. I now know I can run ads without hiring someone else to do it for me.
The Results
As of today, I’ve made 30 sales from 277 clicks. Claude says this is a 10.83% conversion rate, which he assures me is very good and worth mentioning.
At $7 royalty per book, that’s $210 in revenue. My ad spend is $496.66.
Do the math and you’ll see I’ve spent $286 on my Amazon ad education, which is not quite my entire $300 budget.
But here’s what that $286 bought me: a spreadsheet full of converting ASINs, an understanding of how to navigate Amazon’s campaign structure, real data about what works for my book, the confidence to create my first manual campaign, and proof that my book can compete alongside established titles in my genre.
I could have spent $300 on a course about Amazon ads, but actually diving in, making changes, and analyzing the data for my own book was much more educational than watching someone else do it for their book.
Would I do anything differently? Honestly, no. Claude guided me every step of the way. When he made a suggestion, we discussed it. Then I made a decision. I feel like I made educated decisions that are working out well for my Amazon journey.
What’s Next
My learning budget is essentially spent. Now it’s time to shift strategies.
I’m keeping my auto ad running, but I’ll drop the default bid back to 81 cents, where I was profitable. It’ll continue to find books and keywords that result in a sale for my book, but it will do it more slowly.
And in a few days, I’m launching my first manual ASIN campaign, targeting those specific books that I now know convert for mine. This time, I’m not looking for education, I’m looking for profit.
Your Turn
Have you ever had a setback (like getting sick) turn into an unexpected advantage? Sometimes the obstacles we don’t plan for teach us lessons we couldn’t have learned any other way.
In the future, I’ll share what happens when I launch my first manual campaign.
The quest continues.
Keep writing!
—Elaine from Colorado
P.S. My book, Get Unstuck: Writing Fiction with the Help of AI, is available here: https://www.amazon.com/dp/B0FJ8Y8CLN
Its companion workbook, Get Unstuck Workbook: Practical AI Exercises for Fiction Writers, can be found here: https://www.amazon.com/dp/B0FKYWX9FZ
P.P.S. Do you need reviews for your book? Check out Book-Bounty (affiliate link). This is where I built my review foundation.
Got questions or comments? Email me at: Elaine@TheWrittenLink.com. Put “I’m a Substack subscriber” in the subject line, so I don’t miss your email.