A Practical Guide how to use AI to speed up and make it successful.
By Tom Wawer, author of Corporate Communication DNA
I recently launched my first Amazon Ads campaign for my book. I had zero experience with paid advertising. Within a few days, I had three campaigns running and a system in place to optimize them using AI. This post walks you through exactly how I did it, step by step, so you can do the same.
No fluff. No guru talk. Just what works.
Phase 1: Start With an Automatic Campaign (The Learning Phase)
Before you pick keywords or target competitors, you need data. You do not know yet what your readers search for. Amazon does. So let it figure that out for you.
How to set it up
- Go to advertising.amazon.com and sign in with your KDP credentials
- Click Create Campaign
- Choose Sponsored Products
- Name it something clear, like “CC DNA Auto Discovery”
- Set your daily budget to $10 to $15 (enough to collect data, not enough to go broke)
- Under targeting, choose Automatic Targeting
- Amazon gives you four match groups. Leave them all on for now:
- Close match (similar to your book)
- Loose match (broadly related)
- Substitutes (books people might buy instead of yours)
- Complements (books people buy alongside yours)
- Set a default bid of $0.50 to $0.75
- Select your book
- Launch the campaign
Now walk away. Let it run for 7 to 14 days. Do not touch it. Do not panic if you see spend without sales. You are buying intelligence, not conversions. Not yet.
What is actually happening
Amazon is testing your book against hundreds of search terms and reader behaviors. It is showing your ad when someone searches for “business communication book” or “AI prompting guide” or “career development.” You will soon find out which searches actually generate clicks and which ones waste money.
Phase 2: Pull Your Search Term Report (The Gold Mine)
After your automatic campaign has run for at least 7 days, it is time to look at what Amazon learned.
How to download the report
- In the Amazon Ads console, go to Reports in the left sidebar
- Click Create Report
- Report type: Sponsored Products
- Report: Search Term
- Time period: select the dates your campaign ran
- Click Create Report
- Wait a minute, then download the CSV file
What you are looking at
The report is a spreadsheet. Each row is a search term that triggered your ad. The important columns are:ColumnWhat it meansCustomer Search TermWhat the reader actually typedImpressionsHow many times your ad was shownClicksHow many people clicked on itSpendHow much you paid7 Day Total SalesRevenue from purchasesACOSAdvertising Cost of Sales (lower is better)
Some rows will show search terms that got clicks but zero sales. Some will show terms that converted beautifully. Some will show ASINs instead of words. Those are product pages where your ad appeared.
This report is the foundation of everything that follows.
Phase 3: Use AI to Analyze Your Data
Here is where it gets interesting. Instead of staring at a spreadsheet trying to figure out what matters, you hand it to an AI and ask smart questions.
The prompt that does the heavy lifting
Copy your search term report data and paste it into ChatGPT, Claude, or your preferred AI tool. Then use this prompt:
Prompt:You are an Amazon Ads strategist specializing in book marketing. I have attached / pasted my Search Term Report from an automatic Sponsored Products campaign for my book. Book title: Corporate Communication DNA Book topics: corporate communication, DISC personality types, AI prompting, career development, personal branding, writing skills Please analyze this data and provide: 1. TOP PERFORMERS: Which search terms have the best click-through rate and conversion rate? These are my winners to move into manual campaigns. 2. WASTE: Which search terms have high spend but zero sales? These should become negative keywords. 3. KEYWORD CLUSTERS: Group the performing search terms into 3 to 5 thematic clusters. For each cluster, suggest a separate manual campaign name and 10 additional keyword ideas I should test. 4. ASIN TARGETS: Identify any ASINs in the search terms. Look them up and tell me what books they are. Recommend which ones I should target in a manual ASIN campaign. 5. BID RECOMMENDATIONS: Based on the cost-per-click and conversion data, suggest bid adjustments. Where should I increase bids? Where should I decrease? 6. MATCH TYPE STRATEGY: For my top keywords, recommend whether to use Broad, Phrase, or Exact match and explain why. Format your response as a clear action plan I can implement today in my Amazon Ads console.
What you get back
The AI will sort through your data faster than you ever could and hand you a structured plan. In my case, it identified three clear clusters:
- AI and prompting skills (strong performer)
- Career and professional growth (moderate, needs testing)
- Communication and writing (broad, needs tighter keywords)
It also flagged search terms where I was spending money on completely irrelevant clicks. Things like “DNA testing kit” showing up because my title contains the word DNA. Those go straight into a negative keyword list.
Phase 4: Build Manual Campaigns From Your Data
Now you take what the AI told you and build targeted campaigns. Manual campaigns give you control. You pick the keywords. You set the bids. You decide what match types to use.
Campaign structure that works
Create separate campaigns by theme, not one big campaign with everything in it. Here is what I ended up with:
Campaign A: AI Skills Book
- Keywords about AI prompting, ChatGPT for professionals, AI at work
- Daily budget: $15
- Match type: Broad (still exploring)
Campaign B: Career and Professional Growth
- Keywords about career development, personal branding, corporate skills
- Daily budget: $10
- Match type: Phrase (more targeted)
Campaign C: ASIN Targeting
- Targeting specific competing books directly
- Books like “Crucial Conversations,” “How to Win Friends,” AI prompting guides
- Daily budget: $10
How to create a manual keyword campaign
- Campaign Manager, click Create Campaign
- Choose Sponsored Products
- Name it clearly
- Set daily budget
- Choose Manual Targeting
- Select Keyword Targeting
- Click the Enter List tab
- Paste your keywords from the AI analysis
- Select your preferred match type (Broad, Phrase, or Exact)
- Set your bid ($0.50 is a reasonable start)
- Select your book
- Launch
How to create an ASIN targeting campaign
Same steps, but at step 6 choose Product Targeting instead of Keyword Targeting. Then enter the ASINs of competing books. Your ad will appear on their product pages.
A note on match types
This tripped me up at first. Here is the simple version:
- Broad: your ad shows for searches that include your keyword in any order, plus related terms. Widest net.
- Phrase: your ad shows when the search contains your keyword phrase in order, with possible words before or after.
- Exact: your ad shows only when someone types that exact phrase. Tightest control.
Start broad when you are still learning. Move to phrase and exact once you know which terms convert.
Phase 5: Add Negative Keywords (Stop the Bleeding)
This is the step most beginners skip, and it is the one that saves you the most money.
From your search term report, you identified terms that got clicks but never converted. Add those as negative keywordsso your ad stops showing for them.
How to add negative keywords
- Go into your campaign
- Left sidebar, click Negative Targeting
- Click Add Negative Keywords
- Type or paste the terms you want to block
- Choose Negative Exact (blocks that specific term) or Negative Phrase (blocks any search containing that phrase)
- Save
In my case, blocking irrelevant terms like “DNA test” and “genetics book” immediately cleaned up my spend.
Phase 6: The Weekly Review Loop
Marketing is not a “set and forget” activity. Every week, repeat this cycle:
- Download a fresh search term report (covering the last 7 days)
- Feed it to AI using the same prompt above
- Move winners from automatic to manual campaigns
- Add losers as negative keywords
- Adjust bids based on performance
- Pause anything with high spend and zero results after 14 or more days
Over time, your automatic campaign becomes a keyword discovery tool, and your manual campaigns become your money makers.
A second useful prompt for weekly reviews
Here is my Amazon Ads search term report for the past 7 days. Compare it to last week's data (pasted below). Tell me: 1. What improved? 2. What got worse? 3. Any new promising search terms I should add to manual campaigns? 4. Any terms I should add to my negative keyword list? 5. Should I adjust any bids? Up or down, and by how much? Keep your recommendations specific and actionable.
Phase 7: Scale What Works
After 3 to 4 weeks, patterns will emerge. Some keywords consistently convert. Some campaigns have a healthy ACOS (aim for under 50% for books, ideally under 30%). When you find winners:
- Increase bids on high-converting keywords by 10 to 20%
- Increase daily budget on winning campaigns
- Create exact match versions of your best broad match keywords
- Duplicate campaigns for different formats (I copied my Kindle campaign and swapped in my paperback edition, keeping all targeting the same)
When copying campaigns, double check your match types after the copy. I found a bug where match types got scrambled during duplication. Broad became Exact, Phrase became Broad. Caught it before it could do damage, but it is worth checking every time.
What This Has to Do With Corporate Communication DNA
My book teaches a method called ISLE: intent, Simplify, Layer, Elaborate. The idea is that real knowledge comes from actively engaging with information, not passively consuming it.
Running these ad campaigns is ISLE in action:
- Intent: The search term report is raw data coming in – your intent is get the hell all you can and make money.
- Simplify: You save it, organize it, track it over time, don’t look at details.
- Layer: You analyze it (with AI help) and spot patterns – no need to go deep at this stage, but learn slowly how to get more impact.
- Explore and elaborate: You take action by building campaigns, adjusting bids, writing better ads
This is the same loop whether you are learning a new skill, writing a report, or figuring out Amazon Ads for the first time. The method works because it forces you to engage with reality instead of guessing.
Quick Reference: Your Campaign Toolkit
ToolWhat it doesAutomatic campaignDiscovers what readers search forSearch term reportShows you the raw dataAI prompt (analysis)Turns data into action planManual keyword campaignTargets proven search termsManual ASIN campaignTargets competing booksNegative keywordsBlocks wasted spendWeekly review promptKeeps you improving
Final Thought
I spent 35 years in corporate communication before writing my book. I spent zero years in advertising. But the skills transfer. Analyze your audience. Speak their language. Test, learn, adjust. Repeat.
You do not need an advertising agency. You need a search term report, an AI tool, and the willingness to look at the numbers every week. The rest is just clicking buttons in the right order.
Good luck with your campaign. And if you want to learn more about structured thinking, AI prompting, and building your professional skills, the book is called Corporate Communication DNA. You might have already seen the ad.
Tom Wawer is the author of Corporate Communication DNA: Lessons from a 35-Year Career. The book covers corporate communication, personality frameworks, AI prompting, and career development. Available on Amazon in Kindle, paperback, and hardback.