Marketing
How I Set Up Amazon Ads for KDP the Right Way
Ads do not sell your book. They show it to more people faster. Here is the exact order I use to set up Amazon ads for KDP without wasting money.
By Nezir Basar · 2026-06-02 · 7 min read
Ads do not sell your book.
That sounds wrong, I know. You open the Amazon ads dashboard, you see a hundred buttons, and the whole thing whispers one click and the sales roll in. They do not. Amazon ads for KDP only show your book to more people, faster.
So if your cover is weak, your title is confusing, or you have zero reviews, ads just show a bad product to more people, faster. They speed up whatever you already built.
Here is the part almost nobody tells you: the order you launch your campaigns matters more than any single bid or budget. Get the order right and Amazon learns your book. Get it wrong and you pay to teach it nothing.
Let me show you the exact setup I use.
Get your book ready before you spend a cent
I have to say this first, because it saves you money.
Before you run any ads, check three boxes. At least 10 to 15 reviews. A clean cover. A clear title. If even one of those is missing, fix it first.
Ads are not magic. They put your book in front of more people. If the book is good, ads make it better. If it is not ready, no budget will save it.
Box checked? Good. Now you are in the right place.
Why the order of your campaigns matters most
Before you touch the dashboard, you need the plan. The plan is the whole game.
On day one you launch two campaigns: a manual keyword campaign and a manual product campaign. Nothing else. After two weeks you add an automatic campaign. Once you find what works, you build a scale campaign.
Now you are probably thinking the obvious thing. Why not just switch on the automatic campaign right away and let Amazon do the work?
Here is the problem. On day one, Amazon has no idea what your book is. It does not know who should see it. It is guessing with your money.
Run manual campaigns first and you hand Amazon real data. You teach it: these are the buyers who click, and these are the ones who buy. Two weeks later, the automatic campaign launches far smarter.
Think of it like training a new employee. You would not give them full control on day one. You show them how things work, then you let them run.
You know the plan. Now let me show you how to build it.
Campaign one: the manual keyword campaign
This is the most important campaign you will ever run. So get this one right.
Open your KDP dashboard, go to the marketing tab, then the Amazon ads section. Choose your marketplace first. For me that is the US market. Then click through to Amazon ads.
In the sidebar, click create a campaign. Choose sponsored products, then standard ads. Scroll down, leave the ad group name as is, and select the book you want to advertise.
Pick manual targeting and keyword targeting
Keep scrolling to the targeting section. You will see automatic and manual targeting. Choose manual. The automatic one comes in two weeks, not now.
Scroll a little more and you get keyword targeting and product targeting. For this first campaign, pick keyword targeting.
Load your validated keywords, not Amazon's guesses
Here comes the fun part. This is where your homework pays off.
Paste in your validated keywords. I use a set I collected with our keyword research tool, then I drop them straight in. You can also import a CSV or Excel file.
Amazon will suggest its own keywords too. Be careful. They are usually broad and unvalidated, like "journal" or "journal with mom." Choose wisely or skip them.
One rule that keeps your data clean: 15 to 30 keywords per campaign. More than that and the data gets messy. Less and you limit your reach.
Set the match type, the bid, and the budget
Now three settings decide how this campaign behaves.
- Match type: choose broad. Not exact, not phrase. Right now you are in discovery mode. You want to find what works. Exact match comes later, once you know your winners.
- Bid: set every keyword to 30 cents. That is your starting point. You raise the winners and cut the losers once the data tells you which is which.
- Bidding strategy: choose dynamic bids, down only. Amazon lowers your bid when a click is unlikely to convert. It protects your budget so you do not overspend. Fixed bids never worked for me.
Add a few negative keywords too, since broad match casts wide. Words like "free book" or "free ebook" stop Amazon from showing your paid book to bargain hunters.
Name it, budget it, and launch
Name the campaign so future-you understands it at a glance. I use the first letters of the book, then the type. Something like dear_manual_keyword_paperback.
For the daily budget, set $10. Low on cash? Start at $5, or even $3. Most days Amazon never spends the full amount anyway, but a higher cap gives you a better shot at being shown.
Check everything once more. No errors? Click launch campaign. Campaign one is live.
Campaign two: the manual product campaign
Keywords alone are usually not enough. There is a second campaign almost everyone ignores, and it is one of the fastest ways to borrow traffic from your competition.
The process is nearly identical. Create campaign, sponsored products, standard ad, select your book, choose manual targeting. The one difference: pick product targeting instead of keyword targeting.
Find your competitors' ASINs
How do you find competitors? It is simpler than it sounds.
Go to Amazon, search your main keyword, and pick the books you want to advertise next to. Aim for 15 to 30 ASINs, the same sweet spot as keywords. Add them from Amazon's recommendations, search them, or paste a list.
Set the bid to 30 cents again. Keep dynamic bids down only. Name it the same way, swapping "keyword" for "product." Same daily budget, $10 or $5. Launch.
Now your book can show up right beside your competitors. Every time someone studies their listing, yours appears too.
When to add campaigns and how to scale Amazon ads for KDP
Here is the hardest part. You do not touch these two campaigns for two weeks.
I know that feels uncomfortable. But the data you collect in those two weeks is gold, and reacting early just teaches Amazon noise.
Week two: add the automatic campaign
After two weeks, create one more campaign and choose automatic targeting. Same 30-cent bid, same budget. Amazon now knows who clicked and who bought, so it can find buyers you never would have guessed yourself.
Week three or four: find your winners and scale
Now go into your campaigns, open the ad group, and click search terms. Filter for terms with three or more orders. Those are your winners. They actually made you money.
Take those winners into a brand new campaign. This time use exact match, not broad, and raise the bid to 50 or 60 cents. You already know they convert, so you are not guessing. You are doubling down.
That is real scaling. Not spending more on everything. Putting more money behind what already works.
The five-minute weekly routine
Once everything runs, you need about five minutes a week. Here are the rules I follow:
- ACoS over 30%? Lower your bid.
- ACoS under 30%? Raise it by 3 to 5 cents.
- No impressions at all? Your bid is too low or your keywords are not relevant.
- Impressions but no clicks? That is a cover or title problem. Ads cannot fix it.
- Clicks but no sales? Look at your description, your reviews, your A+ content.
And one rule that saves you from yourself: do not optimize too early. Wait until a keyword has at least 1,000 impressions and 10 clicks before you decide anything. Below that, the data is too small to mean a thing.
The whole plan in three lines
Launch two manual campaigns on day one: keyword and product, both broad, both 30 cents. Add the automatic campaign in week two. Find your winners and scale them with exact match in week three or four.
Remember the one truth underneath all of it: ads only put your book in front of more people. A good book gets better. A book that is not ready stays unsold, no matter the budget.
If you want the validated keywords that make campaign one actually work, start with KDP Keyword Research and build your list before you ever open the ads dashboard.