KDP Exact Match Keywords: A Graduation Checklist for Moving Winners Out of Broad Match
InteliAds Team
KDP Ads
You open the search term report on a broad match campaign and find a handful of searches that are quietly selling books. The same report also shows dozens of irrelevant terms taking budget at the same bid. Without a clear rule for when to promote a winner, most authors leave everything in broad match, or they move terms after a single lucky order and then wonder why the new campaign stalls. This guide gives you a graduation checklist for KDP exact match keywords, so each promotion is based on evidence and not on a hunch.
Short answer: Use broad match to find searches you would never have guessed, then move a search term to exact match only after it has enough clicks and orders to show repeatable performance. Check the term against your royalty-based break-even ACOS, not just the order count. Add it as a new exact keyword in a separate campaign, start near the bid that earned the orders, and leave the broad campaign running unless the data says otherwise. Exact match gives you bid control over a proven search, but it will not rescue a term that only looked good for a week.
Why broad match is for discovery and exact match is for control
In Amazon KDP ads built on Sponsored Products, broad match lets your ad appear for searches that contain or relate to your keyword in varied forms. That is useful when you are learning how readers phrase what they want. It also means one bid is applied to many different searches, some strong and some irrelevant.
Exact match narrows your ad to searches that closely match the keyword you entered. The tradeoff is less volume and much more predictable cost. When a search term has proven it converts, exact match lets you set its bid on its own evidence, instead of letting a blended broad bid decide for it.
Think of it as two jobs in one account:
- Broad match campaign: collects search terms, tolerates some waste, and runs at a modest bid.
- Exact match campaign: holds only proven terms, and each keyword gets a bid based on its own results.
- Search term report: the bridge between the two, and the only place graduation candidates come from.
Bid-desk tip: keep exact keywords in their own campaign. When a proven term shares a campaign with discovery keywords, you lose the clean read on what is working.
Reading your search term report for graduation candidates
Download the search term report for your broad campaign and sort by orders, then by clicks. You are not hunting for the highest sales figure. You are looking for terms where the clicks, orders, and cost together show a pattern you could repeat at a controlled bid.
Use these fields to build your candidate list. For a deeper walkthrough of the report itself, see the KDP search term report guide.
| Search term report field | Graduation signal to look for | Exact match decision it supports |
|---|---|---|
| Customer search term | Clearly matches the genre, trope, or reader intent of the book | Whether the term belongs in your exact campaign at all |
| Clicks | Enough volume to rule out a lucky sale | Whether the conversion rate is believable |
| Orders | Two or more orders, ideally across different days | Whether the result is repeatable |
| Average CPC | The cost per click you actually paid in broad match | The starting bid for the exact keyword |
| Spend | Total cost so far compared with royalties earned | Whether the term has paid for its own testing |
| ACOS | At or below your royalty-based break-even ACOS | Whether the term is profitable or only acceptable |
| Royalties from attributed orders | Royalty per order, not just the listed sales price | How much room you have to raise the bid |
| Impressions | A meaningful count showing the search has real volume | Whether exact match can scale or will stay small |
Royalties matter because Amazon Ads sales show the book price, not what you keep. A paperback priced at $14.99 may return around $4.60 to you, so your break-even ACOS is closer to 31% than to 100%. The KDP royalties vs Amazon Ads sales article explains this gap in detail.
Minimum evidence before you promote a search term
There is no official threshold, so treat these as working rules and adjust them to your price and category. The goal is to avoid two errors: promoting a term on thin data, and sitting on a proven term for months because you wanted more certainty.
Run each candidate through this checklist before you create the exact keyword:
- Relevance: would a reader typing this actually want your book?
- Clicks: aim for a sample large enough that one order does not define the result. Many authors use roughly 15 to 30 clicks for lower-priced books and more for expensive ones.
- Orders: look for at least two, preferably not all on the same day.
- ACOS against break-even: compare the term with your royalty-based break-even, not with a generic target.
- Trend: confirm recent clicks are not much worse than the early ones.
- Duplicates: check that the term is not already live as an exact keyword elsewhere.
Not every search term deserves the same response. This table covers the situations you will meet most often. If you need rules for stopping losers, the when to pause or lower a KDP keyword guide covers that side.
| Search term pattern in broad match | Graduate to exact? | Console action |
|---|---|---|
| Many clicks, several orders, ACOS under break-even | Yes, strong candidate | Add as exact in a separate campaign at or slightly above the average CPC you paid. |
| Few clicks, one order, low ACOS | Not yet | Leave it in broad match and recheck after more clicks accumulate. |
| Several orders but ACOS above break-even | Maybe, with a lower bid | Add as exact with a bid below the broad CPC to test whether cost can come down. |
| Many clicks, zero orders | No | Do not graduate. Consider a negative exact keyword in the broad campaign. |
| Relevant term with orders but very few impressions | Yes, expect low volume | Add as exact and use a competitive bid so it can actually serve. |
| Term already running as exact in another campaign | No | Skip it. Duplicate keywords make your own ads compete and muddy reporting. |
Worked example: from a broad campaign to an exact match campaign
Hypothetical scenario for illustration only. An author runs a cozy mystery paperback priced at $14.99 with a royalty of about $4.60 per sale. Break-even ACOS is about 31% ($4.60 divided by $14.99). The broad campaign keyword is "cozy mystery books" at a $0.45 bid.
After four weeks, the search term report shows three notable terms:
- "cozy mystery series for adults": 38 clicks, 4 orders, $15.96 spend, $59.96 in ad sales. ACOS is about 26.6%, with royalties of about $18.40.
- "cozy mystery series": 9 clicks, 1 order, $3.78 spend. The sample is too small to judge.
- "free cozy mystery books": 22 clicks, 0 orders, $9.24 spend. This term is a negative candidate.
Only the first term clears the checklist. It is relevant, has a meaningful click count, shows four orders, and sits below break-even. The average CPC was $0.42, so the author creates a new exact campaign and enters the term with a $0.45 starting bid, close to what already worked. The second term stays in broad match for more data, and the third is reviewed as a negative.
That first term earned about $18.40 in royalties against $15.96 in ad spend, so it is only slightly profitable. This is why the exact campaign is the right place for it. The author can now test small bid changes and read the effect cleanly, without a blended broad bid interfering.
Bid-desk tip: after you add the exact keyword, record the date, the old broad CPC, and the new bid. Without that note, you cannot tell later whether a change in results came from the move or from normal fluctuation.
InteliAds helps you track KDP ads and royalties in one dashboard, so you can compare search term spend with royalty-aware profit instead of relying on sales alone. If you manage several accounts, you can review candidates without switching between consoles. See InteliAds pricing or Create an InteliAds account to try it with your own campaigns.
Once a search term is on your graduation list, the next risk is moving it badly. A wrong starting bid or two campaigns bidding on the same query can erase the gain from a good promotion. These two sections cover the bid and the cleanup.
Setting starting bids for kdp exact match keywords
Do not copy the broad keyword bid by default. The search term report shows the average CPC you actually paid on that query, and that is a better anchor than the bid you set on the parent keyword. Start slightly above that CPC so the new exact keyword can win impressions, then cap it at your break-even CPC.
Hypothetical example for a cozy mystery paperback priced at $14.99 with a $4.20 royalty per sale. The search term "cozy mystery gift for grandma" came from a broad keyword bid at $0.55.
| Graduation input | Hypothetical value | How it shapes the exact bid |
|---|---|---|
| Clicks on the term | 14 | Enough to test, still a small sample |
| Average CPC paid | $0.38 | Anchor for the starting bid |
| Orders / spend | 3 orders / $5.32 | Conversion rate of about 21% |
| Ad sales / ACOS | $44.97 / about 12% | Well under the 28% royalty break-even ACOS |
| Break-even CPC | about $0.90 (21% x $4.20) | Ceiling, not a target |
| Starting exact bid | $0.45 (CPC x 1.2) | Room to win impressions without chasing the ceiling |
The break-even CPC is optimistic here because 14 clicks is a thin sample. Bid-desk tip: cap early bids at about half the calculated break-even CPC until the exact keyword has its own click history. For help separating royalties from ad-attributed sales, see KDP royalties vs Amazon Ads sales.
Console actions when you create the exact keyword
- Open the Sponsored Products campaign you use for exact targeting and add the search term as an Exact keyword.
- Enter the starting bid from your CPC-based calculation, not the broad bid.
- Note the date and starting bid so you can compare exact-match performance later.
- Review after a meaningful number of clicks before raising or lowering the bid.
Preventing overlap and double-bidding after you graduate
When the same query can trigger both the broad keyword and the new exact keyword, your own campaigns compete for the impression. You then pay more than needed and the data gets split across two places. The fix is to decide which campaign owns each query.
| Overlap check | What a problem looks like | Console action |
|---|---|---|
| Graduated term still appearing in broad campaign | Clicks keep landing under the broad keyword | Add the term as a negative exact in the broad campaign |
| Same exact keyword in two campaigns | Two bids for one query | Keep one, pause or archive the duplicate |
| Exact bid below broad bid | Broad keeps winning the auction | Raise the exact bid within your cap, or confirm the negative is live |
| Parent broad keyword underperforming | Many other terms, few orders | Review it separately, do not pause based on one graduated term |
- Add the graduated term as a negative exact in the source broad campaign or ad group.
- Search your exact campaigns for the same keyword text before adding it.
- Check the search term report a week later to confirm the query now reports under the exact keyword.
- Keep a short log of what you graduated so you can reverse it if needed.
If you manage many titles or accounts, this log is where mistakes pile up. InteliAds is built to track KDP ads and royalties in one dashboard, with Sponsored Products rules that offer a dry run preview before anything goes live, royalty-aware profit, and targeting filters to find keywords and search terms by performance. The author approves live changes, and a preview does not guarantee results. For background on reading the report itself, see the KDP ads guide.
FAQ
Q: How many clicks should a search term have before I move it to exact match? A: There is no universal number. A term with at least one order and a handful of clicks is a reasonable candidate, while a term with many clicks and no orders belongs on a review or negative list. Judge it against your price and royalty rather than a fixed click count.
Q: Should I pause the original broad keyword after graduating a search term? A: Usually no. The broad keyword may still surface other useful queries. Pause it only if its overall performance, not one graduated term, shows it is not earning its spend.
Q: Do I need to add the graduated term as a negative exact in the broad campaign? A: It is the cleanest way to stop overlap and keep data in one place. Without it, the broad keyword can keep serving the same query and compete with your exact keyword.
Q: What starting bid should I use when a search term converted at a different bid in broad match? A: Start from the average CPC you actually paid on that term, add a modest margin, and keep it under a conservative share of break-even CPC. The old keyword bid is less informative than the real CPC.
Q: Can a search term with orders but a high ACOS still be worth graduating? A: Sometimes. Compare ACOS against your royalty-based break-even, and consider whether the term brings read-through or series sales. Graduate it with a lower bid so you can control cost while you gather data.
Q: How often should I check my search term report for new exact match candidates? A: A weekly or biweekly review suits most active campaigns. Low-spend campaigns can be checked less often, because the report needs enough clicks to show patterns.
Q: What if a graduated keyword stops getting impressions in the exact campaign? A: Check that the keyword is active and the campaign has budget, then compare the bid with the CPC you saw in broad match. A bid that is too low, or a negative exact placed in the wrong campaign, are the common causes. See Amazon ads not spending for a wider checklist.
Q: Is it ever better to skip phrase match and go straight from broad to exact? A: Yes, when a specific query has already shown orders and you want tight control over its bid. Phrase match is more useful when you still want to discover close variations of a promising query.
To answer the question behind kdp exact match keywords: graduate a term when its own data supports it, set the bid from the CPC you really paid, and use a negative exact so the broad campaign stops competing with it. Compare outcomes against royalties, not just ad sales. When you are ready to manage this across titles, review See InteliAds plans or Create an account. Starter has no trial, and you approve any live changes.
Browse more KDP ads guides on InteliAds.
See InteliAds plans: compare Essential, Pro, and Publisher (Essential has no trial; Pro and Publisher include a 7-day trial). InteliAds is a Sponsored Products manager for KDP authors: royalty-aware profit when KDP Sync is linked, user-written IF/THEN rules with dry run, and targeting filters for keywords, product targets, and search terms. Create an account.
Tools and guides: KDP ads optimizer · break-even ACOS calculator