Back to Blog
GuidesNovember 8, 2026· 13 min read

KDP Auto Campaign Negatives: A 14-Day Plan for Deciding Which Search Terms to Block

I

InteliAds Team

KDP Ads

KDP Auto Campaign Negatives: A 14-Day Plan for Deciding Which Search Terms to Block

You launched an Auto campaign, opened the search term report on day three, and found a handful of strange queries with clicks and no orders. The urge to block everything that looks odd is strong, but early cleanup is where most KDP auto campaign negatives mistakes begin. Block too fast and you cut off the data the campaign needs. Wait too long and obviously wrong terms drain your budget.

This guide gives you a 14-day order of operations for Amazon KDP ads. It covers when to leave the report alone, which terms to block early, and how to handle terms that spend without ordering when you only have a few clicks to go on.

Short answer: In the first three days, leave the search term report mostly alone, because the sample is too small to judge anything. From days 4 to 7, block only terms that are clearly the wrong audience, such as a different genre, a different format, or an unrelated product. From days 8 to 11, judge relevant terms by spend relative to your royalty per sale, not by click count alone. Keep a written reason for every negative so you can reverse it later.

Days 1-3: Why you should mostly leave the search term report alone

An Auto campaign in Sponsored Products tests Amazon's guesses about where your book fits. In the first 72 hours, most of those guesses have not yet produced clicks, and many have not produced impressions. A term with two clicks and no order tells you almost nothing about a book that converts at a typical rate.

The common trap is treating a short report as a verdict. A term that looks odd on day two may be a reader who would buy. A term that looks perfect may be a browser who never will. Neither conclusion is safe yet.

Use days 1-3 for checks that do not require negatives:

  • Confirm the campaign is delivering impressions across the four auto targeting groups: close match, loose match, substitutes, and complements.
  • Record your break-even ACOS by dividing the royalty per sale by the list price.
  • Write down your relevance rules, meaning which genres, formats, and topics are clearly not your reader.
  • Do not add any negative unless a term is both plainly off-topic and already costing money.

Bid-desk tip: a negative added on day two removes a data point you cannot get back. Note suspicious terms in a list and revisit them on day four.

Days 4-7: Spotting the terms that are clearly the wrong audience

By day four, enough queries have appeared to separate two groups. Some are wrong audience terms, where the shopper wants something your book cannot satisfy. Others are unproven terms, where the shopper is plausible but the sample is thin. Only the first group is a candidate for early kdp auto campaign negatives.

Wrong audience terms usually fall into recognizable patterns:

  • Different genre: a cozy mystery matched to a hard sci-fi query.
  • Different format intent: a coloring book matched to "digital planner" or "audiobook free."
  • Different task: a fiction title matched to "how to" or "study guide" queries.
  • Different age group: an adult romance matched to children's reading terms.

For these, add the term as a negative keyword at the ad group level. Use negative exact when only that precise query is wrong. Use negative phrase when a whole word pattern is wrong, such as a format word that never fits your book. Phrase negatives are broader, so reserve them for patterns you would never want.

If the unwanted traffic comes from a specific product rather than a query, the report will show an ASIN. That calls for a different decision, and a single ASIN with one click is not a reason to act in this window.

Bid-desk tip: ask whether you would refuse this reader if they stood in front of you. If the answer is yes, block it. If the answer is maybe, it belongs in the next section.

A sorting guide for search terms: block now, watch longer, or keep

Once you have a week of data, sort every term with spend into one of three buckets. The table below lists the signals to read and the trigger for each bucket. It assumes you know your royalty per sale, because the royalty sets how much a term can spend before it counts as a problem. For more on that relationship, see KDP royalties vs Amazon Ads sales.

Auto campaign signal Earliest day to act on it Block-now trigger Watch-longer trigger
Term relevance to your book Day 4 Wrong genre, format, or task Plausible reader, unclear intent
Clicks on the term Day 8 Many clicks plus an off-topic query Fewer than about 10 clicks on a relevant query
Spend on the term Day 8 Spend well past one royalty with no orders Spend still below one royalty
Orders from the term Day 8 None, and the term is off-topic None, but the term fits your reader
ACOS on the term Day 11 Far above break-even across repeated orders Above break-even on a single order
Royalty per order Day 4 Not a block trigger on its own Use it to set the spend ceiling
Targeting group source Day 8 One group keeps producing off-topic terms A group with too little data to compare

Read the table from left to right. Block now applies when relevance fails, whatever the spend. Watch longer applies when relevance holds but the sample is small. Keep applies to terms that earn orders at or near your break-even, even if some look unusual.

Tools such as Amazon book ads software can put spend, ACOS, and royalty-aware profit next to each term, so the sorting takes minutes instead of a spreadsheet session. The logic stays the same either way.

Days 8-11: Handling terms that spend without ordering when your sample is small

This is the hardest stretch. A relevant term has spent money and produced no order, and you have to decide whether that is bad luck or a bad fit. The answer depends on how much of your royalty the term has already consumed.

Fresh hypothetical: You run a $14.99 paperback cozy mystery with a $4.20 royalty, so break-even ACOS is 28%. The term "small town murder mystery books" has 7 clicks, $2.45 spend, and 0 orders by day nine. At an average CPC of $0.35, that is still below one royalty. The term fits your reader, so you keep watching.

In the same report, "knitting patterns for beginners" has 11 clicks and $3.85 spend with no orders. That query is off-topic, so it needs no more evidence. You add it as a negative exact and note the reason.

For relevant terms with small samples, use this order of checks:

  1. Compare spend to one royalty. Below that, wait.
  2. Check the click count. Under roughly 10 clicks, the result is mostly noise.
  3. Look at the targeting group. A term from loose match or substitutes deserves more suspicion than one from close match.
  4. Decide on a recheck date. Set day 11 or day 14, then act on the larger sample.

Bid-desk tip: if a relevant term passes about 1.5 times your royalty in spend with no order, lean toward blocking. Reverse the call later if your data says otherwise.

For the wider decision framework on lowering bids versus pausing, the KDP ads guide covers how negatives fit alongside bid changes.

Reading ASIN targets in an auto campaign: when a competitor's book page is worth blocking

The search term report in an Auto campaign mixes typed queries with product pages, shown as ASINs. These come mostly from the substitutes and complements groups. Authors often treat every competitor ASIN as an enemy, but a rival book in your genre is frequently the most valuable placement you have. The decision turns on whether that page attracts your reader, not on who owns it.

Fresh hypothetical: Take the same $14.99 cozy mystery paperback with a $4.20 royalty and 28% break-even ACOS. By day twelve, one ASIN from a bestselling cozy mystery in your subgenre shows 6 clicks, $2.10 spend, and 0 orders. A second ASIN, a knitting kit, shows 9 clicks, $3.15 spend, and 0 orders. A third, a thriller from a different genre, has 1 order on 5 clicks and $1.75 spend.

ASIN type in your results Evidence you need first Negative product target? Note to record
Same-subgenre book, spend under one royalty More clicks, or a recheck date No, keep watching "Fits reader, $2.10 of $4.20 used"
Same-subgenre book, spend past 1.5 royalties, no order At least 10 clicks and a second look at the price gap Likely yes "Relevant but not converting after $6+"
Non-book product such as a craft kit None beyond the mismatch Yes, now "Different product type"
Different-genre book that produced an order Compare cost per order with your royalty No, review the cost "Converts, cost to be checked"
Same-author or same-series page Check whether the click led to a page-read or follow-on sale Usually no "Own catalog, protect"

In this example, the knitting kit gets blocked immediately and the thriller stays because it earned an order. The bestselling cozy ASIN stays on watch. Six clicks with spend at half a royalty is too thin to condemn a page that plausibly sells to your reader.

Use these Amazon Ads console actions when working through ASIN rows:

  1. Open the search term report for the Auto campaign and filter to rows that show an ASIN rather than a typed phrase.
  2. Sort by spend and check each ASIN's product page to confirm genre, format, and price.
  3. Add negative product targeting at the ad group level for pages that fail the relevance check.
  4. Write the reason beside each blocked ASIN so you can reverse it if the page's price or category changes.

Bid-desk tip: a competitor ASIN is worth blocking because of what it sells, not because it is a competitor. If readers click it and buy your genre, it is doing the job an Auto campaign exists to do.

Day 12-14: Building a KDP auto campaign negatives list you can reuse for the next book

By day twelve, you have a record of blocked terms with reasons. The last stage is sorting that record into what transfers and what does not. Reusable negatives describe a wrong audience, such as a format word or an unrelated product type. Book-specific negatives describe a wrong fit for this title only, such as a competing author who suits one book but not another.

Sort your list into three groups before you copy anything to a new launch:

  • Universal blocks: terms like "free audiobook" or craft product ASINs, wrong for any fiction paperback you publish.
  • Genre-level blocks: terms that are wrong for cozy mystery but could be right for a sibling genre.
  • Single-book blocks: terms blocked because of this book's price, length, or series position.

Then carry out the day 12-14 review in the console:

  1. Export or copy the negative keyword and negative product target lists for each ad group.
  2. Re-test the watch list with the larger sample and apply the 1.5 royalty rule from earlier.
  3. Reverse any block whose written reason no longer holds.
  4. Save the universal and genre-level groups as a template for the next launch.

Doing this by hand across several books and accounts gets slow. InteliAds offers Sponsored Products rules with a dry run, so you can preview which terms a rule would flag before anything changes. It also shows royalty-aware profit per term and lets you apply targeting filters, such as isolating ASIN rows or one auto targeting group. These tools support your judgment. They do not decide which terms fit your reader, and you still review each flagged term.

FAQ

Q: Is it too early to add negatives if my auto campaign has only had impressions and no clicks by day five? A: Yes, for most terms. With no clicks there is no spend to protect, so blocking removes exposure without saving money. The exception is a plainly off-topic term that Amazon surfaces anyway, which you can block at no cost to your data.

Q: Should I negate a search term that got one click and no sale if my book is a $0.99 ebook? A: Not on that evidence alone. At a low price, one click is a single data point, and your royalty per sale is small, so the spend ceiling is small too. Block only if the term is off-topic, or once spend passes roughly 1.5 times your royalty with no order.

Q: How do I tell whether a term is irrelevant to my book or just hasn't converted yet? A: Ask whether the shopper wants what your book delivers, judging by genre, format, and task. If the answer is no, it is irrelevant regardless of sample size. If the answer is yes or maybe, treat it as unproven and wait for more clicks.

Q: When should I block a competitor's ASIN that appears in my auto campaign results? A: Block it when the page is a different product type or genre, or when a same-genre page has passed about 1.5 royalties of spend with no order across a meaningful number of clicks. Keep same-genre pages that are still under one royalty of spend.

Q: What should I do with a search term that gets orders but at a cost that is above my break-even? A: Do not block it by default. Lower the bid on that target if you can, or accept the cost if the order brings read-through or series sales. Block only if repeated orders keep the ACOS far above break-even with no offsetting value.

Q: Can negating too many terms starve an auto campaign of impressions in its first two weeks? A: Yes. Auto campaigns rely on Amazon testing a wide range of placements. Heavy early blocking, especially with broad phrase negatives, narrows that testing and can leave you with too little data to judge anything.

Q: Should I copy the negatives from my first book's campaign into a new book in the same genre? A: Copy the universal and genre-level blocks, since they describe wrong audiences. Leave behind single-book blocks tied to price, length, or series position. Review the copied list after the first week of the new campaign.

Q: Do the four auto targeting groups (close match, loose match, substitutes, complements) need separate negative lists? A: Negatives apply at the ad group or campaign level, not per targeting group, so you cannot keep a list per group. You can still track which group produces off-topic terms and adjust that group's bid or turn it off if it keeps surfacing poor fits.

The short version of kdp auto campaign negatives is to block wrong audiences early, judge relevant terms against your royalty, and record a reason for every block so the list can be reused. If you want that process inside one dashboard, See InteliAds plans or Create an account. Essential has no trial, while Pro and Publisher include a 7-day trial with a card required.

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

Put these rules to work

Set IF/THEN conditions, run them every hour, and revert any batch when you need to.