Home · Blog

Exact, Phrase, and Broad Match for Book Ads

Match types decide which searches can trigger your keyword. How broad, phrase, and exact match behave on Amazon, when each one makes sense for a book, and a simple campaign layout for a tight budget.

The BookAdsAI team · Campaigns · Published October 5, 2026.

When you add a keyword to a Sponsored Products campaign, Amazon asks you to pick a match type. That choice controls which shopper searches the keyword is allowed to enter. It is the single biggest lever you have over relevance, and it matters more for books than for most products, because readers search in oddly specific ways.

The three match types in plain terms

Broad match

Broad match lets Amazon show your ad for searches that contain your words in any order, plus plurals, close variations, and related terms Amazon judges to be relevant. A broad keyword "space opera" can match "space opera series", "opera in space books", and sometimes "military science fiction". That last kind of stretch is the point and the danger. Broad finds searches you would never have thought of, and it also spends money on searches you would never have chosen.

Phrase match

Phrase match requires the search to contain your phrase in the same order, with other words allowed before or after. "Space opera" in phrase match can trigger "best space opera books 2026" or "space opera kindle unlimited", but not "opera about space". It is a middle setting: still some discovery, much less drift.

Exact match

Exact match triggers only for searches that are the same as your keyword, allowing close variants such as plurals and common misspellings. "Space opera books" in exact match shows for "space opera books" and "space opera book", and not for "space opera books for teens". This is the most controlled setting and usually the most expensive per click, because exact searches tend to be high intent and other authors bid on them too.

How the choice changes your results

Think of match types as a trade between discovery and control.

• Broad: many impressions, more varied searches, lower average click price, and a lower share of buyers. Good for finding new terms. Needs negatives to stay clean.
• Phrase: fewer impressions, searches still on topic, moderate click prices.
• Exact: fewest impressions, the most predictable traffic, and the clearest data for each keyword.

On a small budget, discovery is expensive. Every click on a search that was never going to buy is money that could have gone to a search that does. That is why it usually makes sense to start tighter when cash is tight and loosen up only when you have a reason.

A simple layout for one book

You do not need a dozen campaigns. Four is enough for most authors, and three works well while you are learning:

1. Auto campaign, low bid. Let Amazon find searches and product pages. Its job is research, not profit.
2. Broad or phrase discovery campaign. Ten to thirty keywords drawn from your genre, the words readers of comparable books use, and terms from the auto campaign. Moderate bids.
3. Exact campaign. Only proven terms that produced sales or meaningful page reads in campaigns 1 and 2. Your highest bids live here.
4. Product targeting campaign (optional). Comparable books by ASIN, added once you know which ones convert.

Each week, move terms that convert from the search term report of campaigns 1 and 2 into campaign 3 as exact keywords. Then add those same terms as negative exact in the campaign they came from, so your discovery campaigns stop paying for traffic the exact campaign now handles.

Common mistakes

• Same keyword, all three match types, same bid. The campaigns compete with each other, and Amazon picks which one serves. You lose the clean data that made exact worthwhile. If you run all three, bid exact highest and broad lowest.
• Single-word broad keywords. "Dragons" in broad match will match almost anything with dragons in it, including toys and costumes. Use two- and three-word phrases that describe a reader, not a noun.
• Never adding negatives to broad. Broad without negatives drifts. Read the search term report and add negatives every week or two.
• Judging exact on three days of data. Exact keywords often have little volume. Give them a few weeks before deciding.
• Using author names without care. Targeting another author's name can reach the right readers, but conversion is often lower than you expect, because those shoppers want a specific book. Test with modest bids.

Choosing keywords that fit each match type

Good broad and phrase keywords describe a reader's intent in two to four words: "small town romance", "cozy witch mystery", "productivity for writers". Good exact keywords are full searches you have seen convert: "small town romance kindle unlimited", "cozy witch mystery series".

Avoid stuffing a discovery campaign with hundreds of loosely related terms. A smaller list you can actually read and manage beats a large list that hides its waste.

How much to bid by match type

There is no universal number, because click prices differ by genre and season. A sensible starting point:

• Auto: your lowest bid, often below the suggested range.
• Broad and phrase: near the bottom of the suggested range.
• Exact: near the middle of the suggested range for proven terms.

Then let the data move them. Raise bids on exact terms that convert well under your break-even ACoS, and lower bids on terms that spend without converting. See bid strategies without magical thinking for how to make those changes calmly.

Where BookAdsAI fits

BookAdsAI does not create campaigns or choose match types for you, and it does not harvest search terms into exact match. You set up the structure in the Amazon Ads console. Once a campaign is on your BookAdsAI allowlist, Autopilot can move keyword bids within fixed limits after reports land and pause the book's campaigns when the day's spend reaches your cap. The structure above gives Autopilot cleaner data to act on, because each keyword's results are not blurred by overlapping campaigns.

The short version

Broad finds, phrase narrows, exact controls. On a tight budget, keep most of your money in exact and phrase, use auto and broad as low-bid research, and move proven searches into exact every week with matching negatives behind them.