Amazon Match Types: Broad, Phrase and Exact, Used Properly

Key takeaways: Broad match finds search terms you did not know about, phrase match tests a term's variations, and exact match gives a proven term its own bid. Keep each match type in its own campaign, promote terms from broad to exact when they prove they convert, and negate the promoted term in the campaign it came from. Bid lowest on broad, highest on exact, and never let two match types compete for the same search.

Amazon's three match types are not three levels of the same thing. They are three tools with different jobs. Broad match casts wide, matching your keyword to related searches including synonyms and reordered words, and it is how new search terms are found. Phrase match keeps your words in order and adds words before or after, which tests whether a concept converts across its variations. Exact match limits your ad to the term itself and close variants, which is where a proven term gets its own bid and its own data. Used properly, a keyword moves from broad to exact as it proves itself, and the match types never compete.

How each match type matches

Amazon's broad match has widened over time and can include synonyms and related concepts, not just your words in any order. That makes it a stronger discovery tool and a weaker control tool than it was. Broad match modifiers, where a plus sign locks a word in, give some control back where available.

Exact match includes plurals, common misspellings and minor variants. It does not include added words. That precision is the point: when you bid on an exact term, you know what search you are paying for.

The job of each

Broad: find. A broad-match campaign with modest bids and a capped budget is a search-term generator. Its job is not to produce sales efficiently; it is to produce search terms that then get judged on their own performance. It should be read weekly through the search term report and it should always be losing some money on terms that turn out to be irrelevant. That is what discovery costs, and keeping it inside a fixed share of spend is how the cost stays reasonable, as described in D20 · Discovery campaigns wasting spend.

Phrase: test. Phrase match sits between the two and is the most often skipped. Its value is in testing a concept's variations without opening the door as wide as broad. If insulated water bottle converts in exact, a phrase campaign on it will show whether insulated water bottle for kids, insulated water bottle with straw and large insulated water bottle also convert, and which ones deserve their own exact keyword. Phrase is a cheaper test than broad and a wider one than exact.

Exact: control. An exact-match keyword has one bid, one placement setting through its campaign, and a search term report that shows only that term. It is where rank is built on a head term and where profit is managed on a proven one. Everything that has earned its place ends up here.

How a term moves through the match types

The promotion threshold is a rule you set once. Two or three orders with ACoS inside the band over a rolling 30 to 60 days is a common working rule. The negative step at the end is what stops the match types competing; without it, the broad campaign keeps bidding on the term at a bid never meant for it. This is the routine behind B14 · How a search term earns its own campaign.

Bidding by match type

Bids should rise with precision. Broad match bids are the lowest, because the matched search may be only loosely relevant and the campaign's job is to find rather than to win. Phrase bids sit in the middle. Exact bids are the highest, because you know what you are buying and have data to set the bid from.

Setting the same bid across all three, or bidding higher on broad because "it gets more traffic", inverts the logic. You end up paying the most for the least relevant clicks and starving the terms you know convert.

Mistakes that make match types compete

All three match types for the same keyword in one ad group. Amazon picks whichever match it prefers for a given search, and the exact term never gets its own clean data or its own bid.

No negatives after promotion. The broad campaign and the exact campaign bid against each other for the same search, and the account pays more than either bid alone.

Judging broad on ACoS. Broad's job is to find terms. Judge it on how many new exact keywords it produces per month and on its share of spend, not on its own ACoS, which will always be worse than exact.

Skipping phrase entirely. Going straight from broad to exact misses the middle step that tells you which variations of a concept deserve their own keyword, and it leaves the exact list narrower than it should be.

Exact keywords with no broad or auto feeding them. An account that is all exact match has stopped discovering. The keyword list ages, new search behavior is missed, and competitors find the new terms first.

All of this depends on each match type having its own campaign, which is the structure covered in Amazon PPC Campaign Structure That Scales With Your Catalog.

B14How a search term earns its own campaigncontrol of cost per click, spend and rank on every keyword that has proven it converts.

B15The weekly search-term decision enginea weekly, rule-based clean-up — scale what works, cut what does not — with every change logged in your report.