Amazon keyword match types determine how broadly a manual keyword can match shopping queries. Broad, phrase, and exact offer different boundaries, including eligible close variations. They control matching, not profitability; use search-term reports and product economics to decide which traffic deserves continued investment.
Amazon's match-type documentation includes close variations and explains that automatic campaigns use targeting groups instead. Use the distinctions to organize controlled tests, then let actual search-term evidence guide the next decision.
Amazon broad match: research shopper language
Broad match can reach related language beyond a literal copy of the keyword. That creates room to discover how shoppers describe a need, but it also creates a larger review burden. Set a learning budget you can afford without assuming every query will convert.
For an illustrative travel mug account, a broad seed could reveal interest in commuting, insulation, or a specific lid style. Read those queries for intent. Do not add every term to another campaign simply because it generated a click; decide which question the evidence answers.
Amazon phrase match: explore a more defined need
Phrase match constrains the relationship between words while allowing additional query context. It can be useful when the product need is established but modifiers still matter. Check Amazon's current rules and actual queries rather than importing assumptions from another advertising platform.
A phrase-oriented test might explore size or use-case modifiers around an established product description. Group the evidence into questions your detail page can answer: dimensions, compatibility, quantity, or material. That makes the learning useful for both targeting and the product offer.
Amazon exact match: control bids on deliberate targets
Exact match is a tighter targeting choice, not a guarantee of one literal spelling or profitable conversion. Use it when a query deserves its own bid or review priority. Check reporting for the variations that actually served.
A query with repeated orders and acceptable economics may deserve more deliberate control. One order on a small sample is weaker evidence. Compare the proposed spending with a break-even ACoS model, and retain an observation bucket when the evidence cannot yet support a confident change.
Review the search terms behind each match type
Several targets can be eligible for related shopping queries. Account organization alone does not guarantee the routing you expect. Inspect where spend and orders actually appear, rather than assuming a three-match-type structure will separate demand automatically.
Keep a record of the seed, observed query, advertised product, match type, mature results, and action. Our campaign-structure guide explains how to make discovery and controlled demand readable. Use negative keywords when a specific relevance or routing decision justifies them.
- Choose a purpose for each test.
- Review query relevance and economics together.
- Avoid multiplying targets without a distinct decision to make.
Choose match types for the amount of exploration and control you need. A useful test improves your understanding of shopper intent and supports a clearer next bid decision.
Source material
Amazon Ads: keyword match types and automatic targeting groups
