Match Types
Google Ads match types determine how closely a search must relate to a keyword before the ad can enter the auction. The positive keyword options are broad match, phrase match, and exact match, but none should be read as a literal promise about the words a person will type.
Why it matters
Match type decides how much discovery Google can do beyond the keyword list. That choice changes the search terms an account buys, how quickly budget can move into unfamiliar intent, and how heavily the campaign depends on conversion data and negative keywords. It is not a one-time setup label. Operators need to review the search terms produced by each keyword and change the setting when the account's evidence, offer, or risk tolerance changes.
What people get wrong
The biggest mistake is believing exact means identical, phrase means the phrase must appear exactly, and broad simply adds more words. Google evaluates meaning, intent, context, close variants, and other signals. Exact is the narrowest option, but it can still match a differently worded search. Phrase preserves a concept rather than a fixed string. Broad can move well beyond the visible keyword. The only dependable description of a match type is the search terms it actually bought in that account.
Broad, phrase, and exact match compared
Positive match-type syntax is simple: broad keywords have no punctuation, phrase keywords use quotation marks, and exact keywords use square brackets. The operational differences are not simple because eligibility is based on meaning rather than a mechanical text rule. The comparison below is a starting model; the search terms report is the account-specific record of what Google allowed through.
| Match type | Syntax | What it matches | What can surprise you | When to use it |
|---|---|---|---|---|
| Broad | emergency plumber | Queries Google relates to the keyword's meaning, using account and auction signals | Searches can omit the visible keyword words or approach the service from a different intent | When conversion data is trustworthy, negatives are maintained, and value-based bidding can distinguish useful outcomes |
| Phrase | "emergency plumber" | Queries that include the keyword's meaning, with room for added context and reordered language | The literal phrase does not need to appear, and a close interpretation may pass | When an operator needs discovery around a defined service concept without opening the full reach of broad |
| Exact | [emergency plumber] | Queries Google considers the same meaning or intent as the keyword | Close variants, reordered language, implied words, and differently phrased intent may trigger | When control matters, evidence is limited, or a high-value concept needs a narrow starting boundary |
Exact match now means same intent, not same text
Exact match began as a much more literal control than it is now. Close variants expanded from misspellings and plural forms into reordered words, function words, implied terms, and queries Google classifies as sharing the same meaning. An exact keyword for emergency plumber can therefore enter an auction for urgent plumbing help even though the visible words differ. That may be useful, but it means the brackets do not create a sealed list of query strings.
Treat exact match as a controlled discovery setting. Review the queries, separate useful variants into their own keyword or ad-group structure when they deserve different copy, and block interpretations the business does not serve. If a team promises that brackets buy only the bracketed text, the promise is already outdated.
The search terms report is the real definition
The keyword table shows what the advertiser entered. The search terms report shows what people typed before an ad interaction. Reading the two together reveals what each match type is buying. Start with spend and conversion columns, then inspect the query itself, the matched keyword, campaign, ad group, and match-type classification. A keyword can look efficient in aggregate while a small set of relevant terms carries the useful activity and a long tail consumes the rest.
The report is also where an operator decides whether a match-type change is justified. If phrase match keeps reaching unrelated concepts despite a maintained negative list, exact may provide a better boundary. If exact and phrase repeatedly reveal valuable variations and the account has reliable outcome data, a controlled broad test may be reasonable. Make the decision from query evidence, not from the platform's recommendation score.
When broad match is safe enough to test
Broad match asks the bidding system to decide which related searches are worth entering. It is safest when the campaign has a trustworthy primary conversion, enough recent outcomes for automated bidding to distinguish patterns, a negative-keyword process, and values that separate strong outcomes from weak ones. A raw form fill that includes spam, job seekers, and unqualified requests is not a strong signal merely because it occurs often.
Broad is not a sensible default for a new account with uncertain tracking, a tiny service territory, a narrow offer, or a conversion action that fires before genuine intent is confirmed. In those conditions, phrase or exact provides a smaller area to inspect while the operator builds a clean search-term history. The free conversion tracking audit can check whether core Google Ads and analytics tags are present before broader matching is trusted with the account's feedback loop.
How match types interact with Performance Max
Performance Max does not use a conventional positive keyword list. Search themes can guide it, but they are signals rather than the same eligibility controls used by Search campaigns. When a Search keyword and a Performance Max campaign are both eligible for related demand, campaign priority depends on eligibility, query identity, rank, settings, and Google's auction logic. The result can be Search and Performance Max drawing from the same pool of intent while each reports performance through a different view.
Operators should inspect Performance Max search-term insights alongside the Search terms report, use brand exclusions where appropriate, and apply campaign-level negative keywords when available. The detailed process is covered in the negative-keyword script guide for Search and Performance Max. A script can keep lists synchronized, but it cannot decide the commercial boundary. That decision still belongs to the person who knows which terms deserve dedicated Search coverage.
Negative match types follow different rules
A broad negative blocks a query when all words in the negative are present, even if their order changes. A phrase negative blocks the phrase when those words appear together in that order, with other words allowed around it. An exact negative blocks the query only when the query matches that negative phrase as a whole. Those rules make negative broad more restrictive than many operators expect and negative exact narrower than its positive counterpart.
For example, if the negative keyword is free plumbing course, broad negative can block a query containing all of those words in another order; phrase negative blocks queries containing free plumbing course in that order; exact negative blocks only the complete query free plumbing course. Variations that omit a word or use a synonym may remain eligible. Build exclusions from observed search terms and check for conflicts with services the business does want. The related negative keywords glossary entry explains the maintenance process in more detail.
One keyword through all three match types: a hypothetical
For example, if a Phoenix plumbing company bids on emergency plumber, the broad version could be eligible for searches about urgent plumbing help, a burst pipe company, after-hours drain service, or other queries Google relates to the concept. Some may fit; others may describe work, timing, or geography the company does not accept. Broad gives the system the widest room to use conversion and auction signals.
For example, if the same company uses phrase match for emergency plumber, Google can still allow added context and wording that preserves the concept, such as emergency plumber for a leaking pipe or plumber available for an emergency. The literal phrase is not required. The boundary is tighter than broad, but the operator still has to inspect whether Google's understanding of emergency plumbing matches the offered services.
For example, if the company uses exact match for emergency plumber, queries with the same meaning can still trigger even when the words are reordered or replaced by a close expression. Exact reduces exploration; it does not limit delivery to one typed string. If the business does not offer drain clearing, commercial emergencies, or after-hours calls, those constraints still need campaign structure, copy, schedules, geography, and negatives.
A practical operating cadence
During review, read search terms by spend and business category, not only by the platform's relevance labels. Add negatives for clear exclusions, promote valuable themes when they need dedicated copy or budgets, and record why a match type changes. Compare Search with Performance Max so a movement between campaigns is not mistaken for new or lost demand.
Match types are useful controls, not identities. Exact can discover. Phrase can stretch. Broad can be disciplined when the feedback loop is strong. The operator's advantage comes from making the boundaries explicit, checking what passed through them, and connecting the resulting queries to qualified outcomes instead of trusting the label printed beside the keyword.
Frequently asked
What are the three Google Ads match types?
The positive keyword match types are broad, phrase, and exact. Broad allows the widest interpretation of a keyword, phrase keeps closer to the keyword's meaning, and exact is the narrowest. All three use meaning and intent, so none is a literal text-only rule.
Does exact match only show for the exact keyword?
No. Exact match can show for close variants and searches Google considers the same meaning or intent, including reordered or differently worded queries. It offers the tightest positive-keyword control, but the search terms report is still required.
Should I use broad, phrase, or exact match for a new campaign?
Phrase or exact gives a new campaign a narrower starting boundary while it establishes reliable conversion and search-term data. Broad is safer to test after the primary conversion is trustworthy, exclusions are maintained, and automated bidding has useful outcome signals.
When is broad match safe to use?
Broad match is safest when conversion tracking represents qualified outcomes, automated bidding has current evidence, negative keywords are actively reviewed, and the campaign has clear service and geographic limits. It is risky when the account is new or the main conversion counts weak actions.
How do I see what a match type is actually matching?
Open the search terms report and compare each query with its matched keyword, campaign, spend, and resulting business outcome. Classify terms by accepted service and buyer intent rather than relying only on Google's match-type label.
Do keyword match types control Performance Max?
No. Performance Max does not use the same positive-keyword controls as Search. Search themes guide it, while brand exclusions, available campaign negatives, account structure, and Google's auction rules determine how it overlaps with Search demand.
Do negative keywords use the same match-type rules?
No. Negative broad, phrase, and exact are literal exclusion patterns and do not automatically expand to every close variant or synonym. Negative broad requires all words, negative phrase requires the ordered phrase, and negative exact blocks only the complete query.
Should the same keyword run as broad, phrase, and exact?
Not by default. Running every version can fragment reporting and does not create three isolated auctions. Give one match type responsibility for the concept, inspect its search terms, and widen or narrow it from evidence.