Google Ads Keyword Match Types: Broad, Phrase and Exact

Google Ads keyword match types help express which searches a keyword can be eligible to match, but they are not three permanent levels of “good” and “bad.” Broad, phrase, and exact match now use meaning and context in addition to the visible keyword notation. Choose among them from the search intent you want, the amount of discovery the account can review, the business risk of an irrelevant click, and the quality of the conversion signal.

The practical starting point is not a syntax mark. It is a clear campaign question: which customer need should this ad and landing page answer, which related searches are acceptable, and which searches are outside the offer? Google’s current keyword matching guide describes the platform behavior. For an implementation sequence, see how to set up Google Ads keyword match types.

What are the Google Ads keyword match types?

The three positive keyword match types are broad, phrase, and exact. Their notation is still visible in the keyword list, but the notation should be read as a statement about eligible meaning rather than a promise that the query will contain one exact character sequence.

Google’s current keyword matching documentation explains that the types overlap. Phrase can reach searches that exact can reach and more, while broad can reach related searches using context and signals. The account should therefore be organized around the meaning the ad can serve, not around a claim that one punctuation mark creates a perfect fence.

Broad match

Broad match is a discovery-oriented choice when the account can tolerate testing related demand and has enough context, conversion quality, and review capacity to learn from it. The keyword, ad group, landing page, campaign settings, and available signals help describe what the business considers relevant. A vague keyword in a mixed ad group can still invite the wrong interpretation.

Broad is not automatically careless, and it is not automatically the best starting point. A service with clear eligibility, reliable qualified conversion data, and a useful review process may benefit from discovering queries the planner did not write down. A small budget, ambiguous term, regulated offer, or weak conversion event may require a more conservative starting boundary.

Phrase match

Phrase match is useful when the account wants to preserve a defined concept while allowing relevant changes in wording, order, or surrounding meaning under Google’s current rules. It can be a practical middle position for a theme that is too narrow for a literal list but too risky to open broadly.

Phrase does not remove the need for an ad-group boundary. If one keyword can mean a paid service, a job search, a tutorial, and a free tool, phrase syntax does not decide which business meaning the ad should answer. Separate the themes or add a clearly justified negative where the query is outside the offer.

Exact match

Exact match is useful when a tighter set of closely related searches matters more than broad discovery. It can support a high-value service, a carefully defined product, a controlled test, or a campaign where the page and offer must stay close to one intent. It is still not a guarantee that only the literal characters inside the brackets will match; Google can use same-intent and close-variant behavior under its current rules.

Exact match is not a substitute for good structure. If the keyword is vague, the landing page is generic, or the conversion is unqualified, tighter matching may simply produce a smaller sample of the same measurement problem.

How should you choose a match type?

Choose the type from the relationship between search intent, ad promise, landing page, and business risk. A keyword is a useful control only when the surrounding campaign can answer the searches it reaches and can identify whether the resulting traffic is valuable.

Define the intent before the notation

Write the searcher’s question in plain language. “Buy replacement filters for model X,” “compare enterprise analytics platforms,” and “find free DIY instructions” may share words while representing different offers and economics. Decide which question the campaign serves, which questions are too early or too broad, and which are clearly outside the business boundary.

The keyword should be a compact representative of that intent, not the full strategy. Record the product or service, customer type, location, eligibility, landing page, and conversion that make the query valuable. This record helps the team decide whether a broader match is a reasonable discovery test or whether a tighter starting control is needed.

Match the ad group and page to the type

A match type cannot repair an ad group that combines unrelated intent. Group keywords when the same ad promise, page, offer, and conversion question can serve them. Separate a paid product from a job search, a local service from a national information query, or a high-value plan from a free-resource query when those differences change the customer experience.

For broad match especially, the landing page and ad group provide context for the meaning the account can serve. A generic page with several unrelated offers gives the system and the reviewer less useful information. Improve the structure before concluding that broad match itself caused irrelevant reach.

Treat the first type as a testable hypothesis

Do not write a universal rule such as “always use exact” or “broad always performs better.” State the intended trade-off: more qualified discovery, tighter intent control, a new theme, a safer experiment, or a simpler account structure. Define which evidence would justify keeping, narrowing, expanding, or pausing the choice.

A match-type change can alter reach, query mix, spend, conversion delay, and the learning context at the same time. Keep a change log with the keyword theme, effective date, surrounding changes, and review question. This makes a later result easier to interpret than a silent switch between punctuation styles.

How do negative keywords fit with match types?

Positive match types decide which searches a keyword may be eligible to reach. Negative keywords describe searches that should not receive a campaign or ad group’s ad. They are a separate control layer, and negative matching does not behave exactly like positive broad, phrase, and exact matching.

Google’s negative keyword documentation should be checked before applying a list. Google’s current guidance also notes that negative keywords do not automatically match close variants or other expansions in the same way positive keywords can. The exclusion therefore needs to be tested against the actual query category it is meant to block.

Add a negative for a clear business mismatch

A useful negative answers “why should this search not receive this ad?” The reason might be a different product, an unavailable location, a job-seeking query, a free or do-it-yourself request, or an audience the business cannot serve. Write the mismatch before adding the word. Do not exclude a term only because one query had a low click-through rate; the same word can appear in both relevant and irrelevant searches.

Search terms evidence is usually more dependable than a copied universal list. Review the queries that actually triggered the campaign, group repeatable mismatches, and choose the narrowest exclusion that solves the problem. Then check whether the negative is applied at the campaign or ad-group level that matches the business boundary.

Protect positive coverage

An exclusion can remove qualified demand if it is broader than the mismatch. For example, a word associated with a free resource may also appear in a paid consultation that the business does sell. A location name can be part of an ineligible service area—or the name of a product, customer, or comparison page that should remain eligible.

Review negative changes in both directions: did irrelevant queries decrease, and did valuable query coverage remain? Keep the source query, reason, match type, owner, and review date in the change record. Revisit old negatives when the offer, geography, product catalog, or customer definition changes.

Do not use negatives to repair mixed intent

Negative keywords are useful boundaries, but they should not carry the whole account structure. If one ad group combines unrelated products, audiences, or customer questions, a growing negative list can become difficult to reason about. Split the intent so the keyword, ad, page, conversion, and exclusion share one understandable purpose.

How should you evaluate Google Ads match types?

Evaluate a match type with the search terms it reaches and the business value those searches create. The account should answer four separate questions:

1. Did the keyword receive enough eligible reach for the intended test?
2. Were the searches relevant to the ad, page, product, and customer boundary?
3. Did the resulting visitors take the defined next action?
4. Did those actions become qualified value after sales, billing, or delivery review where applicable?

Google’s Search terms report guidance is the platform reference for reviewing searches associated with ads. Use the report with conversion tracking and the business system that owns the outcome; clicks and platform-attributed conversions alone do not explain search quality.

Classify queries by action, not by emotion

For each useful review, classify queries into a small set of decisions:

• Relevant and served well: keep the theme, ad, and page aligned.
• Relevant but poorly served: improve the ad group, landing page, offer, or conversion path before changing the match type.
• Relevant but in the wrong structure: move or separate the intent so the right ad and budget answer it.
• Clearly outside the offer: add a scoped negative or change the campaign boundary after confirming the pattern.
• Ambiguous: investigate the business meaning before blocking or scaling.

This classification avoids the common mistake of calling every unexpected query “bad.” A query may be unexpected but valuable, or familiar-looking but commercially wrong. The decision should be tied to the customer, offer, and outcome rather than to whether the words look like the keyword.

Measure quality beyond the click

Use the conversion that represents the business result, not only the event that is easiest to trigger. An ecommerce account may need completed orders and accurate value; a lead account may need accepted opportunities or revenue; a local service may need qualified calls or booked jobs. If the account bids on a shallow event, label that limitation in the review.

Separate reach, query relevance, engagement, conversion rate, cost, and qualified value. A narrower type can look efficient because it receives less varied traffic, while a broader type can find valuable themes that need new ads or pages. Neither conclusion is reliable without the outcome definition.

Review changes without confounding the result

Changing match type, keyword, negative list, ad, landing page, budget, bid strategy, and conversion action together makes the report difficult to read. Record the theme, type, effective date, intended hypothesis, and surrounding changes. Review the results after the relevant conversion delay and business cycle rather than reacting to one query or one short reporting period.

If the problem is weak context, improve the ad group or page. If it is a clear business mismatch, use a negative or structure change. If the account needs more qualified discovery and the signal is trustworthy, a broader test may be reasonable. Match type should be the smallest change that addresses the diagnosed problem.

What are common Google Ads match-type mistakes?

• Treating syntax as a literal filter: current matching uses meaning and context, so punctuation does not guarantee one exact query string.
• Using broad match without useful context: a mixed ad group, vague page, or shallow conversion event can make discovery difficult to steer.
• Using exact match for every goal: tighter control can also hide valuable variations and make it harder to discover new demand.
• Mixing unrelated intent: one keyword group should not force the same ad and page to answer product, employment, research, and free-resource queries.
• Copying a negative list blindly: a broad exclusion can block qualified demand, while a narrow one can fail to block a repeatable mismatch.
• Confusing an unexpected query with a bad query: judge the customer need, offer, and qualified result, not just the word overlap.
• Changing every surrounding setting at once: a match-type result is hard to explain when ads, pages, budgets, bids, and conversions all change.
• Measuring clicks instead of value: traffic quality needs conversion and downstream business evidence.

Google Ads keyword match-type checklist

Before adding or changing a match type, confirm:

• The customer intent and campaign boundary are written in plain language.
• The keyword, ad group, ad, landing page, and conversion answer the same question.
• The choice between broad, phrase, and exact states its reach-and-risk reason.
• The account can review search terms and has an owner for negative changes.
• Negative keywords target proven mismatches without blocking positive coverage.
• The conversion signal represents a meaningful business action and can be reconciled with the system that owns the outcome.
• The change log records the hypothesis, effective date, and other settings changed at the same time.

For deeper work, read how to optimize Google Ads match types, common Google Ads match-type mistakes, and Google Ads negative keyword best practices.

Sources

• About keyword matching options
• About negative keywords
• About the Search terms report