Lexicon entryPaid search · Account structure

Match Types

The rules that tell Google which search queries are close enough to your keyword to trigger your ad. The balance between reach and precision is the single largest determinant of wasted spend in most accounts.

Chris Gardner — Digital Marketing Consultant, USA TODAY NETWORK | LOCALiQ. Founder, Signostic. Windsor, ON. Last updated 20 April 2026 · About · LinkedIn · Email

Direct answer

Match types are the rules that tell Google Ads which search queries are close enough to your keyword to trigger your ad. The three current types — broad match, phrase match, and exact match — trade reach against precision, and the balance directly determines how much wasted spend sits in your account.

What it is

The reach-precision tradeoff.

Match types are not a technical setting. They are a commercial decision about how much demand you want to chase versus how tightly you want to control what you pay for.

Every keyword you add to Google Ads is paired with a match type, either explicitly (broad, phrase, exact) or by Google’s default (broad, unless specified otherwise). The match type determines the breadth of queries for which your ad is eligible to appear. Exact match triggers only on queries close to the keyword itself. Phrase match triggers on queries that contain the meaning of the keyword. Broad match triggers on queries Google considers semantically related, which in 2026 includes substantial machine-interpreted context.

The commercial consequence is direct. Broad match without disciplined negative-keyword hygiene routinely produces 15–25% wasted spend in SMB accounts, because Google is incentivised to expand match coverage and charges for clicks from queries that are categorically irrelevant. Exact match eliminates most of that waste at the cost of volume. The right mix is almost never all-broad or all-exact — it is a portfolio decision auditable at the search-term-report level.

How it works

The three match types, ranked by precision.

Google Ads has consolidated its match-type framework to three options as of the 2024 platform update, with broad match redesigned as a machine-learning-driven expansion mechanism rather than the older rules-based definition. Here is how each behaves in practice.

Exact match (keyword enclosed in [square brackets]) matches queries that share the same meaning as the keyword, with Google’s algorithm interpreting close variants — misspellings, plural/singular, close synonyms, same-intent rewordings. A [bookkeeper toronto] exact-match keyword will match “toronto bookkeeper,” “bookkeepers in toronto,” and “best bookkeeper toronto” — but not “accountant toronto” or “bookkeeping software.” Exact is the tightest control.

Phrase match (keyword enclosed in “quotation marks”) matches queries that include the meaning of the keyword, with additional words allowed before or after. A “hvac repair toronto” phrase-match keyword will match “emergency hvac repair toronto,” “toronto hvac repair services,” and “commercial hvac repair downtown toronto” — but not “hvac toronto repair company” if Google interprets the additional context as altering meaning. Phrase is the middle precision tier.

Broad match (keyword with no symbols) is now machine-learning-driven and expands most aggressively. A plumber toronto broad-match keyword can match “leaky faucet fix toronto,” “water heater replacement,” “drain cleaning near me,” and many queries Google judges semantically related. Broad delivers the widest reach and requires the most active negative-keyword discipline to remain efficient.

Negative keywords (keyword prefixed with a minus sign) are not a match type but are often paired with them. Negatives exclude specified queries from triggering ads regardless of which match type is running. In high-volume accounts, a weekly negative-keyword review is essential — particularly in broad-match campaigns where Google’s query-expansion logic otherwise drifts.

Why it matters · SMB operator view

The wasted-spend dimension.

Match-type discipline is the fastest source of recoverable budget in most SMB accounts we audit. Unlike Quality Score improvements that compound over weeks, match-type fixes produce immediate CPC and CPL effects on the next day’s spend.

Insurance. Insurance keywords are especially vulnerable to broad-match drift because the vocabulary is dense with adjacent-but-different meanings: “insurance” matches home, auto, commercial, life, health, and pet when the underlying business only writes personal lines. A commercial broker running broad-match on “auto insurance” without negatives for “cheap,” “quote,” “rental,” “temporary,” and dozens of consumer-intent modifiers typically wastes 20–30% of monthly spend on traffic that will never close.

Retail. Retail keywords suffer a different failure mode. Product-related broad-match keywords routinely trigger on informational queries (“how to” searches, comparison searches, used-market searches) that indicate no purchase intent. A furniture retailer running broad on “modern sofa” without negatives for “diy,” “plans,” “how to make,” “used,” and “kijiji” is paying retail CPCs to capture non-commercial intent. The search-term report exposes this cleanly in a single afternoon of review.

In every vertical, the audit is the same: pull 90 days of search-term data, categorise queries as relevant, marginal, or irrelevant, and quantify the spend sitting in the third category. That number is usually the single largest recoverable expense in the account.

Common misunderstandings

Three things operators get wrong.

Myth

Broad match is dangerous and should be avoided.

Fact

Broad match is not intrinsically dangerous; unmanaged broad match is. Broad match with disciplined negative keywords, smart bidding, and weekly search-term reviews can outperform phrase match on volume at equivalent CPL — and often does, because Google’s machine learning has more signal when match type is broad. The real problem is accounts that run broad match and never review the search-term report.

Myth

Exact match is always the safest choice.

Fact

Exact match is the most precise, but Google’s “close variant” logic means exact-match keywords now trigger on reworded queries and close synonyms — behaviour closer to 2018-era phrase match. Relying on exact as a “safety” setting without auditing search terms still produces waste; just less of it than broad match.

Myth

Modified broad match is still available.

Fact

Modified broad match (+keyword +terms) was merged into phrase match in 2021 and no longer exists as a separate match type. Accounts carrying legacy modified-broad keywords now run as phrase match. If your last match-type audit was before 2022, your account structure probably reflects an older framework.

Frequently asked

Match types, answered.

What are the match types in Google Ads?

As of the 2024 platform update, Google Ads has three active match types: exact match (keyword in square brackets), phrase match (keyword in quotation marks), and broad match (keyword with no symbols). Each trades reach against precision differently. Negative keywords (keyword prefixed with a minus sign) are not a match type but work alongside all three to exclude irrelevant queries. Modified broad match was retired in 2021.

Which match type is best?

No single match type is universally best. Exact match is appropriate for high-intent, high-CPC keywords where every click matters. Phrase match works well for mid-funnel terms where controlled expansion is useful. Broad match with smart bidding and aggressive negative-keyword management works well in accounts with strong conversion tracking and sufficient volume to train Google’s algorithm. Most mature accounts run a mix, not a single type.

What is broad match in 2026?

Broad match in 2026 is a machine-learning-driven query-expansion system, not the rules-based matching of earlier years. Google uses the keyword as a semantic anchor and expands to include queries it judges as related based on search context, landing-page signals, and campaign targeting. This can produce strong volume gains in trained accounts, and substantial waste in untrained accounts. The determining factor is negative-keyword discipline.

Are negative keywords a match type?

No. Negative keywords are an exclusion mechanism that works across all three match types. Adding a negative keyword prevents your ad from triggering on queries containing that term, regardless of whether your underlying keywords are broad, phrase, or exact. Negative keywords have their own match types (exact, phrase, broad) that determine how strictly the exclusion applies. Weekly negative-keyword review is essential in any account running broad match.

Do match types still exist in Performance Max?

Performance Max does not use keyword-level match types in the traditional sense — the campaign is driven by audience signals, creative assets, and landing pages rather than keyword lists. However, Performance Max does respect account-level negative keyword lists (a 2023 platform addition), so negative-keyword discipline transfers from Search to Performance Max even without explicit match types.

Can match types be mixed in the same ad group?

Yes, and many accounts run the same keyword in multiple match types (exact + phrase + broad) within separate ad groups to give Google distinct signals. This practice — sometimes called match-type layering — allows the advertiser to bid differently by precision level and gather cleaner data on which match type is driving conversions. It requires more maintenance but provides better audit visibility than single-match-type campaigns.

Primary sources

Where this definition comes from.

Referenced in this entry

Google Ads Help. About keyword matching options. 2025. support.google.com/google-ads/answer/7478529
Google Ads Help. About search terms report. 2025. support.google.com/google-ads/answer/7531771
Search Engine Land. How Google’s 2024 match-type changes affect spend efficiency. 2024.
WordStream. Match Types and Negative Keywords: The 2025 Playbook. 2025.

Get a diagnosis

If your search-term report is showing broad-match drift and you want a diagnostic that quantifies exactly how much spend is going to irrelevant queries, Chris Gardner reads every audit personally. No templates. No generic recommendations. A diagnostic built on your account data.