Keyword & Query Matching

Broad Match vs Phrase Match: How Wide to Buy

In short: Broad match treats the keyword as a topic hint and matches anything Google judges related, including adjacent subjects the keyword never named. Phrase match keeps the keyword's meaning and its implied word order while allowing words either side. Broad needs automated bidding and a serious negative list; phrase survives either bidding model.

By the AdFlint research team · Fact-checked against current Google and Meta platform behavior · Last reviewed July 2026

Broad Match

Matches a query to anything Google judges related to the keyword's meaning, including synonyms, misspellings, and adjacent topics the keyword never named.

Google reads the keyword as a topic hint rather than a literal string, then uses the landing page, the other keywords in the ad group, and account history to decide which queries qualify. Advertisers reach for it to feed automated bidding more volume once conversion tracking is solid. The usual mistake is running it on manual bids with a thin negative list, which buys wide irrelevant traffic quickly.

Full definition

Phrase Match

Matches queries that carry the keyword's meaning while preserving its implied word order, allowing extra words before or after it.

Since the 2021 rewrite, phrase match keys on meaning plus sequence rather than exact strings, so it absorbed most of what Broad Match Modifier used to do. Accounts use it as the middle setting: tighter than broad, wider than exact, and workable under manual or automated bidding. The common misread is assuming words must appear verbatim, when close variants and reordered paraphrases also qualify.

Full definition

Side by side.

The differences that actually change what happens in your account.

 Broad MatchPhrase Match
What it matchesRelated meanings, synonyms, adjacent topics.The keyword's meaning with its word order preserved.
Extra context usedLanding page, other keywords in the ad group, account history.Query meaning and sequence.
Bidding pairingNeeds automated bidding to price query quality.Workable under manual or automated bidding.
Negative keyword loadHigh and never finished.Moderate.
Volume vs controlWidest reach, least predictable.The middle setting, and the usual default.
Failure modeBroad on manual bids with a thin negative list.Assuming words must appear verbatim, when paraphrases also qualify.

What actually separates them.

01

Broad match uses signals beyond the query - your landing page, the ad group's other keywords, account history - to decide eligibility; phrase match does not reach that far.

02

Phrase match preserves the keyword's implied word order, so it excludes reorderings that change meaning in ways broad match will accept.

03

Since the 2021 rewrite, phrase match absorbed Broad Match Modifier's behavior, which is why BMM was retired rather than replaced.

04

Broad match without automated bidding has no mechanism to pay less for a poor query, which is the single most expensive misconfiguration in this pairing.

05

Both need search terms review, but broad match keeps producing new query shapes, so its negative list is maintenance rather than setup.

Which one should you use?

Use Broad Match when

  • Conversion tracking is verified and the campaign runs on automated bidding.
  • You want the bidder to find query variants your keyword list never anticipated.
  • Volume has plateaued on tighter match types.
  • Someone genuinely reviews search terms and maintains negatives on a schedule.

Use Phrase Match when

  • You need meaningful reach without handing query selection entirely to Google.
  • The campaign is on manual bids or has thin conversion volume.
  • Your product language is specific and adjacent topics are not relevant.
  • You want readable per-keyword data while still capturing longer variants.

Common questions.

Is broad match still risky?

It is risky in the same specific way it always was: paired with manual bids and a weak negative list, it buys wide, cheap, irrelevant traffic quickly. Paired with accurate conversion tracking, automated bidding, and disciplined search-term review, it is how most accounts now find volume beyond their own keyword list.

Do I need broad, phrase, and exact versions of the same keyword?

Rarely, and it fragments your data. Duplicate match types split conversions across keywords that all needed volume to learn from, and Google's matching already overlaps heavily. Most accounts do better choosing one match type per intent and letting negatives, rather than duplication, decide which campaign serves.

How often should I review search terms on broad match?

Weekly while a campaign is new, then on a fixed cadence you will actually keep. Broad match keeps generating query shapes you have not seen, so negatives are ongoing maintenance. The moment nobody is reviewing terms is the moment broad match stops being a volume strategy and starts being a leak.

Or stop choosing between them.

AdFlint picks the setting, writes the ads, and keeps optimizing inside the Google and Meta accounts you already own.

Related comparisons

All comparisons