Skip to main content

Keyword Match Type Tester

What it does

The Match Type Tester takes broad, phrase, and exact match keywords plus a search query, and reports which would match the query under Google Ads’ current match-type semantics. The output is a per-keyword pass/fail, with summary counts per match type. Useful for understanding why a keyword is or isn’t picking up a query, and for testing keyword expansion before launching.

Common situations

You’ve added a new keyword and want to know what queries it’ll trigger. Test a few representative queries against the keyword in different match types: the broader the match, the more queries it triggers, but with less control.

A search term in your search query report is being matched by a keyword you didn’t expect. Test the term against the keyword to confirm the match, then add a negative keyword to block the unwanted variant.

You’re moving keywords from broad match to phrase match (a common quality-improvement step) and want to verify the phrase versions still match the queries you care about. Test representative queries; if any expected matches fail, the phrase needs to be looser or you need to keep the broad version.

You’re auditing why a competitor’s exact-match keyword is appearing on queries that don’t match the literal exact phrase. Google’s “exact match” includes close variants (plurals, word order, function-word changes); the tester demonstrates this behaviour.

You’re explaining to a client why broad match catches more variants than phrase match. The tester is a visual demonstration: paste their queries, show the same keyword set in each match type, see how broad picks up loosely-related variants while exact only matches close-form variants.

What you need to know

Google Ads has three match types, each defining a different breadth of what queries will match a keyword. Understanding the difference is fundamental to campaign control:

Broad match is the loosest. The keyword matches any query Google’s system judges relevant, including synonyms, related searches, intent matches, and “close variants” of all kinds. Since Google’s 2021 expansion of broad match, this is much broader than it sounds. A broad-match keyword “running shoes” might trigger ads for “best trainers for jogging” or “athletic footwear” or “marathon gear” depending on Google’s intent inference.

Phrase match (denoted by quotes: “running shoes”) matches queries that contain the keyword phrase in order, with close variants. The phrase can have words before or after: “buy running shoes online”, “best running shoes 2025”, “running shoes review” all match. Order matters; “shoes for running” might or might not match depending on how strictly Google’s parser is treating order in that auction.

Exact match (denoted by brackets: [running shoes]) matches the keyword phrase or its close variants: plurals, word order rearrangements, function-word changes, common misspellings. So [running shoes] matches “running shoes” but also “shoes running” (word order), “running shoe” (singular), “the running shoes” (function word added). It does NOT match “trail running shoes” or “best running shoes”. Those are different queries.

The tester’s heuristic implementation:

  • Broad: every word in the keyword should be present (or have a close prefix match) in the query. Heuristic for “intent inference” since the actual algorithm is private.
  • Phrase: the keyword phrase appears in order in the query, with close-variant tolerance.
  • Exact: the query has the same set of words as the keyword (set comparison after normalisation), allowing close variants.

Real Google match-type behaviour is more nuanced: it includes semantic relevance, query rewriting, and machine-learned intent inference that can’t be fully replicated locally. The tester is directional. It catches the obvious matches and obvious non-matches but may differ from real Google behaviour on edge cases.

The 2021 broad match expansion made broad match much broader than it had been. Previously, broad match was constrained to plurals and close synonyms. Now it’s a semantic intent matcher. If you’ve been away from Google Ads for a few years, the broad match you remember is not the broad match that exists today.

Frequently asked questions

What’s the practical difference between broad and phrase match?

Phrase match is constrained to queries containing the keyword phrase. Broad match has no such constraint; Google can match any query its system judges intent-relevant. In practice, phrase gives you control over which queries trigger ads, while broad gives Google that control. For most accounts, phrase is the safer default.

Why does Google recommend broad match?

Because broad match generates more impressions (and more spend) for Google. Google’s recommendation reflects their incentives. The pragmatic answer for most advertisers: start with phrase match for control, expand to broad when you have data showing where the spend produces conversions.

Are exact-match keywords still tightly controlled?

Less than they once were. Modern exact match includes close variants: plurals, word reordering, function word changes, common misspellings. So [running shoes] matches “running shoe”, “shoes running”, “the running shoes”. It still doesn’t match unrelated terms like “trail running shoes”; those would be picked up by phrase or broad versions of the keyword.

What was modified broad match (the +modifier)?

A retired match type. Modified broad (e.g. +running +shoes) was Google’s middle option between broad and phrase, requiring all modified words to appear. Google retired it in February 2021, folding its functionality into phrase match. If you have legacy +modifier campaigns, they were converted automatically.

Can the same query match multiple match types?

Yes. A query can match all three simultaneously if you have keywords in all three match types. The auction picks one based on relevance, bid, and Quality Score. The keyword chosen is reported in your search term report.

How does this help me avoid wasted spend?

By revealing which match types are picking up irrelevant queries before you launch. If broad match for “running shoes” matches “tennis trainers” in your testing, broad is too loose for this keyword: switch to phrase, or add negative keywords to block the variants you don’t want.

Why is the tester sometimes inconsistent with my actual Google Ads search query reports?

Because Google’s matching algorithm uses signals (semantic similarity, intent inference, account history) that the tester can’t observe. The tester catches the deterministic cases; for the marginal cases, real data is the canonical source.

Do close variants apply to all match types?

Yes. Google’s “close variants” (plurals, word order, function word changes, etc.) apply to phrase and exact match. Broad has been close-variant-tolerant since its inception. The 2021 update extended close variants to include semantic similarity for broad match specifically.

Common problems

Problem: Search term report shows a query that doesn’t seem to match my keyword.

Google’s broad match expansion includes intent matching that isn’t visible from outside. A keyword “running shoes” might match “athletic footwear” through Google’s intent model, even though no shared tokens exist. Add the unwanted query as a negative; Google may continue matching similar queries until enough negatives accumulate.

Problem: Phrase match keyword isn’t matching queries I expected.

Phrase requires the phrase in order. “Running shoes” (phrase) won’t match “shoes for running” because the order differs. Use broad match for those variants, or phrase versions in different word orders.

Problem: Exact match keyword matched a query that’s clearly different.

Close variants for exact match include word order, plurals, and function words. If the surprise match is one of those (e.g. [running shoes] matched “shoes running”), it’s working as designed. If the surprise is a genuinely different query, it’s likely Google’s close-variant tolerance has expanded. Review and add negatives if needed.

Problem: Switching from broad to phrase dropped impressions by 80%.

That’s the expected magnitude. Broad match’s intent-matching captures many more variants than phrase. The question is whether the dropped impressions were converting; usually phrase keeps the high-converting variants and drops the noise. Real conversion data after the switch confirms or refutes.

Problem: Tester says match but real Google Ads didn’t show ad on that query.

Match type determines eligibility, not impression. An eligible keyword can still lose the auction (low bid, low Quality Score, exhausted budget) or be filtered (audience targeting, geographic restriction, time-of-day rules). The tester only checks match-type matching; real Google Ads has many more filters.

Tips

  • Start with phrase match by default. Move to broad only when you have data showing conversion patterns warrant the expansion.
  • Use the tester to check expansion variants before adding them. If a new broad keyword matches queries you don’t want, add the variants as negatives at the same time.
  • Audit search term reports weekly. Real query data reveals match-type issues that pre-launch testing can miss.
  • Don’t conflate match type with intent matching. Phrase match doesn’t always preserve intent. “Buy running shoes” and “running shoes review” both match phrase “running shoes” but represent very different intents.
  • The 2021 broad match changes made broad much broader than older guides describe. If your match-type intuition is from before 2021, retest your assumptions.

Related tools in this suite

The Negative Keyword List Builder is the natural follow-up: once the tester reveals match-type behaviour, negatives are how you constrain the unwanted variants. The Quality Score Estimator provides upstream context, since match type contributes to ad relevance scoring per query. Both sit within the Paid Search Suite.

What this looks like at scale

For a single keyword, the tester is fine. For an account with thousands of keywords across many match types, match-type discipline should be structural: naming conventions per match type, automated negative-keyword sharing across campaigns, and search-term-report reviews built into the regular optimisation workflow. That’s part of what an ongoing paid search engagement covers. The Knowledge Center has more background on campaign structure principles.

Take it further

If your account has chronic search-term issues (many irrelevant queries triggering ads, or expected queries not matching), the structural fix is usually a match-type and keyword-organisation review. Talk through the account and we can scope what the audit looks like.