Keyword & Query Matching

Exact Match

By the AdFlint research team · Last reviewed July 2026

Matches only queries with the same meaning as the keyword, including close variants such as plurals, misspellings, and reordered function words.

Exact match still triggers on close variants, so it is a same-intent filter rather than a literal string filter. You use it on high-value terms where you want tight control over cost per conversion and clean per-keyword data. The frequent error is treating it as airtight and skipping the search terms report, which is exactly where you find the same-meaning queries you never wanted to buy.

Key takeaways

  • Exact match is a same-intent filter, not a literal string filter - plurals, misspellings, and reordered function words still qualify as matches.
  • Reserve exact match for high-value, already-validated terms where you want tight cost control and clean per-keyword reporting, not as your only match type.
  • Never skip the search terms report on exact match keywords - close variant matching means unwanted queries can still slip through.
  • Use exact match for controlled comparisons (isolating a bid change or ad test) since it introduces the least query-level variance of the three match types.

In practice.

Exact match filters on same meaning, not identical text, which is the single most common misunderstanding advertisers carry into this setting. A keyword set to exact match will still trigger on plurals, common misspellings, abbreviations, and queries with function words reordered or added ("the," "for," "a"), as long as Google judges the underlying intent to be the same. "Running shoes" as an exact-match keyword can match "running shoe," "running shoos," and "shoes for running" - none of which are the literal string "running shoes" - because all three express the same search intent.

It interacts with bid strategy the least of the three standard match types, because the variance it introduces is smallest. Exact match works cleanly under Manual CPC, which is part of why it is the default choice for diagnostic periods when you want to isolate one variable (a new ad, a new landing page, a bid change) without match-type-driven query variance muddying the read. It also works fine under Smart Bidding once volume supports it, but the lift from automation is smaller here than it is for broad match, since there is less query-level variance for the algorithm to price differently.

It matters most on high-value terms - branded searches, your core service or product name, terms you have already validated convert - where you want tight control over cost per conversion and clean per-keyword data for reporting. It matters least as a way to control your whole account's traffic, because exact match alone caps your reach at whatever close variants Google recognizes; it will not surface new query ideas or adjacent demand the way phrase or broad match will. An account running exact match exclusively tends to plateau in volume even when there is unmet budget and unmet demand sitting one match type away, and that plateau can be mistaken for a saturated market when it is really just a targeting ceiling.

The frequent error is treating exact match as airtight and skipping the search terms report entirely, on the logic that "it's exact match, so I already know what's triggering it." That logic breaks down precisely because of close variant matching - the report is exactly where you find the same-meaning queries you never explicitly agreed to buy, including close variants that carry unwanted intent (a misspelling that happens to also mean something unrelated, or a plural that shifts meaning in a way singular did not). The second common error is over-relying on exact match early in an account's life, before there is enough data to know which close variants and paraphrases are actually converting, which caps learning before it starts.

In reporting, exact match gives you the cleanest per-keyword read of the three standard match types, which is its main reporting advantage: cost per conversion, CTR, and conversion rate at the keyword level are trustworthy because the query variance underneath each keyword is narrow. Use that cleanliness deliberately - exact match is the match type to reach for when you need a controlled comparison between two keywords or two ad variations, not just a scaling tool.

Worked example

Isolating a bid test with exact match

Suppose an e-commerce brand wants to test whether raising the max CPC on their top converting keyword from $2.50 to $3.25 improves impression share without tanking ROAS. They run it as an exact-match keyword under Manual CPC specifically so the query mix stays stable across the two-week test window - no new broad match expansion could show up mid-test and confound the comparison.

Week one at $2.50: 400 clicks, $1,000 spent, 20 conversions, $50 cost per conversion. Week two at $3.25: 380 clicks, $1,235 spent, 26 conversions, $47.50 cost per conversion. Because exact match kept the underlying query set essentially unchanged between the two weeks, the brand can attribute the improvement to the bid change itself rather than to a shift in which queries were being bought - a comparison that would be far muddier under broad match, where the query mix could have shifted for reasons unrelated to the bid.

Exact Match compared with

The settings this gets confused with, and how to tell them apart.

Common questions.

Can exact match keywords still get impressions from misspelled searches?

Yes. Google's close variant matching explicitly includes common misspellings, so an exact-match keyword for "mattress" can still trigger on a search for "matress" without you adding the misspelling separately. This is standard behavior, not a targeting leak.

Why does my exact match keyword have more impressions than I expected for such a specific term?

Close variant matching is the usual explanation - exact match covers plurals, misspellings, abbreviations, and reordered function words in addition to the literal phrase, so the effective set of matching queries is wider than the keyword text alone suggests. Pull the search terms report to see the actual variant breakdown.

Should every high-value keyword eventually move to exact match?

Not automatically - the usual pattern is to start terms on broad or phrase match to discover which specific queries convert, then promote the proven winners into their own exact-match keywords for tighter bid and budget control. Moving straight to exact match without that discovery step means you are guessing at exact query wording instead of letting the data tell you.

You should not need to know this to advertise.

AdFlint handles the settings for you, inside the Google and Meta accounts you already own.

Try AdFlint free