Google Ads Match Types in 2025: What You Learned in 2019 Is Now Wrong

Google Ads match types control which search queries trigger your ads. Broad match can show your ad for loosely related searches. Phrase match triggers for queries that include the meaning of your keyword. Exact match only triggers when the search is closely related to your exact keyword. There used to be four match types; today, there are effectively three — and all three behave differently than they did five years ago.

If you learned match types before 2021, you probably learned a version of this:

Broad match = dangerous. Broad match modifier = the safe broad. Phrase match = moderate control. Exact match = precision.

That mental model is now mostly wrong. The match type you were trained to use as a "safe broad" — broad match modifier (BMM) — was retired in July 2021. The others have had their matching behavior expanded significantly. And Google's automated bidding has become a variable that interacts with match types in ways the old guides never accounted for.

This isn't a Google-is-evil rant. Some of these changes are genuinely good for advertisers. But you can't manage what you don't understand, so let's build an accurate picture.

---

The 2021 Earthquake Nobody Noticed

On February 4, 2021, Google's Ads Liaison Ginny Marvin posted an announcement that most practitioners either missed or underestimated: Broad Match Modifier (BMM) was being phased out, with phrase match absorbing most of its function.

BMM keywords used a + prefix on each required word — +keyword +match +type — to force those words (in any order) to appear in the triggering query. It was the workhorse match type for many agencies because it gave coverage without the risk of pure broad match.

Google's rationale was that modern phrase match could handle the same use cases while adding intelligence about word order and meaning. That was true in theory. In practice, the migration had two significant side effects:

1. Phrase match got broader. The old phrase match required the keyword phrase to appear in order within the query. The new phrase match — incorporating BMM behavior — triggers for queries that include the "meaning" of your phrase, which Google's language models interpret much more liberally. Queries that phrase match would never have triggered in 2020 now regularly match.

2. Advertisers who relied on BMM saw traffic spikes. When BMM was automatically converted to phrase match in mid-2021, campaigns that had been carefully managed under BMM suddenly started triggering for broader queries. Many accounts saw impression volumes jump 20-40% with corresponding drops in conversion rate. For advertisers not watching Search Terms reports closely, this was invisible until budget overruns showed up.

If you inherited a Google Ads account and the campaign notes say "these are BMM keywords," the + symbols were likely stripped during migration, and the keywords are now running as phrase match with expanded behavior. Check your search terms report for the last 90 days against your keyword list — the delta between what you're bidding on and what's actually triggering your ads is probably larger than you think.

---

The Three Match Types, Explained With Fishing

The right mental model for match types is fishing methods. Different tools for different conditions, different yields, different costs.

Broad match is a trawl net.

You drag it through the ocean and pull up everything: the fish you wanted, fish you didn't want, an old boot, some seaweed. In volume, nothing beats it. In precision, nothing's worse. With broad match, Google will show your ad for any search it deems semantically related to your keyword — including synonyms, tangentially related queries, and sometimes searches that make you genuinely wonder what Google was thinking.

Broad match is most powerful when paired with Smart Bidding (Target CPA or Target ROAS). Without automated bidding, pure broad match is usually wasteful. With it, Google's bidding algorithm acts as a quality filter — it only bids aggressively when it predicts a conversion is likely, which means low-quality broad match traffic gets depressed bids automatically. The trawl net still catches everything, but it only hauls in the expensive catch; it lets the rest go.

Phrase match is a gill net.

A gill net is set at a specific depth and mesh size to catch a target species. You still catch some bycatch, but you've narrowed the pool significantly. Phrase match requires the searcher's query to contain the meaning of your keyword phrase. Word order matters but isn't absolute — Google's language understanding will match queries that are meaningfully the same even with slight reordering.

Phrase match is the current workhorse for most mid-to-large accounts. It gives enough coverage to generate volume while maintaining meaningful targeting. For most commercial keywords, phrase match + Smart Bidding is the starting configuration.

Exact match is spearfishing.

You know exactly what you're going after. You won't catch anything else. Exact match triggers only when the search query is essentially the same as your keyword — Google allows for close variants (misspellings, reordering, functional words) but the intent must match tightly.

Exact match gives you maximum control over which queries trigger your ads, which is important for high-value keywords where you're paying $15-50+ per click and can't afford mismatched intent. It also gives you the cleanest data for bid optimization because you know precisely what triggered the ad.

The tradeoff: exact match limits your reach. A campaign running only exact match keywords will miss real buyers who use slightly different phrasing. Most accounts benefit from a layered approach — exact match for high-value terms, phrase match for discovery.

---

The Match Type Behavior Table (2025 Reality)

"Close variants" deserve a note. Google introduced close variant matching for exact and phrase match keywords in 2017-2018. Before then, exact match meant exactly that keyword. Now, close variants include:

• Misspellings (campain tracking = campaign tracking)
• Singular/plural forms (utm builder = utm builders)
• Stemmings (running shoes = run shoes)
• Abbreviations
• Accents

This was controversial. Many advertisers building negative keyword lists to sculpt traffic were frustrated that close variants sometimes bypassed exact match negatives. This is still a known limitation in 2025: a negative exact match keyword won't necessarily block all its close variants.

---

The Smart Bidding Variable Nobody Talks About

Match types alone don't determine your traffic. Match types + bidding strategy together determine your traffic.

This is the biggest shift in how match types work compared to five years ago, and it's underappreciated.

With manual CPC bidding, match type is the primary filter. You set bids, Google matches queries according to the match type rules, and you pay what you bid. Simple.

With Smart Bidding (Target CPA, Target ROAS, Maximize Conversions), Google's algorithm adjusts your bid in real time based on predicted conversion probability. This means:

• A broad match keyword with tCPA bidding might generate better quality traffic than a phrase match keyword with manual CPC — because the algorithm is filtering by conversion intent, not just query similarity
• Exact match keywords with tCPA might actually reach fewer high-value users than you expect, because the algorithm is constrained by match type precision even when it could otherwise identify a convertible user through a broader query

The implication: the "match type is your quality filter" mental model only holds for manual bidding. With Smart Bidding, the algorithm itself is a quality filter, and match type becomes more of a scope control — how wide a net do you want to cast for the algorithm to find conversions within.

A 2022 analysis by Merkle (one of the largest paid search agencies in the US) found that for accounts spending over $50k/month on branded terms, broad match + Smart Bidding drove 10-15% more conversions at similar CPA compared to exact match + Smart Bidding. The counterintuitive result: by giving the algorithm more room to operate, it found more buyers.

This doesn't mean broad match is always right. In categories with high irrelevant-query risk (e.g., legal, medical, financial terms that share words with non-commercial queries), phrase match still provides meaningful protection even with Smart Bidding.

---

A Real Campaign Failure Worth Learning From

In early 2020 (pre-BMM sunset), a personal injury law firm in the US was running phrase match keywords for "car accident attorney" and related terms. Roughly 65% of their budget was going to these keywords.

A new account manager switched the keywords from phrase to broad match to "increase reach." No Smart Bidding — they were on manual CPC.

Within two weeks, their search terms report was full of queries like "car accident settlement amount," "how to negotiate with insurance," and "what to do after car accident" — all informational intent, zero conversion intent. Impressions tripled. Clicks doubled. Conversions dropped 40%. CPA nearly tripled.

The issue wasn't broad match itself. Broad match without Smart Bidding is like a trawl net without a sorting system — you haul everything up and manually sort on deck, which is expensive and slow.

The fix: switch to tCPA bidding + broad match, set conservative tCPA targets, and give the algorithm 2-3 weeks to learn. The same keywords on Smart Bidding generated 15% more leads at a lower CPA than the original phrase match / manual CPC setup.

The lesson: match type strategy is incomplete without a bidding strategy. They're a pair, not independent levers.

---

When to Use Each Match Type in 2025

Use broad match when:

• You're running Smart Bidding (tCPA, tROAS, Max Conv.) and have enough conversion data (30+ conversions/month) for the algorithm to learn
• You want to discover new query variations you haven't thought to bid on
• You're in a category with diverse synonym usage where exact/phrase misses real buyers
• You're willing to monitor search terms reports weekly for irrelevant traffic

Use phrase match when:

• You want coverage beyond exact match without the unpredictability of broad match
• You're on manual CPC or have limited conversion data for Smart Bidding
• Your keywords contain high-risk words that broad match could take in dangerous directions (e.g., legal, medical, finance)
• You're in the early stages of a campaign before you have enough conversion data to trust broad + Smart Bidding

Use exact match when:

• The keyword is extremely high-value (>$20 CPC) and you need to control exactly what triggers it
• You're building a brand keyword campaign and want precise control over brand vs. non-brand query matching
• You're testing ad copy and landing page combinations and need clean, consistent query data
• You want to isolate specific high-intent terms from discovery keywords in the same campaign

The practical default for most accounts starting fresh: phrase match + Smart Bidding, with exact match duplicates for your 10-20 most important keywords. After 60-90 days of data, evaluate whether broad match + Smart Bidding outperforms phrase on core terms.

---

Negative Keywords Still Matter More Than Most People Think

All the match type sophistication in the world doesn't help if you're not maintaining negative keyword lists.

In a broad or phrase match campaign, you will trigger for irrelevant queries. The question is how quickly you identify and negate them. A weekly search terms report review where you add negatives for:

• Informational queries (how to, what is, define, meaning of)
• Competitor brand names (unless you're intentionally targeting them)
• Unrelated industries that share terms with your category
• Job-seeker queries (jobs, careers, salary, interview)
• Free/cheap modifiers (if you're not a freemium product)

...will consistently improve campaign efficiency by 15-25% over accounts that set-it-and-forget-it.

Use the keyword match type tool on MarketerTools to check how a given keyword would behave under each match type before adding it to a campaign.

---

The One Framework That Survives Every Google Update

Google will keep changing match type behavior. What hasn't changed — and likely won't — is the underlying logic:

More control = less reach. Less control = more algorithmic dependence.

The art of match type strategy is deciding, for each campaign and keyword, how much you trust Google's algorithm to find the right buyers versus how much you need manual control. That answer depends on your conversion volume, your bidding strategy, your category, and your tolerance for variance in spend.

Accounts with mature Smart Bidding setups and high conversion volume should lean toward broad match — they're giving the algorithm the data it needs to outperform manual control. Accounts with limited data, high CPC, or niche categories should lean exact and phrase — they can't afford the algorithm's learning curve.

Every account is somewhere on that spectrum. Your match type strategy should reflect where you actually are, not where a 2019 blog post said you should be.

---

FAQ

What are Google Ads match types? Match types are settings on keywords in Google Ads that determine how closely a user's search query must match your keyword for your ad to be eligible to appear. There are three match types: broad match (widest reach), phrase match (moderate control), and exact match (highest control). A fourth type — broad match modifier — was retired by Google in July 2021 and its behavior was folded into phrase match.

Which match type should I use in 2025? For most accounts, phrase match combined with Smart Bidding (Target CPA or Target ROAS) is the safest starting point. For high-value terms or mature accounts with sufficient conversion data, broad match + Smart Bidding often outperforms phrase match. Exact match is best for brand keywords, very high-CPC terms, or situations where you need precise control over which queries trigger your ads.

What happened to broad match modifier? Google retired broad match modifier (BMM) in July 2021. BMM keywords used a + prefix to require specific words to appear in the triggering query. Google rolled its behavior into phrase match, which now handles more of the use cases BMM was used for. Existing BMM keywords were automatically converted to phrase match. The practical effect was that phrase match traffic increased and became slightly less predictable for advertisers who relied on BMM's tighter matching.

Does match type matter with Smart Bidding? Yes, but differently than with manual bidding. With Smart Bidding, match type controls scope (how wide a net to cast) while the algorithm controls quality (which impressions within that scope to bid on). Broad match with Smart Bidding can outperform exact match with Smart Bidding in high-volume accounts because the algorithm has more opportunities to find converting users. With manual CPC, match type is the primary quality control mechanism.

What are close variants in Google Ads? Close variants are searches that Google considers meaningfully similar to your keyword, even if the exact words don't match. For exact and phrase match keywords, close variants include misspellings, singular/plural forms, stemming variations, abbreviations, and accents. Google introduced close variants for exact match in 2018 and has expanded what counts as a "close variant" several times since. This means exact match no longer means exactly that keyword — it means that keyword and queries Google determines have the same meaning and intent.

How do I find out what queries are triggering my ads? In Google Ads, go to Keywords → Search Terms report. This shows you the actual user queries that triggered your ads. Filter by match type, impression volume, or conversion data to identify both valuable new keyword opportunities and irrelevant queries to add as negatives. Review this report at minimum weekly for active campaigns.