Google Ads gives you three keyword match types, broad, phrase and exact, and every one of them now matches on meaning rather than exact wording. The quick rule: use exact match when you need tight control on low volume, phrase match while you’re testing new territory, and broad match once you’ve got reliable conversion data feeding Smart Bidding. Google Ads Help confirms this meaning based shift, and Google’s own Smart Bidding guidance is built around it.
TL;DR:
- Broad, phrase, and exact match now all target search intent, with Google matching on meaning rather than literal wording, including close variants and synonyms.
- Using negative keywords effectively is key, with negative exact for specific queries, negative phrase for categories, and negative broad as a blunt tool to block unwanted searches.
- Pair broad match with automated bidding strategies like Target CPA or Target ROAS when you have more than 30 conversions monthly to maximize learning and efficiency.
- Avoid mixing match types within the same ad group to ensure clear performance insights and better control over keyword relevance.
- Regularly review search terms and update negative keyword lists to prevent budget waste and maintain campaign accuracy.
Table of Contents
- What are Google Ads match types, really?
- How Google actually decides which keyword triggers your ad
- Negative keywords: your real control lever
- Match types and Smart Bidding: picking the right pairing
- Rolling out match-type changes without breaking your account
- What we see going wrong in account audits
- How match types affect Quality Score and Ad Rank
- Search Network versus Display: match types don’t behave the same way
- How to test and optimise match-type performance properly
- The misconceptions that cost advertisers money
- Our take: match types are a dial, not a switch
- Get your match-type strategy audited before you spend another dollar on guesswork
- Sources
- FAQ
What are Google Ads match types, really?
Forget what you learned about match types five years ago. Broad, phrase and exact match all target the intent behind a search now, not the literal string of words you typed into the keyword field. Google states outright that ads can show on searches that share “the same meaning or same intent as the keyword”, even for exact match, which used to be the strictest, most literal option in the toolkit.
Here’s how the three behave and how you write them in your account:
- Broad match (written with no punctuation, e.g.
running shoes): the widest net. Google can show your ad for searches it judges related to your keyword’s meaning, including synonyms, related categories, and searches with no words in common at all. - Phrase match (written in quotation marks, e.g.
"running shoes"): a middle ground. Google keeps the core meaning of your phrase intact but allows extra words before, after, or occasionally inside it. - Exact match (written in brackets, e.g.
[running shoes]): the tightest steering. Google still matches close variants and same-intent searches, but the query has to align closely with the keyword’s specific meaning.
A worked example makes this concrete. For the keyword running shoes, broad match might trigger on “best trainers for marathon training” or even “athletic footwear sale”, because Google has decided the underlying intent overlaps. Phrase match on "running shoes" would likely catch “cheap running shoes Melbourne” or “running shoes for flat feet”, where the core phrase survives. Exact match on [running shoes] narrows this further, generally to the search itself and its closest variants, like “running shoe” or “running shoes” (a common misspelling).
One historical note worth clearing up: broad match modifier, the old plus-sign syntax, was retired in 2021. If you’re still writing +running +shoes anywhere, that syntax no longer does anything special. Phrase match absorbed most of BMM’s old behaviour, so there’s no reason to chase it in 2026.
How Google actually decides which keyword triggers your ad
Every search you run through Google Ads goes through a matching pipeline before an ad ever appears, and understanding the steps stops you relying on the old “literal string” mental model that no longer applies.
- Interpretation. Google reinterprets the raw search first, correcting spelling, expanding abbreviations, and factoring in the searcher’s recent activity to build what it calls a retrieval query.
- Eligibility. Every keyword that could plausibly match gets checked against budget, geotargeting, device settings, negative keywords, and policy approval. Anything that fails drops out before selection even starts.
- Selection. Among the eligible keywords left standing, Google prioritises an identical match to the search term first. If nothing is identical, it weighs relevance, then uses Ad Rank as the tie-breaker between competing advertisers.
That’s the sequence Google Ads Help lays out in its explanation of the matching system, and it explains a lot of confusing account behaviour. Close variants, plural forms, misspellings, abbreviations, and paraphrased rewordings, are treated as legitimate matches at every match type, not just broad. It also explains why two advertisers with the “same” keyword can get wildly different results: your ad group’s other keywords, your ad copy, your landing page, and your assets (sitelinks, images, callouts) all feed into the relevance signal Google uses during selection. Landing page relevance isn’t a footnote here, it’s part of the mechanism.
Negative keywords: your real control lever
If match types are a reach dial, negative keywords are the brake. Because broad, phrase and exact all now match on meaning, negatives are genuinely the only reliable way to stop Google showing your ad on a specific query you don’t want.
The three negative types work in reverse to their positive counterparts:
- Negative exact blocks only that specific search (and its close variants), useful for surgical removal of one bad query.
- Negative phrase blocks any search containing that phrase, good for cutting off a whole category, like “free” or “jobs”.
- Negative broad blocks searches containing all the words in any order, the bluntest and riskiest tool of the three.
Run your search terms report on a fixed 14-day cadence and add negatives as patterns emerge, don’t wait for a monthly review to catch waste.
Pro Tip: Build a shared negative list for obvious junk (jobs, reviews, free, DIY) and apply it across every campaign in the account. It saves you rebuilding the same exclusions every time you launch something new.
Match types and Smart Bidding: picking the right pairing
Smart Bidding changes the calculus on match types more than most advertisers realise. When Google sets bids automatically at auction time using signals like device, location, time of day and audience, it needs a wide enough stream of queries to learn from, and that’s exactly what broad match supplies.
Google’s own guidance recommends pairing broad match with Target CPA, Target ROAS, or Maximise Conversions, because the algorithm can only optimise as well as the data it’s fed. Feed it a narrow stream of exact match queries and you starve it of the volume it needs to find efficient auctions.
Here’s a practical framework for pairing match type with bidding strategy and conversion volume:
- Under 10 conversions a month: start with exact match. Smart Bidding struggles without enough signal, so manual control matters more than reach.
- 10 to 30 conversions a month: phrase match paired with Maximise Conversions or an early-stage Smart Bidding strategy. You get more query variety without losing the plot.
- 30+ conversions a month: broad match paired with Target CPA or Target ROAS becomes viable, this is where broad match and Smart Bidding together typically perform best.
One combination to actively avoid: broad match with manual CPC. Without an automated bidding layer interpreting each auction, broad match’s wide net just burns budget on loosely related searches nobody is bidding intelligently for.
None of this works if your conversion tracking is unreliable. Smart Bidding trained on noisy or duplicate conversion data tends to optimise for the wrong thing entirely, so audit your conversion actions before you widen match types, not after.
Rolling out match-type changes without breaking your account
Match-type changes are cheap to make and expensive to get wrong at scale. A structured rollout keeps the damage contained if an experiment doesn’t pan out.
The pre-flight checklist:
- Confirm your conversion actions are firing correctly and deduplicated.
- Draft a negative keyword list before you go live, not after the wasted spend shows up.
- Set a calendar reminder for a 14-day search terms review.
- Watch CPA or ROAS daily for the first week, then weekly after that.
Three examples show how this plays out differently depending on the goal:
- Brand protection. A retailer bidding on its own name uses exact match almost exclusively,
[brand name], to keep tight control over who and what triggers on branded searches. There’s no upside to reach here, only downside risk from competitors or resellers. - Expanding a proven winner. An ad group with a phrase-match keyword converting well for months gets tested on broad match, with a strict negative list applied on day one and performance compared over a fortnight before any budget shift.
- Cold prospecting. A new campaign launches on broad match paired with Target CPA and a heavier starting negative list than usual, because prospecting queries are inherently unpredictable and the algorithm needs room to learn safely.
Keep one dominant match type per ad group. Mixing broad, phrase and exact in the same group makes it near impossible to read performance cleanly, because Google’s selection logic will quietly favour whichever keyword it judges most relevant for each query. Structure ad groups by intent and landing page instead, and use Keyword Planner to scope volume before you commit. You change match type directly in the keyword’s row in the Ads and Extensions or Keywords section of the UI, no need to pause and rebuild.
Working through this process is exactly what a Google Ads audit is designed to catch before it costs you real budget.
What we see going wrong in account audits
The same three mistakes show up again and again. Conversion actions are misconfigured or duplicated, which quietly poisons any Smart Bidding strategy layered on top. Ad groups mix match types, so nobody can tell which keyword is actually earning the clicks. And negative keyword lists get built once at launch, then never touched again.
The fixes are unglamorous but effective: clean up conversion tracking first, run controlled broad-match experiments with a negatives list ready on day one, and commit to a fortnightly review cadence rather than an occasional one.
How match types affect Quality Score and Ad Rank
Match type doesn’t directly set your Quality Score, but it shapes the conditions Quality Score is calculated under. Quality Score is built from expected click-through rate, ad relevance, and landing page experience, all measured against the actual search query that triggered your ad.
Broad match exposes your keyword to a wider spread of queries, some closely aligned with your ad copy, some more tangential. If those tangential matches pull down your click-through rate or make your ad feel less relevant to the searcher, your Quality Score for that keyword can drift lower over time. Exact match, by staying closer to a tighter cluster of intent, tends to produce more consistent Quality Score readings because the query-to-ad relationship stays predictable.
Ad Rank matters more directly. As the matching pipeline shows, when multiple advertisers are eligible for the same query, Ad Rank breaks the tie, and Ad Rank itself is a function of bid, Quality Score, and expected impact of ad extensions. A broad match keyword pulling in loosely related traffic that tanks your click-through rate can drag your Ad Rank down for genuinely valuable searches too, since Quality Score is measured per keyword, not per query.
The practical takeaway: if you widen a keyword to broad match, watch its Quality Score column for a few weeks. A steady decline is your signal that the query mix has drifted too far from what your ad and landing page actually deliver, and it’s time to tighten with negatives or pull back to phrase match.
Search Network versus Display: match types don’t behave the same way
Match types as most advertisers understand them, broad, phrase, exact, negative, apply to the Google Search Network, where campaigns target actual typed queries. On the Display Network, the concept works completely differently, and this trips up a lot of advertisers moving between campaign types.
Standard Display campaigns don’t use search keywords to trigger placement in the same way. Instead, targeting relies on audience signals, topics, placements, and contextual relevance, Google’s system matches your ad to a webpage or app based on its content and the user’s inferred interests, not a search string. Keywords can still be used as a contextual targeting signal on Display, helping Google understand what kind of content or audience is relevant to your ad, but they don’t function with the same broad/phrase/exact matching logic that governs Search.
This distinction matters for budget planning. A Search campaign running broad match against high-intent queries is fundamentally a different reach mechanism to a Display campaign layering keyword contextual targeting on top of audience signals. If you’re used to Search-style match-type control and move into Display expecting the same precision, you’ll find the levers simply aren’t there in the same form. Performance Max campaigns blend both networks automatically, which is part of why match-type control feels looser inside PMax, you’re not setting keyword match types directly; Google’s systems are making those calls for you across Search, Display, YouTube and more.
How to test and optimise match-type performance properly
Match-type testing works best as a controlled comparison, not a gut-feel swap. Duplicate the ad group you want to test, change only the match type in the new version, and pause the original rather than deleting it, so you can revert instantly if the test underperforms.
Give any test a minimum of two full weeks before judging it, ideally aligned with your 14-day search terms review, so you’re assessing a genuine sample rather than a few noisy days. Track cost per conversion and conversion rate side by side, not just click volume, broad match will almost always win on clicks and lose the comparison if you stop there.
Segment your search terms report by match type periodically. It shows you exactly which queries each match type is actually pulling in, and it’s often the fastest way to spot when broad match has drifted into territory your negatives list hasn’t caught yet. Pair this with intent-based keyword research upfront, understanding the intent clusters behind your keywords makes it much easier to predict how broad match will behave before you switch it on.
The misconceptions that cost advertisers money
The biggest misconception is treating exact match as literal. It hasn’t been literal for years, exact match still triggers on close variants and same-intent searches, it’s simply narrower than broad or phrase. Advertisers who assume [running shoes] only ever matches “running shoes” are routinely surprised by their own search terms report.
The second misconception runs the other way: assuming broad match is inherently wasteful. Broad match paired with Smart Bidding and a properly maintained negative list is a legitimate growth lever, not a budget leak, the waste comes from pairing it with manual bidding or skipping the negatives review, not from broad match itself.
A third trap is chasing volume before conversion tracking is trustworthy. Widening match types without clean conversion data just means Smart Bidding optimises faster toward the wrong outcome. And a fourth, subtler mistake is mixing match types within one ad group, which muddies your ability to read which keyword variant is actually doing the work.
Our take: match types are a dial, not a switch
Most advice on this topic still treats match types like a one-time decision you set and forget. That’s backwards. The evidence points the other way, match type is a dial you adjust as your conversion data matures, not a fixed setting you pick at campaign launch and leave alone.
The conventional wisdom oversells exact match as the “safe” choice for every advertiser, when in reality it starves Smart Bidding of the signal it needs once you’ve got enough conversions to justify automation. Equally, plenty of guides oversell broad match as the future for everyone, when accounts without clean conversion tracking or a disciplined negatives cadence just bleed spend on it.
What we’d prioritise first, before touching match types at all, is conversion accuracy. Get that right, commit to reviewing search terms every fortnight, and the match-type decision becomes far less fraught. It’s less a strategic bet and more a natural consequence of the data you’re already collecting.
— MYBMC
Get your match-type strategy audited before you spend another dollar on guesswork
Reading about broad, phrase and exact match is one thing, seeing exactly where your own account is leaking spend on the wrong pairing is another. Mybusinessmarketingcoach runs hands-on Google Ads management for Melbourne business owners who want a marketing partner that treats their ad account like their own money, because it is.
Every engagement starts with a proper look under the hood, conversion tracking, ad group structure, negative keyword hygiene, before a single match type gets touched. If you’d rather have someone check the mechanics than second-guess every keyword change yourself, book a Google Ads audit and find out exactly where your account stands.
Sources
For the primary sources behind this guide, see Google Ads Help on keyword matching, Smart Bidding guidance, and the KeywordMatchType developer reference.
- About keyword matching options – Google Ads Help
FAQ
What are the different types of Google Ads keywords and campaigns?
Google Ads runs several campaign types, Search, Display, Shopping, Video, Performance Max, App, Demand Gen and Local, but keyword match types (broad, phrase, exact, plus negatives) apply specifically within Search campaigns. People sometimes conflate campaign types with match types, but they’re separate concepts covering different parts of account structure.
What is a match type in Google Ads?
A match type controls how closely a search query needs to align with your keyword’s meaning before your ad becomes eligible to show. The three positive types are broad, phrase and exact, running from widest reach to tightest control.
What is AI Max in Google Ads?
AI Max is a Search campaign setting that expands keyword matching further using Google’s AI, tapping into your assets and landing pages to find additional relevant searches beyond your existing keywords. It’s built on top of the same meaning-based matching principles that already govern broad, phrase and exact match.
What are negative keyword match types in Google Ads?
Negative keywords come in the same three flavours as positive keywords, negative exact blocks one specific search, negative phrase blocks any search containing that phrase, and negative broad blocks searches containing all the specified words in any order. They’re the main lever for keeping ads off unwanted queries now that all match types work on meaning rather than literal text.
Which match type should I use if I’m just starting out?
Start with exact match if you’re running under 10 conversions a month, since Smart Bidding needs volume to work well and manual control matters more early on. Move to phrase match around 10 to 30 conversions, then consider broad match paired with Target CPA or Target ROAS once you’re consistently above 30 conversions a month with clean tracking.