Lecture 6: Match Types Explained: Broad, Phrase, Exact, and Negative Keywords

Short answer: Match types tell Google and Microsoft how loosely or tightly a keyword should be interpreted before your ad is allowed to enter the auction. Broad match casts the widest net and now leans heavily on Smart Bidding and audience signals rather than the literal words you typed. Phrase match requires the core meaning and word order of your phrase to be preserved, though wrapper words are allowed on either side. Exact match once meant "only this exact string" but today it matches close variants, reworded phrases, and even implied intent that the algorithm judges to mean the same thing. Negative keywords are the fourth, often-overlooked lever: they tell the auction which searches should never trigger your ad, and they are the single fastest way to stop budget leaking onto irrelevant clicks. Mastering all four together is what separates accounts that scale profitably from accounts that just spend.

What You'll Learn in This Lecture

• How broad match actually works in 2026, and why it's no longer "match anything remotely related"
• How phrase match preserves meaning and word order while allowing flexible wrapper text
• Why exact match no longer means literally exact, and what "close variants" includes
• The historical timeline of how match type behavior has loosened since 2014
• How to choose match types based on campaign goal, funnel stage, and bidding strategy
• Why negative keywords protect budget and improve Quality Score and ROAS
• A repeatable process for building a negative keyword list from search terms and team input
• Common negative exclusions by industry (legal, ecommerce, SaaS, local services, healthcare)
• How match types apply differently to negative keywords than to regular keywords
• How to build and apply shared negative keyword lists across an entire account
• A weekly and monthly cadence for auditing search terms to refine match types and negatives
• The most common match type mistakes that quietly waste ad spend
• A practical framework for testing match type changes without breaking a stable campaign

Broad Match Explained

Broad match is the default match type and the loosest of the three. In its original form (pre-2010s), broad match simply meant your ad could show for searches containing your keyword's words in any order, including synonyms and related terms determined by simple linguistic rules. In 2026, broad match has evolved into something closer to an intent-matching engine than a keyword-matching engine. When you enter a broad match keyword like running shoes for beginners, Google no longer just looks for those words — it looks at the searcher's full query, recent search history, device, location, the landing page content, other keywords in your ad group, and — critically — the bid and budget signals coming from your Smart Bidding strategy (Target CPA, Target ROAS, or Maximize Conversions). The result is that broad match can trigger your ad for searches sharing almost no words with your original keyword, provided the algorithm believes the underlying intent and conversion likelihood are strong enough.

This is why Google now strongly recommends broad match almost exclusively be paired with automated bidding that optimizes to a conversion or value goal, rather than manual CPC. Without conversion data to learn from, broad match with manual bidding is the riskiest combination in the platform — you are handing matching decisions to an algorithm with no signal about what a "good" click looks like yet. With Target CPA or Target ROAS active and enough historical conversion volume (generally 30+ conversions per month as a practical minimum), broad match becomes a genuine discovery tool: it surfaces converting query variations you would never have thought to add manually, including long-tail and conversational phrasing.

The tradeoff is control. Broad match in 2026 requires an aggressive, well-maintained negative keyword list and close monitoring of the search terms report, because the expanded matching logic will occasionally pull in searches that are topically adjacent but commercially wrong — for example, matching a "buy" query to a "how to" informational search because the algorithm judged the audience signals similar.

Phrase Match Explained

Phrase match sits between broad and exact. A phrase match keyword, written with quotation marks such as "running shoes for beginners", requires that the meaning of the phrase be preserved and that words appear in a similar order, but it allows additional words before, after, or occasionally in the middle of the phrase, as long as those additions don't change the core meaning. Since the 2021 phrase match update (which folded the old "broad match modifier" into phrase match), phrase match has also picked up meaning-based flexibility: it can match reworded versions of your phrase, synonyms for individual words, and different word forms (plurals, verb tenses), provided intent stays the same.

What phrase match will not do is match a search where the core meaning changes. A phrase match on "running shoes for beginners" should not match "running shoes for marathon runners," since that changes the target audience implied, even though most words overlap. This makes phrase match the workhorse for most mid-funnel campaigns: meaningfully more reach than exact match, without the wide-open exposure of broad match.

Exact Match Explained

Exact match, written in brackets such as [running shoes for beginners], was originally the strictest match type — your ad would only show when the query matched your keyword word-for-word, with no extra words, reordering, or synonyms. That literal definition has not been true since Google's "close variants" expansion began in 2014 and deepened through updates in 2017, 2018, and the 2021-2022 close-variant-for-exact-match update. Today, exact match includes: word order changes that preserve meaning, plurals and singulars, misspellings and typos, abbreviations, stemmings (run/running/runs), implied words, and paraphrases that Google's language models judge to carry the same intent.

In practical terms, [running shoes for beginners] can now legitimately trigger for "beginner running shoe," "running shoes new runner," or "best starter running shoes" — none of which are the literal string you entered. Google frames this as serving "the same intent," but from an advertiser's perspective, exact match is best understood in 2026 as "tightest available control," not "literal string match." It remains the match type with the smallest reach and highest average relevance, which is why it's still preferred for high-intent, high-value, or brand-protection keywords needing maximum predictability.

Example: Take the keyword waterproof hiking boots. As broad match (waterproof hiking boots, no punctuation), it can trigger for "best boots for rainy trails," "shoes that keep feet dry hiking," or even "gifts for a hiker who complains about wet feet," because Smart Bidding judges these searchers convert similarly. As phrase match ("waterproof hiking boots"), it can trigger for "waterproof hiking boots for wide feet" or "buy waterproof hiking boots online," but should not trigger for "waterproof boots for construction work," since that changes the intended use case. As exact match ([waterproof hiking boots]), it can trigger for "waterproof hiking boot," "hiking boots that are waterproof," or a minor misspelling like "waterproof hikeing boots," but it will not trigger for unrelated reworded searches about boots for a different activity.

How Match Type Behavior Has Changed Over the Years

Understanding the trajectory matters because advertisers who learned match types a decade ago often carry outdated assumptions into 2026 accounts. The rough timeline: before 2014, all three match types behaved close to their literal definitions, and "broad match modifier" (the plus-sign syntax) was introduced as a middle ground requiring specific words to be present. Between 2014 and 2018, exact match began quietly absorbing close variants — first plurals and misspellings, later reordering and paraphrasing. In 2019, Google extended close variants to phrase and broad match modifier too. In July 2021, Google retired broad match modifier entirely and folded its behavior into phrase match, consolidating five match types down to three plus negatives. Through 2022-2023, phrase and exact match both gained deeper semantic understanding as matching systems moved from rule-based synonym lists to transformer-based language models, so matching decisions increasingly reflect "what does this query mean" rather than "which words are present."

From 2024 into 2026, the biggest shift has been the tightening coupling between broad match and automated bidding. Google now actively nudges advertisers toward broad match plus Smart Bidding as the default recommended setup for new campaigns, and positions "keywordless" or audience-signal-driven matching (as seen in Performance Max) as a natural extension of broad match logic. Phrase and exact match have also grown more convergent — both incorporate close variants and paraphrase matching, with exact match simply applying a narrower similarity threshold rather than a fundamentally different mechanism. The practical upshot: no match type today is "literal," and every match type increasingly depends on the quality of your conversion data and negative keyword hygiene rather than the exact syntax you choose.

Choosing Match Types by Campaign Goal and Bidding Strategy

Because match types no longer have fixed, predictable reach, the right choice depends heavily on what the campaign is optimizing for and what bidding strategy is attached to it. A few concrete scenarios:

Brand campaigns: Use exact match (and tightly scoped phrase match) on your own brand terms to guarantee the ad shows for genuine brand searches without leaking spend into broader discovery. Manual CPC or Target Impression Share often pairs well here since the goal is coverage and defense.

Lower-funnel, high-intent commercial campaigns (e.g., "buy," "price," "near me," model numbers): Favor exact and phrase match with Target CPA or Target ROAS. These searches already carry strong purchase intent, so precision beats expansion — broad match here risks pulling in cheaper but lower-intent traffic that dilutes ROAS.

Mid-funnel consideration campaigns: Phrase match is usually the sweet spot — it captures intent-adjacent variations without opening the floodgates. Pair with Maximize Conversions once you have enough volume for the algorithm to learn from.

Discovery and scaling campaigns with mature conversion tracking and healthy historical data: broad match with Target ROAS or Maximize Conversion Value is appropriate, because you are explicitly trying to find new converting query space and you have the bidding guardrails and negative list needed to control it.

New accounts or thin conversion data: start narrower (phrase and exact) while you build the 20-30+ monthly conversions most Smart Bidding strategies need, then graduate portions of the account to broad match once you trust the signal.

A useful mental model: match type controls how wide the door is, and bidding strategy controls how carefully the algorithm decides who is allowed to walk through it. The two decisions should always be made together, never in isolation.

Negative Keywords: Why They Matter

If match types decide what your ads can show for, negative keywords decide what they absolutely must not show for. This is not a minor housekeeping task — negative keyword management is one of the highest-leverage activities in most accounts, because every irrelevant click is money removed from your budget that could have gone to a relevant one. A single unmanaged broad or phrase match keyword can silently accumulate hundreds of dollars in spend on searches that were never going to convert, and because that spend is spread thin across many small irrelevant queries, it often doesn't show up as an obvious problem in top-line reporting — it just quietly drags down your average CPA and ROAS.

Negative keywords also sharpen the relevance signal Google uses to calculate Quality Score. When your account systematically excludes irrelevant queries, the remaining traffic that triggers your ads is more topically aligned with your ad copy and landing page, which tends to improve expected click-through rate and relevance scores — and in turn can lower your effective cost per click across the board. Negative keywords are not just defense; they are an active lever for making your paid traffic and your Quality Scores better at the same time.

Building a Negative Keyword List

Negative keyword lists should be built from several complementary sources rather than a single pass of guesswork:

1. Search term reports. Your primary, ongoing source. Every search that triggered an ad is visible in the Search Terms report, broken out by campaign and ad group. Sort by cost or clicks with zero conversions to find the highest-waste offenders first.

2. Pre-launch brainstorming by category. Before a campaign goes live, list obvious irrelevant categories for your business. Common cross-industry exclusions: "free," "diy," "jobs," "salary," "course," "definition," "meaning," "wikipedia," "reddit," and "review" (unless review content is genuinely part of your funnel).

3. Industry-specific exclusions. Legal services: "law school," "legal aid," "pro bono," "how to become a lawyer." Ecommerce: "used," "refurbished," "return policy," "class action." SaaS: "open source," "free alternative," "api documentation," "careers." Local services (plumbers, electricians, contractors): "diy," "how to fix," "certification," "training course." Healthcare: "symptoms," "home remedy," and any terms that could trigger medical-advice searches rather than appointment-booking intent, plus compliance-driven exclusions specific to your vertical.

4. Competitor and adjacent-brand terms you do not want to bid on (or want isolated into their own dedicated campaign rather than blended into generic campaigns).

5. Search Term Insights. Google's automated category grouping can surface irrelevant themes faster than manually scanning thousands of individual queries.

6. Sales and support team input. Ask the people fielding inbound leads what unqualified inquiries look like — this often reveals negative categories a search term report alone would never show.

Match Types for Negative Keywords

Negative keywords use the same three match type syntaxes as regular keywords, but the logic they enforce is inverted, and the practical implications are different enough that they deserve separate treatment.

Negative broad match (plain text, e.g., free) blocks any search containing all the words in your negative keyword, in any order — but unlike positive broad match, it does not expand to synonyms or close variants; it is a literal "contains all these words" block. This makes negative broad the most powerful and most dangerous negative match type: a careless single-word negative broad like free will block every search containing that word, including ones you might have wanted, such as "toll-free number for support."

Negative phrase match (e.g., "free shipping") blocks any search containing that exact sequence of words in that order, with other words allowed around it. This is usually the safest default for multi-word exclusions.

Negative exact match (e.g., [free]) blocks only the search query matching that exact term and nothing else — no close variants, no extra words. This is the most surgical option, useful for blocking one specific query without affecting anything adjacent.

A critical technical point: negative keywords, unlike positive keywords, do not benefit from close-variant matching at any match type. Your negative list must explicitly include misspellings, plurals, and reworded versions if you want to block those too. This asymmetry catches many advertisers off guard and is a common source of "why is this obviously irrelevant search still triggering my ad" support tickets.

Shared Negative Keyword Lists Across Campaigns

Rather than rebuilding the same exclusions campaign by campaign, Google Ads and Microsoft Ads both support shared negative keyword lists at the account level, applicable to any number of campaigns — essential for consistency and maintenance time as an account grows. A typical structure includes at minimum: a universal "always exclude" list ("free," "jobs," "salary," "wikipedia," "reddit," and other terms almost never relevant regardless of campaign); an industry-specific list tailored to your business category; a competitor list if you deliberately want to exclude (or isolate) competitor brand terms from generic campaigns; and a campaign-type list distinguishing exclusions appropriate for Search but not Shopping or Performance Max, since query behavior differs across formats.

Shared lists should be reviewed on their own cadence, separate from campaign-level negatives, because a single addition can affect dozens of campaigns simultaneously — a powerful efficiency, but also a risk if a badly chosen negative blocks a valuable traffic segment account-wide. Always check recent search term volume for a term across the account before adding it to a shared list, not just in the campaign where you first noticed it.

Auditing Search Terms Regularly to Refine Match Types and Negatives

Match types and negative lists are not a "set once" configuration — they require an ongoing audit rhythm, because query behavior shifts as your ads gain data, as broad match's Smart Bidding signals mature, and as seasonal or trending searches introduce new irrelevant traffic. A practical cadence:

Weekly: scan the search terms report for any new query with meaningful spend (even $10-20 in a smaller account) and zero conversions. Add clear irrelevant matches to negatives immediately rather than letting them accumulate.

Bi-weekly to monthly: review search terms that are converting well but aren't yet their own keywords — candidates to "harvest" into dedicated exact or phrase match keywords so you can control bids and ad copy directly, rather than leaving them entirely to broad match discovery.

Monthly: reassess match type mix at the campaign level. If a broad match campaign's search terms drift toward irrelevant themes despite an actively maintained negative list, that signals a need to shift some keywords back to phrase or exact, or to tighten the bidding strategy's conversion goal (e.g., switching from Maximize Conversions to Target CPA once enough data exists).

Quarterly: audit the shared negative lists themselves for outdated entries, overly broad negative-broad terms silently blocking valuable traffic, and gaps revealed by new product lines added since the last review.

This rhythm turns match type management from a one-time setup decision into a continuous feedback loop — exactly how the algorithm expects to be used, since Smart Bidding and broad match both improve as they receive cleaner conversion signals over time.

Common Match Type Mistakes That Waste Budget

A few recurring mistakes account for the majority of match-type-related waste seen across accounts:

Running broad match with Manual CPC and no negative list. The single most expensive combination possible — no automated bidding to constrain who the ad shows to, and no negatives to catch the overflow.

Assuming exact match means "no expansion." Advertisers who stop watching the search terms report are often surprised to find close variants and reworded queries still triggering broad, sometimes off-target matches.

Using single-word negative broad match carelessly. A term like free or jobs can block far more traffic than intended, including relevant searches that happen to share that word. Prefer negative phrase or exact for anything not unambiguously universal.

Forgetting that negatives don't get close-variant matching. Teams add one version of a negative and assume misspellings or plurals are automatically covered — they are not, and this gap quietly lets irrelevant spend back in.

Letting match types stagnate as an account matures. A campaign launched cautiously with exact match should often graduate budget to phrase or broad once conversion data accumulates; the reverse mistake — never tightening a broad match campaign that keeps drifting off-topic — is just as common.

Duplicating the same keyword across match types without bid or budget alignment, causing internal competition between your own ad groups rather than genuine incremental reach.

Not applying shared negative lists consistently across new campaigns, so every launch reintroduces waste already solved elsewhere in the account.

Mastering match types and negative keywords is what allows every other lecture in this course — bidding strategy, ad copy testing, landing page alignment, budget allocation — to actually work, because none of those levers matter if the wrong searches trigger your ads in the first place. In the next lecture, we shift from controlling which searches you match to understanding what those searches actually mean, through search intent classification and SERP analysis.

Related Lessons Across SEO, AEO, GEO, SEM, and PPC

Use these connected lessons to move through organic search, answer engines, generative AI visibility, paid search, and PPC without losing the bigger strategy.

• Lecture 1: What Is SEM? How Search Engine Marketing Works in 2026 (SEM) - return to the course foundation when you need the big picture.
• Lecture 1: What Is PPC? How Pay-Per-Click Advertising Works (PPC) - separate SEM strategy from PPC execution.
• Lecture 38: PPC Strategy Roadmap: Bringing Search, Social, Retail, and AI Together (PPC) - connect SEM with the full PPC channel roadmap.
• Lecture 4: Keyword Research Fundamentals (SEO) - connect organic keyword research with the same demand signals.
• Lecture - 7: Prompt Research: Finding What People Ask AI Tools About Your Topic (GEO) - turn search queries into AI prompt research.