For most of Google Ads’ history, keyword match types were one of the advertiser’s most important controls.
The principle was straightforward. You selected a keyword, chose how closely a person’s search needed to match it, wrote an advert and set a bid. Exact match provided precision. Phrase match allowed some flexibility. Broad match increased reach.
Campaign management involved anticipating the different ways prospective customers might search and building keyword lists to cover them. Singulars, plurals, misspellings, synonyms and alternative word orders could all be added separately. The advertiser decided which searches were acceptable, while Google mainly followed the instructions.
That relationship has changed.
Google Ads increasingly interprets the meaning, context and commercial intent behind a search rather than relying only on the words entered. Exact match is no longer strictly exact. Phrase match no longer requires the keyword to appear as an unbroken phrase in every case. Broad match can reach relevant searches that do not contain any of the words in the original keyword.
Keywords have not disappeared, but their role has evolved. They are becoming signals that help Google understand the market you want to enter rather than rigid commands defining every search you are willing to pay for.
This creates an important tension for advertisers.
AI-driven matching can identify valuable searches that no human could realistically list in advance. It can understand language variations, consider additional signals and respond to changes in the way people search. However, the same flexibility can also introduce irrelevant traffic, low-quality leads and wasted budget when campaigns lack reliable conversion tracking, negative keywords and commercial oversight.
The future of keyword strategy is therefore not completely manual or completely automated. It is a partnership between Google’s ability to process intent at scale and the advertiser’s understanding of customers, profit, lead quality and business priorities.
This guide explains how that partnership works—and why match types still matter in an increasingly AI-driven Google Ads platform.
What Is a Google Ads Keyword Match Type?
Google Ads keywords are words or phrases selected by an advertiser to help connect Search adverts with prospective customers.
A keyword is what the advertiser targets. A search term is what the user actually enters into Google.
Match type controls how closely the meaning of the search must relate to the keyword before the advert can be considered for the auction.
Google Ads currently supports three positive keyword match types:
| Match type | Example format | Relative reach | Strategic role |
|---|---|---|---|
| Exact match | [emergency plumber] | Narrowest | Stronger steering around a defined meaning or intent |
| Phrase match | "emergency plumber" | Medium | Controlled expansion around the meaning of the keyword |
| Broad match | emergency plumber | Widest | Wider discovery using intent, context and additional signals |
The match types overlap. Phrase match can reach the searches covered by exact match and additional related searches. Broad match can reach searches covered by both exact and phrase match, as well as further queries Google considers relevant.
Match type does not guarantee that your advert will appear. The search must still satisfy the campaign’s location, audience, budget, bid, policy and Ad Rank requirements.
The easiest way to understand the modern system is to view match types as different degrees of freedom.
Exact match gives Google the least freedom.
Phrase match gives Google more freedom.
Broad match gives Google the most freedom.
What Google does with that freedom depends heavily on the campaign’s structure, bidding strategy, landing pages and conversion data.
The Evolution of Keyword Match Types in the Age of AI
Before machine learning became central to Google Ads, keyword targeting was much more literal.
Advertisers often created large keyword lists because the platform needed explicit instructions. If potential customers could search for “accountant London”, “London accountant”, “accountants in London” or “London accounting firm”, each variation might be added separately.
This gave advertisers substantial control, but it also created complexity. Accounts could contain thousands of near-duplicate keywords divided across highly fragmented ad groups. Campaign managers spent considerable time expanding keyword lists, adjusting bids and moving search terms into ever more specific structures.
As Google’s language models improved, the platform became better at recognising close variants, synonyms and equivalent meanings. The boundary between keywords and search terms gradually became less literal.
Google can now consider signals such as:
The meaning and intent of the search
Close variants and alternative wording
Other keywords in the ad group
The content and purpose of the landing page
The user’s location and location intent
Recent search activity
The device, time and wider auction context
Predicted conversion probability and value
This changes the role of the advertiser.
The old challenge was to predict every relevant search and add it as a keyword. The modern challenge is to give Google a sufficiently clear theme, a commercially meaningful goal and accurate feedback about the people who become valuable customers.
In other words, Google Ads management is moving from controlling every individual input towards designing the system within which AI makes decisions.
That does not mean the human role has become less important. It means the work has moved up a level.
Instead of spending most of the time adjusting individual keyword bids, advertisers increasingly need to decide:
Which market and search intent the campaign should pursue
Which products, services and locations deserve separate control
Which searches should be excluded
Which landing pages best satisfy each intent
Which conversion actions should guide automated bidding
How Qualified Leads, sales and revenue will be returned to Google Ads
How much freedom the system should receive at each stage of campaign maturity
AI can process more auction signals than a person can evaluate manually. However, it does not automatically know your profit margins, sales capacity, ideal customer profile or definition of a Qualified Lead. Those commercial decisions must still come from the business.
The New Reality of Keyword Match Types
The names exact, phrase and broad remain, but their meanings have changed.
They should no longer be interpreted as fixed rules describing where the words must appear. They describe how broadly Google may interpret the relationship between the keyword and the search.
Exact Match Is Not Strictly Exact
Exact match uses square brackets:
[emergency plumber]
It provides the strongest degree of steering and normally the lowest reach of the three match types. However, it does not require a character-for-character match.
The keyword may potentially match searches such as:
emergency plumber
emergency plumbing service
plumber for an emergency
urgent plumber near me
These searches use different words, but Google may judge them to communicate the same basic intent.
Exact match is still useful when:
A keyword represents proven commercial demand
The budget is limited
The campaign has little conversion history
Search relevance must be tightly controlled
Manual CPC or Target Impression Share is being used
Brand traffic requires separate control
Different services or locations have materially different values
Exact match should therefore be understood as precise steering around a meaning—not a guarantee that the user typed the exact keyword.
Phrase Match Has Become Meaning-Based
Phrase match uses quotation marks:
"emergency plumber"
It provides a middle ground between exact and broad match. The search must contain the meaning of the keyword, but it may include additional words or express that meaning differently.
The keyword could potentially match:
emergency plumber in Bristol
local emergency plumbing company
call a plumber for an urgent leak
24-hour plumbing repair
Phrase match is no longer a simple instruction requiring the complete keyword to appear in a fixed sequence. Word order can still matter when it changes the meaning, but Google now interprets the overall intent.
Phrase match is particularly useful when:
You want more reach without immediately adopting broad match
The campaign is new or has limited conversion history
Prospective customers express the same commercial need in many ways
You want to discover longer-tail searches around a defined theme
You are using manual bidding or Maximise Clicks and require more control
You are expanding from a proven exact-match foundation
For many new lead-generation campaigns, exact and phrase match provide a sensible starting point. They allow the advertiser to learn from real search behaviour without immediately giving Google the widest targeting freedom.
Broad Match Is an AI-Driven Discovery Tool
Broad match uses the keyword without punctuation:
emergency plumber
It offers the widest reach and allows Google to consider more contextual signals. The search does not need to contain the same words as the keyword.
The keyword could potentially match:
urgent boiler leak repair
water pipe burst help
24-hour plumbing company
who fixes a leaking bathroom pipe
Broad match can uncover valuable demand that would be difficult to predict manually. It can learn from the landing page, other keywords in the ad group and the wider auction context.
However, discovery and risk arrive together.
The same flexibility that helps Google find a new customer can also expose the campaign to ambiguous, informational or unsuitable searches. Broad match is therefore most effective when supported by:
Accurate conversion tracking
Conversion-based Smart Bidding
Sufficient conversion volume
Realistic CPA or ROAS targets
Strong negative keyword coverage
Clearly themed ad groups
Relevant landing pages
Enough budget to explore additional auctions
Reliable lead-quality or revenue feedback
Our guide to broad match for lead-generation and service businesses explores these requirements in more detail.
The Shift From Keyword Matching to Intent Matching
The most important change is not that Google has made match types broader. It is that the system increasingly tries to understand why the person is searching.
Imagine two people enter different searches:
“emergency plumber near me”
“someone to stop a burst pipe tonight”
The words are different, but the underlying need is similar. A literal keyword system might treat these as unrelated. An intent-based system can recognise that both users may need the same service immediately.
This is where AI-driven matching can improve performance. Human language is too varied for any advertiser to list every possible expression of the same need. New products, changing terminology, voice search and conversational searches continually create variations.
However, intent is not always easy to interpret.
Consider a keyword such as business automation.
The searcher might want:
An automation consultant
Business automation software
Examples for academic research
A training course
A job in automation
Free templates
A definition of the term
All of these searches may be linguistically related, but they do not have the same commercial value.
This is the central challenge of modern keyword matching: relevance is not the same as profitability.
Google may correctly understand the topic while still choosing a search that is unsuitable for the business. The advertiser must therefore provide additional guidance through match types, negative keywords, campaign structure, bidding goals and conversion data.
Why Do We Still Use Match Types at All?
If Google increasingly understands meaning and predicts which users will convert, it is reasonable to ask why match types are still necessary.
The answer is that AI needs boundaries and direction.
Match types do not provide the rigid control they once did, but they still determine how much freedom Google receives. They remain valuable for several reasons.
Controlling Risk
Not every campaign is ready for the widest possible reach.
A new advertiser may have no conversion history, a limited budget and little understanding of which searches produce good customers. Exact and phrase match can reduce the cost of learning while the campaign establishes a reliable foundation.
An established advertiser with hundreds of accurate sales each month can give Google more freedom because the bidding system receives far more evidence about what success looks like.
Protecting High-Value Intent
Some searches deserve deliberate control because they represent strategically important products, services or audiences.
Exact match can help an advertiser steer campaigns around proven commercial themes, control brand searches or isolate services with different margins and budgets.
Testing Expansion
Match types provide a way to test how much additional reach improves performance.
An advertiser can begin with exact and phrase match, establish a baseline and then introduce broad match through a controlled experiment. The important question is not whether broad match produces more conversions, but whether it produces incremental Qualified Leads, customers and revenue at an acceptable cost.
Maintaining Insight
Keywords still provide useful reporting themes. They help advertisers understand which concepts are entering auctions and how broader searches relate to the campaign’s original targeting.
The Search Terms report then reveals the actual queries that generated traffic, allowing the advertiser to refine both targeting and exclusions.
Combining Human and Machine Intelligence
Google’s AI is strong at evaluating patterns across enormous datasets. The advertiser is better placed to understand customer fit, business strategy, operational limitations and profit.
Match types allow these two strengths to work together. The advertiser sets the direction and degree of freedom. Google evaluates individual auctions within that framework.
How Match Types Work With Bidding Strategies
Match type cannot be considered separately from bidding.
Broad match determines which additional auctions Google may enter. The bidding strategy determines how aggressively it competes in each one.
This relationship explains why the same match type can perform very differently in two campaigns.
Manual CPC
Manual CPC gives the advertiser direct control over keyword bids, but it does not automatically evaluate the predicted conversion value of each auction.
Exact and phrase match normally provide a safer structure in this environment. Broad match with manual CPC can expose the same bid to searches with very different levels of intent.
Manual CPC can still be useful for brand campaigns, low-volume markets or situations where direct bid control is required.
Maximise Clicks
Maximise Clicks aims to generate as many clicks as possible within the available budget and any bid limit. It does not primarily optimise for sales, Qualified Leads or revenue.
Combining Maximise Clicks with broad match can therefore give Google wide targeting freedom while asking it to favour click volume. Exact and phrase match usually provide a safer foundation.
Maximise Conversions and Target CPA
These strategies optimise towards the conversion actions included in the campaign’s goals.
They can work effectively with broad match when those conversions are accurate and commercially meaningful. However, the bidding system will pursue the outcome it is given.
If every form submission is recorded as a primary conversion, Google will try to generate form submissions. It will not automatically know which submissions became Qualified Leads or customers.
Maximise Conversion Value and Target ROAS
Value-based bidding allows Google to distinguish between outcomes with different values. This is particularly useful for ecommerce revenue, products with different margins or service businesses that import CRM stages and sales revenue.
Broad match can then explore additional demand while the bidding strategy evaluates the predicted value of each auction.
The strength of this approach depends on accurate values, sufficient volume and consistent data. Learn more in our guide to Google Ads bidding strategies.
The Increasing Importance of Online and Offline Conversion Tracking
As keyword matching becomes broader, conversion data becomes more important.
This is one of the biggest changes in modern Google Ads.
In the past, an advertiser could use restrictive keywords to control which searches entered the campaign. As Google receives more freedom to interpret intent, the platform needs reliable feedback about which users eventually create value.
Standard website conversion tracking can record actions such as:
Form submissions
Telephone calls
Appointment requests
Online purchases
Live-chat enquiries
These signals are useful, but for lead-generation businesses they describe only the beginning of the sales journey.
Two people may submit the same form. One could be a spam enquiry outside the service area. The other could become a customer worth £20,000. If Google receives only the original form submission, both appear to have equal value.
This creates a dangerous feedback loop.
Broad match reaches a wider range of searches. Smart Bidding looks for patterns among the people recorded as conversions. If the campaign counts poor-quality leads as successful outcomes, Google can learn how to generate more poor-quality leads efficiently.
The dashboard may show a falling cost per lead while the sales team reports that performance is getting worse.
Google Ads offline conversion tracking closes this gap by connecting the original advertising interaction with later outcomes recorded in a CRM, sales platform or payment system.
A modern lead-generation journey might look like this:
Google Ads click → Website enquiry → Qualified Lead → Sales Opportunity → Customer → Revenue
Enhanced Conversions for Leads can improve the accuracy of this connection by using consented, hashed first-party data alongside relevant advertising identifiers.
This allows advertisers to move beyond metrics such as:
Cost per click
Number of form submissions
Cost per lead
And evaluate:
Cost per Qualified Lead
Cost per Sales Opportunity
Customer acquisition cost
Sales revenue
Conversion value
Return on advertising spend
When these deeper outcomes are accurate and sufficiently frequent, they can also improve automated bidding. Google begins learning from the people who became valuable customers rather than everybody who completed a form.
This is why online and offline tracking are not merely reporting features. They are targeting inputs.
The wider Google’s matching becomes, the better the feedback loop needs to be.
How to Decide on the Right Match-Type Strategy
There is no universal match-type ratio.
An 80/20 or 70/30 rule may sound practical, but the number of keywords does not reflect their traffic, cost or commercial importance. One broad keyword can generate more impressions and spend than hundreds of exact-match keywords.
The correct balance depends on the maturity and economics of the campaign.
Start With the Business Objective
Decide what the campaign is genuinely expected to achieve.
Is the goal to generate visibility, website traffic, leads, Qualified Leads, sales or revenue? A campaign pursuing impression share requires a different match-type strategy from one using Target ROAS to maximise profit.
Assess the Quality of the Data
Ask whether the conversion data accurately represents business value.
If tracking is incomplete, duplicated or based only on low-value actions, wider matching can amplify the problem. Exact and phrase match provide stronger control while measurement is improved.
Consider Conversion Volume
AI requires evidence.
An ecommerce campaign generating hundreds of sales can learn much faster than a specialist B2B campaign producing a handful of opportunities. Low-volume campaigns often need more human steering because the bidding system receives fewer reliable outcomes.
Consider Search Ambiguity
Some markets have naturally clear commercial searches. Others overlap heavily with jobs, training, DIY, research, free information or unrelated meanings.
The more ambiguous the market, the more cautiously broad match should be introduced.
Consider Budget and the Cost of Learning
Broad matching needs room to explore. A campaign with a very small budget may spend too much of that budget testing auctions that never produce enough data to guide future decisions.
Where budget is limited, exact and phrase match can concentrate investment around the most defensible intent.
Consider the Bidding Strategy
Broad match becomes more credible when it is paired with conversion-based Smart Bidding and accurate outcome data. It is generally more risky with manual CPC, Maximise Clicks or visibility-based bidding.
Consider Lead Quality and Sales Feedback
For lead generation, do not assess match types using cost per lead alone.
Compare:
Lead-to-Qualified-Lead rate
Lead-to-opportunity rate
Lead-to-customer rate
Cost per customer
Revenue per lead
Revenue and profit by campaign or search theme
A search generating fewer, more expensive enquiries may be more valuable than one producing a high volume of cheap, unsuitable leads.
A Practical Match-Type Progression
The following approach works well for many new Search campaigns, although it should be adapted to the business.
1. Begin With High Commercial Intent
Start with the products and services most closely connected with revenue. Avoid broad informational terms simply to increase traffic.
2. Use Exact and Phrase Match for Initial Control
Create compact, closely themed ad groups using the strongest exact and phrase keywords. This gives the campaign enough variation to learn without immediately opening the widest reach.
3. Build Negative Keyword Coverage
Add known exclusions before launch, including unsuitable services, locations, job-related intent, training, DIY and other irrelevant themes.
4. Verify Conversion Tracking
Ensure conversion actions fire correctly, are not duplicated and are classified as primary or secondary appropriately. Low-value micro-conversions should not guide bidding accidentally.
5. Learn From Search Terms and Sales Outcomes
Review which queries generate clicks, conversions, Qualified Leads, customers and revenue. Add valuable themes where they deserve specific control and exclude unsuitable searches.
6. Introduce Broad Match as a Test
Once the campaign has a suitable bidding strategy, reliable data and enough budget, test broad match deliberately. Where practical, use an experiment or a defined comparison period.
7. Measure Incremental Commercial Value
Do not ask only whether broad match generated more conversions.
Ask whether it generated additional profitable customers that the existing targeting would not otherwise have reached.
8. Retain Human Oversight
Continue reviewing search terms, lead quality, locations, landing pages, conversion actions and budget allocation. Automation changes the management task; it does not eliminate it.
Negative Keywords Are the Counterbalance to Broader Matching
Positive match types tell Google which searches may be relevant. Negative keywords identify searches the campaign should avoid.
As positive matching becomes more semantic, negative keywords provide an essential layer of commercial control.
Google Ads supports negative broad, phrase and exact match, but these do not behave like their positive equivalents.
Negative keywords are generally more literal and do not automatically expand to close variants. If you exclude job, you may also need to add jobs, career, vacancy and other unwanted variations.
| Negative match type | Example | General behaviour |
| Negative broad | plumbing course | Blocks searches containing all terms, even in a different order |
| Negative phrase | "plumbing course" | Blocks searches containing the terms together in that order |
| Negative exact | [plumbing course] | Blocks that specific search without additional words |
Negative keywords can exclude themes such as:
Jobs, careers and salaries
Free products or services
Courses and training
DIY and informational research
Unsupported locations
Products or services the business does not provide
Low-value enquiries the business cannot serve profitably
The objective is not to restrict every possible variation. It is to prevent clearly unsuitable intent from consuming budget and contaminating the campaign’s conversion data.
Read our complete negative keyword guide and negative keyword list for further examples.
The Search Terms Report Connects Strategy With Reality
Keywords represent the advertiser’s targeting strategy. Search terms reveal what happened in the marketplace.
The Google Ads Search Terms report should be reviewed to identify:
Irrelevant searches to exclude
High-value searches that deserve their own keyword
New services, products or customer needs
Themes requiring separate adverts or landing pages
Unsupported geographic intent
Research, employment or DIY searches
Unexpected brand or competitor traffic
Differences between conversion volume and lead quality
However, the report should not be interpreted using clicks and website conversions alone.
Imagine one search term generates ten leads at £30 each, while another generates four leads at £50 each. The first appears more efficient.
If none of the first ten becomes qualified but two of the second group become customers, the more expensive search is commercially superior.
This is why modern Google Ads optimisation must combine advertising data with CRM and sales data.
How AI Max Changes the Match-Type Conversation
AI Max for Search takes intent-based targeting beyond conventional keyword matching.
Its search-term matching can use broad-match technology and keywordless matching to find additional relevant queries. Other features can adapt advert text and select a landing page that Google considers more relevant to the user’s intent.
This does not mean keywords or match types have become obsolete.
Existing keywords, ad-group themes, landing pages, brand controls, exclusions and conversion goals still help Google understand what the campaign is trying to achieve.
AI Max should therefore be viewed as an additional expansion layer. It increases the importance of:
Clear website structure and landing-page content
Accurate conversion tracking
Suitable Smart Bidding
Negative keywords and brand controls
Reliable lead-quality measurement
Controlled testing
The mistake would be to treat wider automation as a substitute for strategy.
Broad match and AI Max can find more opportunities, but they cannot independently decide whether those opportunities fit your margins, sales process or operational capacity. Learn more in our guide to making Google Ads AI work for your business.
Common Match-Type Mistakes
Assuming Exact Match Is Literal
Exact match can reach close variations and searches Google considers to have the same meaning. It still requires Search Terms review.
Using Broad Match Before Measurement Is Ready
Broad match amplifies the quality of the campaign’s inputs. Poor conversion data can produce poor optimisation at greater scale.
Combining Broad Match With Maximise Clicks
This gives Google wide targeting freedom while asking it to prioritise click volume rather than customers or revenue.
Treating Every Lead as Equal
Google cannot infer the sales team’s judgement unless meaningful CRM stages and revenue are returned to the platform.
Separating Every Match Type Into Different Campaigns
Duplicating the same themes across separate match-type campaigns can fragment data and create unnecessary management complexity. Separate campaigns when budgets, locations, goals, bidding or commercial priorities genuinely differ.
Adding Every Search Term as a Keyword
Modern matching does not require exhaustive keyword lists. Add a search term when it deserves distinct control, reporting, messaging or a dedicated landing page—not simply because it appeared once.
Applying Recommendations Without Commercial Review
Google may identify an opportunity to expand reach, but it cannot fully account for margins, lead quality, sales capacity and operational constraints. Recommendations should be evaluated and tested rather than applied automatically.
The Future of Google Ads Match Types
Google Ads is clearly moving towards broader interpretation, intent-based matching and greater automation.
However, it would be premature to declare that match types no longer matter or will inevitably disappear.
Their practical role is changing. Match types are becoming steering controls within a larger AI-driven system rather than rigid instructions defining every eligible search.
Advertisers are also gaining other ways to guide that system, including:
Conversion goals and values
First-party customer data
CRM lifecycle stages
Brand inclusions and exclusions
Negative keywords
Landing-page content
Audience signals
Location controls
The strongest advertisers will not be those who resist every form of automation or accept every automated recommendation.
They will be the advertisers who understand when to maintain control, when to expand reach and how to supply Google with reliable commercial feedback.
Every advertiser can access the same Google AI. Competitive advantage comes from giving it a clearer strategy, stronger data and better-defined outcomes.
Frequently Asked Questions
Which Google Ads Match Type Is Best?
There is no universally best match type. Exact provides the strongest steering, phrase offers controlled expansion and broad provides the greatest discovery potential. The right choice depends on campaign maturity, bidding, budget, conversion volume and data quality.
Is Exact Match Still Exact?
No. Exact match can reach close variants and searches Google interprets as having the same meaning or intent. It remains the narrowest positive match type, but it is not strictly literal.
Should a New Campaign Use Broad Match?
Not automatically. Exact and phrase match often provide a safer starting point for a new or low-data lead-generation campaign. Broad match becomes more credible when tracking, Smart Bidding, negative keywords and commercial outcome data are reliable.
Does Broad Match Require Smart Bidding?
Google permits broad match with other bidding strategies, but it is designed to work most effectively with conversion-based Smart Bidding. Using broad match with manual CPC or Maximise Clicks normally introduces greater risk.
Can Exact, Phrase and Broad Match Be Used Together?
Yes. However, it is not necessary to duplicate every keyword across all three match types. Each version should have a strategic purpose.
Are Negative Match Types the Same as Positive Match Types?
No. Negative keywords behave more literally and do not automatically match close variants in the same way. Plurals, synonyms and alternative spellings may need to be added separately.
Has AI Max Replaced Keyword Match Types?
No. AI Max expands Search targeting through broader search-term and keywordless matching, but exact, phrase and broad keywords remain available. Match types, campaign themes and controls still influence performance.
Conclusion
Google Ads match types have evolved from rigid word-matching rules into controls over how much freedom Google receives to interpret search intent.
Exact match provides the strongest steering, but it is not strictly exact. Phrase match expands around the meaning of a keyword. Broad match gives Google the greatest freedom to discover additional demand using context and AI.
That wider reach can be powerful, but only when the campaign provides the right foundations.
Start with clear commercial intent. Use exact and phrase match where control matters. Introduce broad match when the campaign has accurate conversion tracking, meaningful data, suitable Smart Bidding and enough budget to learn. Support every approach with negative keywords, Search Terms analysis, relevant landing pages and CRM feedback.
The goal is not to choose between human judgement and artificial intelligence.
Google is better at evaluating millions of auction signals. The advertiser is better at understanding customers, profitability and business priorities.
The strongest Google Ads strategy combines both.