Targeting in Google Ads – How to Reach and Convert Customers

Google Ads targeting is no longer simply a matter of choosing keywords, locations and audiences. Modern campaigns combine advertiser controls with Google’s machine learning, using search intent, first-party data, product information, landing pages, creative assets and conversion signals to decide when an ad should enter an auction.

That creates more opportunities to reach potential customers, but it also makes campaign structure and measurement more important. If the inputs are weak, automated campaigns can optimise towards the wrong searches, audiences or conversions. If the targeting is too restrictive, a campaign may never collect enough data to perform effectively.

The objective is therefore not to make every campaign as narrow as possible. It is to give Google Ads enough freedom to find demand while retaining the controls needed to protect relevance, budget and lead quality.

This guide explains the main Google Ads targeting options, how they differ by campaign type, and how to combine manual targeting with automation across the awareness, consideration and conversion stages of the customer journey.

What Is Google Ads Targeting?

Google Ads targeting determines which auctions a campaign can enter, which people or contexts it can reach, and which signals Google can use to predict whether an impression is likely to produce the desired result.

Depending on the campaign type, targeting may be based on:

• The words and meaning within a person’s search

• Products in a Google Merchant Center feed

• A person’s interests, behaviour or purchase intent

• Previous interactions with your website, app, YouTube channel or adverts

• Customer data supplied through Customer Match

• Website content, apps, YouTube channels or individual videos

• Geographic location, language, device and time of day

• Search themes, audience signals, landing pages and creative assets

• Conversion data and value signals used by Smart Bidding

Some of these settings impose firm restrictions. Others are signals that guide Google’s systems without limiting delivery to the selected audience. Understanding that difference is fundamental.

Targeting, Observation and Signals Are Not the Same

Google Ads uses several mechanisms that can look similar in the interface but behave differently.

With Search and Shopping campaigns, audiences can often be added in Observation mode. The campaign continues to reach people who meet its other criteria, while the audience data becomes available for analysis. Selecting Targeting instead narrows eligibility to members of the chosen audiences. Google provides a fuller explanation of Targeting and Observation settings.

Performance Max audience signals work differently. They help Google identify promising patterns and may accelerate learning, but the campaign can serve beyond the supplied segments when it predicts that doing so will achieve the conversion objective.

Match Targeting to the Customer Journey

The traditional AIDA model is still useful for copywriting, but Google Ads planning is often clearer when campaigns are organised around three commercially meaningful stages: awareness, consideration and conversion.

Awareness

At the awareness stage, the potential customer may fit the target market but is not yet actively searching for the product or service. Useful options include affinity segments, custom segments, YouTube content, selected placements, broad demographic criteria and Demand Gen campaigns.

The measurement focus is usually qualified reach, video engagement, brand search growth and assisted conversions rather than immediate last-click sales alone.

Consideration

At the consideration stage, users are researching solutions, comparing providers or showing relevant commercial intent. In-market segments, custom segments, product-led creative, non-brand Search campaigns, Demand Gen and remarketing can all play a role.

This stage benefits from more specific messaging: comparisons, benefits, proof, case studies, product features and reasons to choose the business.

Conversion

At the conversion stage, the strongest signals often come from high-intent searches, Shopping inventory, remarketing, Customer Match and first-party lifecycle data. Campaigns should be aligned with meaningful conversion actions such as qualified leads, sales opportunities, completed purchases and revenue—not simply every form submission or button click.

These stages overlap. A Search campaign may introduce a new brand and convert demand in the same interaction, while remarketing can support both consideration and conversion. The framework is intended to clarify the role of each campaign, not force every user into a perfectly linear funnel.

Search Campaign Targeting

Search advertising reaches people based primarily on what they are searching for. It is one of the strongest channels for capturing existing demand because the user reveals intent directly through a query.

Keywords and Search Intent

Google Ads keywords tell Google which searches may be relevant to an ad group. They are not the same as search terms: a keyword is selected by the advertiser, while the search term is what the user actually entered.

The three principal match types are:

• Exact match: provides the greatest control, although it can match close variants and searches with the same meaning or intent.

• Phrase match: reaches searches that include the meaning of the keyword, allowing additional words before or after it.

• Broad match: gives Google the widest interpretation and uses additional signals to identify related intent.

Broad match can work well with reliable conversion tracking, sufficient conversion volume and Smart Bidding. It should not be treated as an automatic upgrade for every account. Lead-generation campaigns with limited data, mixed lead quality or expensive clicks often benefit from establishing control with exact and phrase match before testing broader expansion.

The search terms report remains essential. It shows which user searches triggered adverts and helps identify new keywords, negative keywords, changes in intent and areas where match types are expanding too far.

Negative Keywords

Negative keywords prevent adverts from showing for unsuitable searches. They are critical for separating services, filtering research traffic, excluding jobs and training searches, and preventing budget from being spent on products or locations the business does not serve.

Negative keyword match types do not behave exactly like their positive equivalents, so exclusions should be reviewed carefully. An excessively aggressive shared negative list can suppress valuable demand across multiple campaigns.

Audience Targeting on Search

Audiences can be layered onto Search campaigns using either Observation or Targeting settings.

Observation is usually the safer starting point for a standard Search campaign because it preserves keyword reach while revealing how existing customers, website visitors, in-market users and other segments perform. Targeting is appropriate when the campaign is deliberately restricted—for example, a search remarketing campaign designed only for previous site visitors.

AI Max for Search Campaigns

AI Max adds a suite of expansion and creative features to Search campaigns, including broader search-term matching, text customisation and final URL expansion. It can help reach relevant queries beyond the account’s conventional keyword coverage, but it also changes how much control the advertiser delegates to Google.

It should be tested as a strategic expansion layer, not enabled solely because it is recommended in the interface. Before testing, ensure that conversion tracking represents genuine business outcomes, landing pages are suitable for final URL expansion, brand controls are configured where required, and search-term performance can be evaluated against lead or revenue quality. See Google’s current overview of AI Max for Search campaigns.

Shopping Campaign Targeting

Shopping campaigns do not use conventional positive keywords. Google primarily uses the Merchant Center product feed, landing page and user query to decide whether a product is relevant.

Important targeting inputs include:

• Product titles and descriptions

• Google product categories and product types

• Brand, GTIN, MPN and other identifiers

• Price, availability and condition

• Product images and landing pages

• Custom labels used to organise campaign structures

• Negative keywords where supported

This makes Google Shopping feed management a targeting discipline as much as a technical one. A vague title or incorrect product category can reduce relevance before bidding or creative optimisation even begins.

Standard Shopping provides more direct structural control, while Performance Max uses the feed alongside broader automation across Google’s inventory. The right choice depends on conversion volume, feed quality, budget, reporting requirements and the value of retaining campaign-level control.

Performance Max Targeting

Performance Max campaigns can serve across Google Search, Shopping, YouTube, Display, Discover, Gmail and Maps. They use goals, product feeds, creative assets, audience signals, search themes, landing pages and historical conversion data to predict where an advert is most likely to achieve the desired outcome.

Audience Signals

Audience signals can include:

• Customer Match lists

• Website and app visitor data

• Custom segments

• Detailed demographics

• In-market and affinity segments

These inputs help Google begin with a stronger understanding of likely customers. They do not normally limit the campaign to those people.

Search Themes

Search themes allow advertisers to provide words and phrases that customers use when looking for their products or services. They are optional, additive signals rather than a replacement for keywords, and Google currently permits up to 50 per asset group. Google’s search themes guidance explains how they interact with other Search campaigns.

Use tightly relevant themes that contribute information Google may not infer easily from the feed or landing page. Adding every possible keyword merely creates noise.

Asset Groups and Landing Pages

Asset groups should represent coherent products, services or customer needs. Combining unrelated offers in one asset group weakens the relationship between the audience signal, search theme, creative and destination page.

Final URL expansion can help Google match users with a more relevant page, but unsuitable URLs should be excluded. Websites with thin pages, outdated content or unclear conversion paths should be improved before they are used as inputs for broad automation.

Conversion Goals as Targeting Inputs

In Performance Max, the conversion goal is effectively part of the targeting. If the campaign is optimised for low-value actions, it will seek more people likely to complete those actions—even if they rarely become customers.

For ecommerce, pass accurate purchase values and margins where possible. For lead generation, connect the advertising platform with CRM outcomes such as qualified lead, sales opportunity and closed customer. This allows Smart Bidding to distinguish easy-to-generate enquiries from commercially valuable prospects.

Display Campaign Targeting

Display campaigns can target either people, content or a combination of both across websites, apps and other Google Display inventory.

Audience Targeting

Audience targeting focuses on who the person is or what their behaviour suggests. Options can include affinity segments, in-market segments, detailed demographics, custom segments and first-party data.

Content Targeting

Content targeting focuses on where the advert appears. The principal options are:

• Topics: groups of pages, apps or videos about a selected subject.

• Placements: specific websites, apps, YouTube channels or videos.

• Display or video keywords: words used to identify relevant content.

When multiple content methods are added to the same ad group, check how Google combines them. A campaign can become far narrower—or broader—than intended depending on its settings.

Optimised Targeting

Optimised targeting allows Google to reach people beyond manually selected audience segments when it predicts they are likely to convert. It can generate incremental conversions, but it should be evaluated separately from the seed audiences supplied by the advertiser.

For a tightly controlled placement campaign or a campaign where brand safety is particularly important, review whether optimised targeting is compatible with the objective. Google explains the current behaviour in its guidance on Display campaign targeting.

Demand Gen and YouTube Targeting

Demand Gen campaigns are designed to create and convert demand across visually led Google surfaces, including YouTube, Discover, Gmail and the Google Display Network. Audience options can include in-market segments, affinity segments, custom segments, detailed demographics, first-party data and lookalike-based signals.

Lookalike functionality in Demand Gen is evolving towards a signal-based model. A first-party seed still tells Google what a valuable customer looks like, but delivery may extend beyond a fixed similarity threshold to pursue the campaign goal.

YouTube targeting can also use topics, keywords, channels and individual videos. However, combining content restrictions with narrow audiences can severely limit reach. Build separate ad groups when you need to compare audience-led and placement-led strategies clearly.

For video, the creative itself performs part of the targeting. The opening seconds, message, format and call to action determine who continues watching and who self-selects out. Media targeting cannot compensate for an advert that is poorly matched to the viewer’s stage of awareness.

Google Ads Audience Segments

Audience availability varies by campaign type, objective, account eligibility and privacy requirements, but the main categories include the following.

Detailed Demographics

Detailed demographics extend beyond basic age and gender to characteristics such as parental status, education, homeownership or employment-related groups where available. Google may infer some attributes, and many users remain in an “Unknown” category.

Do not exclude Unknown automatically. Doing so can remove a substantial number of otherwise qualified users.

Affinity Segments

Affinity segments group people according to longer-term interests and habits. They are generally more suitable for awareness and broad consideration than for tightly controlled direct-response targeting.

In-Market Segments

In-market audiences are intended to identify people actively researching or planning a purchase within a category. They usually indicate stronger short-term commercial intent than affinity audiences, although they should still be validated against actual conversion quality.

Life Events

Life event segments cover major transitions such as moving home, graduating or getting married. They can be useful when the timing of a product or service is closely connected to that event.

Custom Segments

Custom segments let advertisers describe an audience using relevant search terms, websites or apps. The behaviour varies with campaign type: these inputs help Google interpret the intended audience, but they do not necessarily mean the user searched for those exact terms or visited the named websites.

This is the current successor to older “custom intent” terminology used in the original version of this article.

Your Data Segments

Previously known as remarketing lists, your data segments contain people who have interacted with your website, app, YouTube channel or other owned properties. They can support Google Ads remarketing, exclusions, retention and audience analysis.

Segment users by meaningful behaviour where volume permits. Someone who viewed one blog post should not automatically receive the same message as a user who abandoned a basket or requested a quotation.

Customer Match

Customer Match uses consented first-party customer data, such as email addresses or telephone numbers, to build audience segments. It can support retention, cross-selling, exclusions and automated campaign signals.

A carefully selected list of profitable customers is more useful than a mixed database containing customers, unqualified leads, suppliers and old enquiries. Data quality and consent are more important than list size alone.

Location Targeting

Location targeting can use countries, regions, cities, postcodes and radius targeting, subject to availability and minimum geographic requirements.

The most important setting is often the advanced location option:

• Presence or interest: can include people in the location as well as people who have shown interest in it.

• Presence: focuses on people in or regularly in the targeted location.

Google uses Presence or Interest as a broad default for supported campaigns. This can suit travel, education, property and other sectors where users research locations remotely. For a plumber, dentist or other local service provider that can only serve customers within a defined area, Presence is usually the more appropriate starting point.

Location signals are inferred and are not perfectly precise. Review the matched-location reports, exclude areas that repeatedly generate unsuitable traffic, and compare geographic performance with actual customer records. Google documents the distinction in its advanced location options.

Language, Device and Ad Schedule Settings

Language Targeting

Language targeting is based on signals such as the language a person uses and understands, not simply their physical location. Avoid selecting every language without considering whether the adverts, landing pages and sales process can support them.

Device Targeting

Device reporting normally covers computers, mobile phones, tablets and, for eligible campaign types, TV screens. Some campaigns support bid adjustments or more detailed device controls, while highly automated campaigns provide less manual control.

Do not exclude a device solely because it has a weaker last-click conversion rate. Mobile may initiate calls or research that later converts elsewhere. Compare lead quality, cross-device behaviour, page experience and call tracking before making a decision.

Ad Scheduling

Ad schedules control the days and times when eligible campaigns can serve. They are useful when enquiries cannot be handled outside business hours, conversion rates vary materially by time, or budget needs to be concentrated in the strongest periods.

Allow for account time zone, delayed conversions and limited datasets. A few conversions are not enough to conclude that an entire day or hour should be excluded permanently.

Exclusions Are Part of Targeting

Good targeting is defined as much by what a campaign excludes as by what it includes. Depending on campaign type, exclusions may cover:

• Negative keywords and brand exclusions

• Unsuitable geographic areas

• Existing customers when acquisition is the objective

• Job seekers, suppliers or irrelevant audience groups

• Mobile app categories or individual placements

• Sensitive content themes and inventory types

• Poor-quality landing pages through URL exclusions

• Products with low margins, poor availability or unsuitable economics

Use account-level exclusions for genuinely universal rules. Keep campaign-specific exclusions at campaign level so that one strategy does not accidentally suppress another.

How Smart Bidding Changes Targeting

Smart Bidding evaluates auction-time signals such as device, location, time, query context, audience characteristics and predicted conversion likelihood. This means bidding and targeting are increasingly interconnected.

The system can only optimise towards the conversion data it receives. If every lead is reported as equally valuable, it cannot know that one campaign generated sales-ready enquiries while another generated spam or low-value prospects.

For lead generation, improve the signal progressively:

1. Track the initial enquiry accurately.

2. Capture click identifiers and attribution data in the CRM.

3. Record qualified lead, sales opportunity and customer stages.

4. Return offline conversions through Google Ads CRM integration.

5. Include revenue or representative values where the data is reliable.

For ecommerce, use accurate purchase values, refunds and margin-aware data where feasible. Better commercial signals allow automation to target profitable outcomes rather than the largest possible number of nominal conversions.

A Practical Google Ads Targeting Process

1. Define the Commercial Objective

Specify the outcome the campaign must create: profitable sales, qualified leads, booked appointments, store visits, subscriptions or another measurable result. Avoid using traffic or impressions as a substitute when the actual objective is revenue.

2. Choose the Correct Campaign Type

Use Search to capture expressed demand, Shopping to match product inventory with product searches, Display and Demand Gen to build or recapture demand, YouTube for video-led reach and persuasion, and Performance Max when cross-channel automation is justified by the data and objective.

3. Map the Available Signals

List the assets available to guide targeting: keywords, product data, customer lists, website audiences, CRM lifecycle stages, locations, landing pages, videos and creative. Identify missing or unreliable inputs before launch.

4. Establish Controls

Configure locations, advanced location options, languages, negatives, exclusions, brand controls, budgets and conversion goals. Check Google’s automated defaults rather than assuming they match the business model.

5. Align Message and Landing Page

Targeting identifies an opportunity; the advert and landing page must convert it. Match each ad group or asset group to a coherent need and send users to the most relevant page—not automatically to the homepage.

6. Validate Tracking Before Scaling

Test forms, calls, purchases, values, enhanced conversions and CRM hand-offs. Remove duplicate or low-value conversion actions from bidding goals.

7. Expand in Controlled Tests

Once the foundation is reliable, test broader match types, new audiences, optimised targeting, AI Max or Performance Max in a way that preserves a useful comparison. Do not change every targeting and bidding variable simultaneously.

8. Optimise Using Business Outcomes

Review search terms, placements, locations, audience insights, product performance and lead quality. Optimise towards profit, revenue or qualified pipeline rather than click-through rate or platform-reported conversion volume alone.

Common Google Ads Targeting Mistakes

• Treating audience signals in Performance Max as strict targeting

• Using Search audiences in Targeting mode when Observation was intended

• Enabling broad expansion before conversion tracking is reliable

• Optimising lead-generation campaigns for all enquiries regardless of quality

• Using Presence or Interest for a local service that cannot serve remote users

• Failing to inspect search terms and placement reports

• Excluding the Unknown demographic category without sufficient evidence

• Combining too many unrelated products or services in one asset group

• Assuming a named website in a custom segment is a guaranteed placement

• Using automated recommendations without assessing commercial risk

• Making exclusions from small or statistically weak samples

• Allowing remarketing, prospecting and customer retention audiences to overlap without a plan

Frequently Asked Questions

What Is the Best Google Ads Targeting Option?

There is no universal best option. Search keywords are often strongest for capturing explicit demand, Shopping feeds for retail products, and first-party data for retention or high-quality automation signals. The best combination depends on the customer journey, campaign objective, available data and conversion volume.

Are Keywords Still Important in Google Ads?

Yes. Keywords still provide important control in Search campaigns, even as Google uses more intent-based matching and AI-driven expansion. Search terms, negative keywords, landing pages and conversion data now work alongside the keyword rather than making it obsolete.

Do Performance Max Audience Signals Restrict Who Sees Ads?

No. Audience signals generally guide Performance Max towards useful starting points, but Google may serve adverts beyond those audiences when it predicts that doing so will achieve the campaign goal.

Should Search Audiences Use Targeting or Observation?

Observation is usually appropriate when you want audience reporting without restricting keyword reach. Use Targeting when the campaign is intentionally designed to reach only the selected audience segments.

Is Broad Match Better Than Exact or Phrase Match?

Not automatically. Broad match provides greater reach and can perform strongly with Smart Bidding and high-quality conversion data. Exact and phrase match remain valuable when control, query relevance, limited data or lead quality are major concerns. Test expansion based on account evidence rather than treating it as a default requirement.

How Often Should Targeting Be Reviewed?

Monitor new or changed campaigns frequently enough to catch wasted spend, but allow for conversion delays and normal variation. Mature campaigns should still receive regular reviews of search terms, placements, locations, audiences, products, exclusions and conversion quality.

Final Thoughts

Effective Google Ads targeting is a balance between control and learning. Manual controls protect relevance and budget; automation can evaluate more signals and auctions than a person could manage individually. Neither works well without accurate measurement and a clear commercial objective.

Start with the strongest evidence of intent, build clean campaign structures, supply reliable first-party and conversion data, and expand only when performance supports it. This creates a targeting system that does more than generate clicks: it helps Google Ads find the customers most likely to create genuine value for the business.

If you need help reviewing targeting, search terms, conversion tracking or campaign structure, One PPC can carry out a comprehensive Google Ads audit and identify where budget, data and automation can be improved.