AdTech Targeting Methods: A Practical Guide

Before launching a campaign, advertisers need to decide which targeting methods to deploy. Behavioural targeting? Demographic targeting? A mix of both? Choosing wrong — or choosing nothing deliberately — means wasting budget on the wrong audience.

Advertising technology has given marketers an extensive toolkit, but the real skill lies in knowing how to define a target audience and how to source the data needed to reach it accurately. No single method works in isolation; most effective campaigns rely on a combination of tactics. What follows is a thorough walkthrough of the most widely used AdTech targeting approaches, how each one works, and when it makes sense to use it.

Contextual Targeting

Contextual targeting is one of the most fundamental methods of segmenting audiences for advertising campaigns. Ads are matched to their surrounding content, which tends to produce a more cohesive, less disruptive user experience.

There are several levels of contextual targeting, each with distinct uses and advantages.

Site category targeting: At the broadest level, ads can be targeted based on the site category of the publisher's site or application. The Interactive Advertising Bureau (IAB) has outlined an extensive list of categories and subcategories that most reputable publishers use to define their content.

Domain/app-specific targeting: A campaign can be aimed at a single domain or application. The reach is relatively narrow, but it is useful for reaching visitors to sites with a direct connection to the advertised product or service.

Page-content targeting: This involves honing in on specific pages or sections within a site or across multiple sites. It applies to text ads as well as rich-media campaigns, and is typically executed through keyword matching:

• Text content — page keywords, title tags, or alt text for images
• Rich-media content — video titles, tags, or the presence of video and other rich media

While Google AdWords is the most recognizable example of keyword-based contextual targeting, virtually all demand-side platforms (DSPs) allow advertisers to insert keyword lists and target page content or specific sections based on those keywords.

Time- and Event-Based Targeting

For products tied to specific occasions, consumption moments, or promotional windows, targeting based on time or external conditions — weather, major sporting events, seasonal factors — can be highly effective.

• A beer brand might run a weekend promotional campaign timed around football broadcasts.
• A car repair service might serve ads immediately following a significant storm.

Programmatic media buying and selling has enabled time- and event-based targeting at a level of precision that simply wasn't possible before automated infrastructure existed. Ignoring these capabilities is leaving real campaign performance on the table.

Behavioural Targeting

Reaching customers based on past behaviour — and increasingly, predicted future behaviour — remains the most valued form of audience targeting for many marketers. No other method provides better insight into customer interests, preferences, and customer intent: for example, signals indicating when a person may be ready to book a flight, reserve a hotel, or make a large purchase.

Behavioural targeting also demands the most data and the most advanced data management infrastructure to execute properly.

For sites and apps with high traffic volume, behavioural targeting based exclusively on first-party data is both feasible and highly valuable. Companies with lower traffic volumes, however, will need to supplement with third-party data vendors to achieve the reach their campaigns require.

When brands want to use behavioural data they haven't collected themselves, they access and purchase it through a DSP in predefined bundles categorized along lines such as:

• Interests → Sports
• Intent → Hotel booking

The data vendor building these segments draws on a variety of behavioural signals. The main drawback is that the advertiser typically has limited visibility into the exact data used to build a segment, meaning the match quality between the segment and the brand's ideal customer is not always guaranteed.

First-Party Data Targeting

Segment-based targeting using first-party data involves targeting users based on a single condition or a group of conditions — often behavioural data, but frequently a combination of behavioural and location information.

First-party data segments can be exported directly from web analytics tools such as Google Analytics, Piwik PRO, or WebTrends, and activated through a data management platform (DMP) into a DSP, or through a hybrid DMP/DSP system.

Depending on the volume and richness of data available, first-party segments can be quite granular. The main data sources break down as follows:

1) Interaction with site or app content

Behavioural measurement starts with recording how visitors interact with content. At the general level this includes:

• Pageviews
• Product views
• Session duration

More specific signals include:

• In-site search queries
• Scrolling behaviour (for written or visual content)
• Video view duration

Even more concrete data comes from on-site or in-app conversions, which fall into two categories:

• Macro conversions: downloading a whitepaper, signing up for a newsletter, making a purchase, or redeeming a coupon
• Micro conversions: clicking a "related posts" link on a blog, viewing contact information on a company's site, and similar smaller engagement actions

2) Ecommerce behaviour

Sites and apps that sell products or services online offer especially strong signals of customer interest and intent. Relevant data points include:

• Items viewed (pageviews)
• Items added to shopping cart
• Average order value

3) Interaction with ad content

Clicking through (or declining to click) display ads, search ads, or in-app ads, and viewing video ads beyond a certain duration — this information can be sent back to a brand via a conversion pixel, which captures click-through and view data and aggregates it with other behavioural data in a DMP.

4) CRM data

Beyond online first-party data, brands often hold valuable customer information in CRM platforms. A customer who has consistently made large in-store purchases, for example, becomes a high-value target for an online promotional campaign. Because CRM data can be personally identifiable, it must be anonymized before activation — a process commonly handled using tools such as LiveRamp.

Demographic Targeting

Demographic targeting — based on location, gender, language, age, income, ethnicity, or life status — can appear less sophisticated than behavioural approaches, but it is no less important. Combined with behavioural data, campaigns aimed at specific demographic groups can be particularly powerful. And without demographic information, an ad can be entirely misdirected.

A German-language whitepaper served to someone based in Tennessee illustrates the point well enough.

Gender and Age

Many products and services are designed for specific genders or age groups, making these dimensions fundamental to effective targeting.

For desktop browsers, age and gender are harder to determine directly, though both can be inferred from behavioural signals such as ad conversions and product views. On platforms like Facebook or Pinterest, users supply this information during registration, making it readily available to advertisers.

Some platforms do not explicitly collect gender — Snapchat is a notable example — but may infer it through behavioural data and third-party data services.

Language

Language is a basic but critical targeting criterion. For desktop users, it can be drawn from browser settings; for mobile users, from the app's language setting.

Income

Income data — typically provided in ranges — can be derived from several sources with varying degrees of accuracy. Location-based estimation (by city, country, or neighbourhood) offers a rough proxy, though precision is limited. More accurate data can come from ISPs or cable TV service providers, who know which service packages customers have subscribed to.

Other Demographic Segments

Targeting can also reflect ethnicity and life status (e.g., single, married, married with children), to the extent consumers have provided this information. Ethnicity-based targeting in particular has drawn scrutiny and regulatory attention in the context of housing, employment, and credit advertising.

Geolocation Targeting

"Location, location, location" hasn't lost its relevance. Geolocation is among the most dynamic targeting conditions available — it can change rapidly, particularly with the growth of mobile internet usage — and it is especially valuable for local businesses.

IP-based geolocation: A user's IP address provides a general sense of location and can be used effectively for country-, city-, or region-level targeting.

Hyperlocation (mobile apps): For mobile applications, GPS-based targeting enables far greater precision. When users enable location services, it is possible to serve location-specific ads relevant to their immediate surroundings. Hyperlocation often employs geofencing to narrow targeting to a specific area — down to a few hundred metres in some cases.

Geolocation is also a form of contextual targeting: ads for certain products can be shown based on known characteristics of a user's surroundings, such as the average household income of the neighbourhood.

Device Targeting

Device targeting is particularly useful for mobile campaigns. It enables advertisers to serve locally relevant ads, contextual ads, or promotional coupons tied to stores or restaurants near the user's current location.

A closely related tactic is targeting by connection type — distinguishing between users on Wi-Fi and those on a GSM (cellular) data connection. This distinction matters less for identifying high-value customers than for determining what ad creative is appropriate to serve.

Facebook, for example, offers advertisers differentiated targeting options for mobile users on GSM connections: 2G, 3G, and 4G. A user on a slow 2G connection is generally not an ideal target for a media-heavy, high-bandwidth ad creative.

Ad Blocker Targeting

One of the newer — and somewhat counterintuitive — targeting methods involves focusing specifically on users who have installed ad-blocking software. Several factors make this segment commercially interesting:

• Self-identification: Users with ad blockers installed effectively identify themselves. Given that large publishers actively work to detect and work around ad blockers, the presence of blocking software is itself a reliable signal.
• Demographic skew: A majority of ad blocker users tend to be younger and more tech-savvy, placing them automatically into a distinct, targetable segment.
• Preference signal: The use of an ad blocker signals a preference for non-intrusive, higher-quality ad experiences. Publishers can leverage this by offering premium ad placements to advertisers with the budgets to produce top-quality creatives.

Some publishers — Forbes is a well-documented example covered by the WSJ — have adopted a "bargaining" approach: offering ad-blocker users access to content in exchange for an ad-light experience, then charging premium rates for the ads served within that experience.

Retargeting

Retargeting means serving ads to users who have previously interacted with a brand in some way. Its primary focus is customer retention and converting potential customers who have already engaged — which is why it tends to be effective for businesses of all sizes, particularly when used alongside acquisition-focused campaigns.

Retargeting is typically implemented by setting a cookie when a visitor completes a specific action on a site or engages with content in a defined way. The threshold for triggering retargeting can be set at a single interaction or a sequence of interactions, depending on how the brand wants to evaluate potential customer value.

Platforms such as Facebook extend retargeting further by enabling advertisers to target users based on email address lists — collected from newsletter subscriptions, ecommerce account registrations, or in-person interactions later entered into a CRM database.

The targeting methods covered here represent the core toolkit available to programmatic advertisers today. In practice, the most effective campaigns rarely rely on any single approach. Combining contextual signals with behavioural data, layering demographic filters, or pairing geolocation with device type — these combinations are where campaign performance tends to compound. Understanding each method individually is the prerequisite for knowing how to combine them intelligently.