Targeting is the practice of choosing which group of people sees an ad or a piece of content, based on traits such as location, age, interests or past browsing behavior, so it only reaches the people most likely to respond to it.
The barn doors decide where the light does not fall
The targeting types in use today, why behavioral targeting keeps losing ground to contextual targeting now that Chrome has decided to keep third-party cookies, and what EU law requires before you switch either one on.
The barn doors decide where the light does not fall
What targeting means
Targeting comes from the English word target and means deciding which group of people an ad or a piece of content reaches, instead of showing it to everyone who lands on a page or watches a video. That decision rests on traits such as age, location, language, stated interests or prior browsing behavior.
It gets confused often with segmentation, but the two are separate steps in the same process: segmentation splits an entire audience into groups that share traits, targeting picks which of those groups a specific campaign actually goes after. A store might segment its visitors by age and location, then target only the group of women aged 25 to 34 in Chicago for a single ad.
It's not personalization either, which adapts content to one individual in real time within the same visit. Targeting decides, before a campaign launches, which group it reaches at all, personalization decides afterward, during that same visit, which exact version a given person sees. In practice campaigns tend to combine both, guided by the buyer persona that defines who a brand wants to reach in the first place.
Targeting versus segmentation and personalization
The three terms get thrown around interchangeably in briefs and meetings, but each one answers a different question within the same marketing process.
Splits the whole audience into groups with shared traits
Analysis stage, ahead of any campaign
Targeting
Picks which specific group a campaign reaches
While setting up the campaign, before launch
Personalization
Adapts content to one person within that visit
In real time, while the person browses
The most common mistake is treating targeting and segmentation as interchangeable: a poorly built segmentation drags its flaws into any targeting built on top of it, no matter how sharp the rest of the campaign looks.
How targeting works
There are several targeting types, and most campaigns combine more than one. Demographic targeting relies on demographic data such as age, gender or education level. Geographic targeting, or geo-targeting, limits an ad to a country, a region, or a radius around a physical store. Contextual targeting looks at the content of the page where the ad appears, without knowing who's reading it: according to Google Ads Help, the system checks each page's main topic to match a relevant ad to it. Behavioral targeting draws on one person's browsing history across multiple sites, and retargeting or remarketing shows an ad again to someone who already visited a site or worked through a CRM flow without finishing a purchase.
Behavioral targeting leans hardest on third-party cookies to recognize the same person across different sites. That model has been eroding for years because Safari and Firefox block those cookies by default. Chrome, the browser with the largest share, held out longer, until Google announced on April 22, 2025 that it would neither remove third-party cookies from Chrome nor roll out a separate prompt to manage them. They remain active there for now.
Even so, the industry keeps shifting toward contextual targeting and toward interest signals computed on the device itself, such as Google Ads's Topics API, which assigns broad interest topics without building an individual profile or handing a full browsing history to every advertiser.
Why targeting matters
Precise targeting cuts the ad spend wasted on people who will never use the product, and it improves metrics like click-through rate or conversion rate, two KPIs that show up in nearly every campaign. The tighter the group, the cheaper it gets to earn the attention of someone who actually fits the product.
There's a legal limit too. The EU's ePrivacy Directive requires a user's prior consent before storing or reading information on their device, the technical basis behind most cookie-driven behavioral targeting. On mobile, Apple has required since April 2021, with iOS 14.5, that every app ask explicit permission before accessing the device's advertising identifier, and most users decline when asked directly.
That legal limit isn't just paperwork: targeting that ignores a user's consent exposes a brand to fines, and targeting that's too precise breeds distrust in people who notice an ad following them from page to page. The goal isn't tracking as much as possible, it's getting the call right with less data.
Best practices
Combine contextual targeting with first-party data instead of relying solely on third-party cookies, which keep getting less reliable outside Chrome.
Configure consent mode before turning on any behavior-based targeting, so the data reflects users' actual choices.
Review which audience segments each live campaign actually uses, instead of leaving whatever the original template defaulted to.
Test geographic and demographic targeting before behavioral targeting on new campaigns, since it needs less data infrastructure and carries less legal risk.
Measure results per segment against a concrete metric, such as cost per conversion, instead of trusting the feeling that "the ad fits".
Document the legal basis for every active targeting type, especially anything relying on advertising identifiers or cookies.
Common mistakes
Confusing targeting with segmentation and building a campaign on a segmentation nobody has reviewed in months.
Still betting the whole budget on behavioral targeting through third-party cookies, with no backup plan outside Chrome.
Turning on remarketing without a frequency cap, until the same ad follows a user for weeks.
Launching targeting campaigns before consent mode is configured, which artificially inflates or shrinks the available data.
Treating interest-based Topics API targeting as if it delivered the same level of detail as an individual, cookie-built profile.
Manuel Riveiro RodriguezCEO & Digital Strategist
A technical audit covers this and everything else in one pass.
What's the difference between targeting and segmentation?
Segmentation splits the whole audience into groups sharing traits like age or location. Targeting is the next step: deciding which of those groups a specific campaign goes after. Segmentation gets built once and reused, targeting changes with every campaign depending on its goal.
Are third-party cookies still active in 2026?
In Chrome, yes. Google announced on April 22, 2025 that it wouldn't remove them from the browser or roll out a separate choice prompt for them. Safari and Firefox, by contrast, have blocked them by default for years, which is part of why contextual targeting has gained ground.
Do I need user consent for behavioral targeting?
Yes. The EU's ePrivacy Directive requires prior consent before storing or reading information on a user's device, the technical basis behind most tracking cookies. On apps, iOS has required an additional explicit permission since 2021 before accessing the advertising identifier.
What is contextual targeting?
It picks where an ad appears based on a page's content, not on who's reading it. A running shoe ad next to a marathon article is contextual targeting. It needs neither cookies nor identifying the user, which is why it's gained ground over behavioral targeting.
Does the Topics API fully replace cookie-based behavioral targeting?
Not at the same level of detail. It computes broad interest topics right on the device, without building a full browsing history or handing it entirely to every advertiser. That trades some precision compared to an individual profile for exposing far less personal data.
