Contextual advertising
Contextual advertising places ads based on the content and environment where they appear, such as the topic of a page, video or app, rather than on a profile of the individual user. A cookware ad next to a recipe article is the classic case.
What is contextual advertising?
Contextual advertising is one of the oldest forms of ad targeting, now revived by privacy changes. Instead of asking who the user is, it asks what the user is engaging with right now. Modern contextual systems go beyond keywords to classify topics, sentiment, imagery and video scenes, often using standard taxonomies so signals travel through OpenRTB bid requests. Context also supports Brand suitability, helping buyers avoid content that conflicts with their values. It is commonly used alongside or instead of audience targeting.
How contextual advertising works
Content is analysed with keyword, category and natural language or image analysis to assign topics, sentiment and suitability labels. Those signals are passed in bid requests or used in deal packages. Buyers then target or bid on placements whose context matches their brand and message.
- The page, app screen or video is analysed to identify topics, entities, sentiment and suitability.
- The resulting labels are attached to the inventory, often using standard content taxonomies.
- Labels are passed to buyers in bid requests or built into contextual deal packages.
- Buyers target, bid on or exclude placements according to those contextual signals.
Why it matters
Contextual targeting does not rely on tracking individuals across sites, which makes it resilient to privacy rules and identifier loss. Relevant context can also improve how audiences receive an ad.
For advertisers
Contextual targeting reaches people in relevant moments without depending on personal identifiers.
For publishers
Rich, accurate content signals let publishers package and price inventory for buyers who value specific environments.
Practical example
An outdoor gear brand targets articles and videos about hiking and camping, without using any user-level data.
Key considerations
- Keyword-only approaches can misread meaning, so semantic analysis gives better accuracy.
- Context indicates interest in the moment, not purchase intent over time.
- Video and CTV context is harder to analyse than text and may rely on publisher-supplied metadata.
Frequently asked questions
Does contextual advertising use cookies?
It does not need them. Targeting is based on the content being viewed, though it can be combined with other signals where consent allows.
How is contextual different from behavioural targeting?
Contextual targeting uses the current content, while behavioural targeting uses a user’s past activity across sites or apps.
Can contextual advertising work on CTV?
Yes, using programme metadata such as genre and rating, and in some cases scene-level analysis of the content.