Context Studio is Coactive’s application for translating advertiser briefs into contextual ad packages. It enables ad sales teams to automatically match brand intent to the right content moments across a large video library.
Advertisers don’t buy demographics, they buy moments. A beer brand wants to reach viewers during sports highlights. A travel company wants to be next to aspirational adventure content. But traditional contextual targeting relies on coarse genre and show-level metadata that doesn’t capture what’s actually happening on screen.
The result is an intent gap: the distance between what a brand wants and what ad tech can actually find.
Context Studio closes that gap by using Coactive’s Multimodal AI Platform (MAP) to analyze video content at the shot level — understanding visual scenes, spoken dialogue, and on-screen context together — and then surfacing the exact moments that match an advertiser’s brief.
With Context Studio, you can:
Context Studio is designed for two audiences who work together:
Ad Sales Teams use Context Studio to build and pitch contextual packages to advertisers. It gives you the language and data to articulate why a piece of content is right for a brand. It’s not just saying “sports content” but giving the opportunity for the user to say I want to find “high-energy moments with crowd celebration during championship games” in my content library.
Ad Operations / Technical Teams configure the underlying targeting logic, set confidence thresholds and aggregation methods, and manage the export pipeline to your ad server.
Coactive organizes video content into four levels:
Context Studio can target at the Video level or the Scene level. Scene-level targeting requires publishers to provide ad break timecodes via the Segment API — see the Scene Segmentation Guide.
Dynamic Tags are AI-powered content classifiers created and managed in Coactive’s Video-Level Analysis (VLA) tool. Each tag is a visual or transcript-based concept — for example, outdoor_adventure, brand_safety_violence, or luxury_items.
Tags are the building blocks of every Context Studio package. Before you can use a tag in Context Studio, it must be in Validated status in VLA.
When you review a tag’s performance in VLA and confirm its results look accurate, you mark it Validated (shown with a purple indicator). This is a quality gate because it signals that the tag has been reviewed and is ready for use in ad targeting packages.
Tags that are not yet Validated will not appear in Context Studio. If a tag you need is missing, check its status in VLA first.
Each tag in a package has an aggregation method — the rule that determines how shot-level confidence scores are combined to decide whether a video or scene “passes” for that tag.
You set the aggregation method and its parameters (threshold %, minimum seconds) when adding a tag to a package.
The Match Percentage threshold controls how many of your inclusion tags a given video must satisfy to be included. For example, if you have 5 inclusion tags and set the threshold to 60%, a video that matches 3 of the 5 tags will qualify.
This is useful when you want broad reach and some tag flexibility, versus strict mode where content must satisfy every inclusion criterion.
Context Studio is an application layer built on top of VLA (Video-Level Analysis). Before you can build a Context Studio package, you need:
If you’re setting up VLA and Dynamic Tags for the first time, see the Dynamic Tags guide before returning here.
Once a package is built and reviewed in Context Studio, you export an Export Bundle. This is a structured JSON file containing all matching videos, their metadata, and the specific tag moments that triggered the match. This bundle can automatically be saved to an S3 bucket to be sent to your ad server.
Ready to build your first package? Head to the Context Studio How-To Guide for a step-by-step walkthrough.
For targeting at the scene level (individual ad slots within an episode), see the Scene Segmentation Guide.