Methodology

📘

Scope3's Methodology

Scope3’s brand suitability classification is built on full transparency, client accessibility, and AI-driven decisioning. It evaluates every advertising placement using a multi-layered classification framework, ensuring brands have complete control over their suitability settings.

Scope3's methodology is built to empower advertisers allowing them to define their own brand alignment preferences and continuously refine classification models.

Scope3's differentiate vertical-specific models, bespoke models from common sense brand safety. Common sense brand safety is a natural prerequisite excluding any content not suitable for any advertising by default (ie. CSAM, explicit violence, terrorist material) while maintaining neutrality and protect legitimate discourse.

Scope3's AI-driven capabilities allow nuanced analysis of all content elements (text, images, audio, video) to provide a comprehensive safety and suitability assessment for advertisers. Using Scope3's framework both advertisers and publishers benefit from greater transparency, content is no longer flagged without a clear explanation and decision which can be overruled on both ends. Advertisers can decide what is aligned or not with their brand preferences at any stage of the process from agent creation to model calibration and fine tuning. Publishers can see why a specific is article is flagged and report misclassification in Scope3's platform.

Content Ingestion

Scope3 ingests and processes content from multiple sources to provide a complete coverage across the digital ecosystem. This is performed in real-time, when possible, or asynchronously. Among the sources utilized by Scope3 we can list the following ones:

  • Real-Time Data Provider (RTDP) logs — Captures live content signals from (bid-stream) websites and apps (open web).
  • JavaScript (JS) — Captures live content signals from websites and apps via Scope3’s measurement trackers live in the wild.
  • Manual Uploads — Allows advertisers to submit artifacts and/or specific URLs or any content for direct classification.
  • API connections — Captures ad adjacent content signals from Walled Gardens and Social Platforms where Javascript measurement is not allowed.

Ensuring that the Scope3 Crawler can properly see content

We strongly recommend reaching out to Scope3 to set up a publisher account, which includes full transparency into brand safety and vertical model classifications. It also includes, optionally, the ability to integrate into your ad server for direct-sold campaigns and to integrate directly into your CMS to avoid the need for crawling entirely.

We also recommend adding the Scope3 Crawler to your list of approved crawlers:

"Scope3/2.0 (scope3.com) "