Contextual Whitelists

Contextual Whitelists

PubMag introduces a new tool for building specialized whitelists — based not on formal categories or subjective judgment, but on actual audience interests.

At its core is the analysis of search patterns: how users phrase their queries, which topics dominate, and which websites consistently appear in relevant contexts.

The result is topic-focused site lists that are genuinely aligned with user intent. No fluff, no irrelevant matches — only those domains that truly "live" in the relevant audience.

An additional advantage is technological transparency. Every site in the whitelists is enriched at the infrastructure level using previously collected data. Context as the entry point.

PubMag’s proprietary tools enable you to:

  • HBTracker: shows through which SSPs (Header Bidding adapters) inventory is available on specific sites;
  • LibTracker: detects which advertising and analytics libraries are used on a site;
  • Search patterns: help identify where inventory for specific topics is concentrated.

All of this creates opportunities for more precise media buying, contextual targeting, and reducing inefficient impressions.

Example: the "Cars" cluster (full category list here).

Websites were gathered across hundreds of user interests — from purchasing an electric car to comparing insurance providers. The final selection highlights strong domains that consistently appear in the context of these queries. Random or irrelevant domains are filtered out. What’s left is an ideal base for targeted ad campaigns.

Who benefits from this:

  • Agencies — looking for inventory tailored to briefs and audience segments,
  • Ad platforms — setting up contextual access to inventory,
  • Media researchers — mapping audience interests and advertising landscapes.

PubMag is an independent project. No magic. Just data work and attention to detail.

If you see value in this for your needs — I'd be happy to talk, show examples, and explore potential collaborations.

Autonomous Operation of HBTracker
Battleboard: February 2025 — Two Kings and Their Court
January Results and Plans for February
January Holidays
How I Linked Data from HBTracker and LibTracker and What Came Out