WhoisXML API, LUMS & UC Davis: Adding Website Context to Meta Pixel Privacy Research
About
Researchers Abdullah Ghani from Lahore University of Management Sciences (LUMS) and Yash Vekaria and Zubair Shafiq from the University of California, Davis developed PixelConfig, a framework for analyzing how Meta Pixel is configured across websites.
Their study examined how Meta Pixel tracking features and privacy restrictions evolved over time, with particular attention to differences between health-related websites and the broader web. WhoisXML API’s Website Categorization API helped them understand the types of websites represented in their general-web control group.
- Location
- U.S., Pakistan
- Products used
Highlights
- Website Categorization API classified the researchers’ Top 10K control websites by business or content category.
- The categorization showed that the control dataset included substantial representation from commercial and general-web categories, while Health accounted for 2.04% of the total websites.
- This category context supported the researchers’ comparison of Meta Pixel configurations on health-related websites and the broader web.
Adding Context to a General-Web Control Group
The researchers wanted to understand how Meta Pixel configurations had changed over time and whether they differed between health-related websites and popular websites more generally.
For this comparison, they assembled 18,327 U.S.-focused health websites and used the Tranco Top 10K websites as a control group. To better understand what that control group represented, they needed a scalable way to classify the websites by their primary business or content categories.
Categorizing the Top 10K with WhoisXML API
The researchers used WhoisXML API’s Website Categorization API to classify the 10,000 websites in the control group.
Each website was queried through the API, which returned one or more standardized category labels when it could confidently classify the site. The researchers then analyzed these categories and consolidated several health-related labels, including Healthcare Industry, Pharmaceutical Industry, Nutrition, Wellness, and Eldercare, into a broader Health category.
This gave them additional context on the composition of the control dataset used in their Meta Pixel analysis.
Insights Enabled by WhoisXML API Data
Website Categorization API enabled the researchers to characterize the types of websites represented in their Top 10K control group.
Characterizing the Control Dataset
The analysis found that 64.24% of the websites remained uncategorized. Among the categories identified, Business and Finance accounted for 32.36%, while Shopping and Retail Industry each accounted for 27.11%. Websites could receive more than one category.
The consolidated Health category accounted for only 2.04% of the total Top 10K websites.
Supporting the Broader Comparison
The category distribution showed that the control group contained substantial representation from business, retail, clothing, technology, and other non-health sectors.
This helped establish that the researchers were comparing healthcare websites against a broader general-web baseline rather than another health-heavy dataset, strengthening the context for their analysis of whether Meta Pixel tracking features and privacy protections differed on healthcare websites.
Learn how WhoisXML API intelligence can help you achieve success