The rise of GDPR, recent government regulations, and public sentiment demanded a change in how ads are targeted. That's why we asked "How can we deliver relevant, helpful ads at scale without personal data?" First piloted in 2019, Pegasus collects articles from a list of partner publications that include multidimensional contextual tags, such as keywords, visuals, content, enriched data, tags, and metadata. These contextual tags, rather than users’ personal data, determine which ad version gets served, based on the creative strategy for the particular campaign. For this campaign, we expanded the portfolio of titles we worked with to 20 Vox and Conde Nast publications compared with just five publication partners during previous campaigns.

Our target audience consisted of Vox and Conde Nast publication readers aged 18-54. By using metadata inside our partners’ content, we could dynamically serve users with relevant advertisements for a range of products without relying on data signals that users are increasingly finding invasive. The approach demonstrates to publishers and marketers that even as the regulatory and technological landscape changes, powerful, efficient, and contextually relevant digital advertising can be delivered at scale, without personal data. Creative strategy for ads was based on aligning our 25 creative versions in multiple formats to article content data for four Shopping verticals of Home & Garden, Electronics, Apparel & Accessories, and Health & Beauty (See execution).

Working with an established above-the-line creative campaign, shopping verticals and product features, we developed a creative strategy and designed flexible display templates for Google Shopping. The creative strategy for serving the display ads was based on aligning our 25 creative versions to article content data for four Shopping verticals of Home & Garden, Electronics, Apparel & Accessories, and Health & Beauty. This way, readers were served ads based around article content rather than personal data or publication alignment. For example, if a viewer was on Glamour reading an article about electronics, the ad shown would be related to electronics (the article content). Conversely, if they were on Wired reading about fashion, the ad served would align to fashion rather than electronics. The campaign ran during prime holiday shopping time in November and December.

Pegasus proved that contextual advertising can be delivered at scale without relying on privacy-invasive data. Our Pegasus Shopping campaign showed marketers and publishers a way forward in a world where third-party cookies are on the way out, while beginning to build infrastructure for ad-based digital media in a privacy-first world. Not only that, but ads delivered via Pegasus more than doubled our goal of 3% lift in consideration with 6.6% lift.