>> perspectives

Measuring Outcomes and Audiences in a (Mostly) Cookieless Mobile World

Yieldmo Dragonfly™ enables more accurate campaign measurement and efficiency by powering our ability to attribute conversions (via a Yieldmo pixel placed on check-out pages, product pages, etc.) to Yieldmo ad exposures and our ability to analyze which audience segments drove the largest share of performance. Together, this information drives stronger campaign performance by making optimizations possible at multiple levels—audience, ad format, ad creative, specific ad placement, etc.

This technology is powered by a combination of mobile device identification techniques. For in-app advertising, it uses Apple’s Identifier for Advertisers (IDFA) and Android’s Advertiser ID. The process for mobile web advertising is more complicated. In the mobile web, Dragonfly utilizes third-party cookies when possible, in combination with the user’s mobile web browser cache.

Both of these techniques are useful, but not sufficient for many use cases, especially given Apple’s iOS 11 update which limits the longevity of cookies in the Safari browser. To overcome this challenge, Dragonfly additionally relies on the unique attributes of each HTTP request that accompany every impression. Examples of such data points include the user’s IP address, browser type, language settings, and much more. None of these elements are unique on their own, but the aggregation of everything taken together creates a unique device identifier. While cookies are 100% accurate, our testing reveals that Dragonfly’s unique device identifier is 94% accurate without relying on cookie data. This high level of accuracy in a cookie-less environment is critical for driving campaign effectiveness.

Posted by: Teddy Jawde, Head of Product