BlueSky Thinking Summary

Florian Zettelmeyer and Brett Gordon, of Kellogg School of Management, identify the real challenge in measuring the effectiveness of digital advertising in today's algorithmically targeted world.

Traditional metrics, such as last-click conversions, they believe, almost never capture the real causal impact because of confounding variables.

This baseline impact is found by running randomized controlled trials, and then these researchers proxy metrics—like last-click data—to more reliably predict ad efficacy.

They estimate that analyzing nearly 1,000 RCTs of Facebook ads established that a small number of RCTs can significantly enhance predictions, thus allowing an advertiser to evaluate ad performance precisely without having to run an RCT for every campaign.

This not only reduces cost but also efficiently works in pre-existing data, hence it's an extremely scaleable approach to improvise marketing ROI in a data-driven landscape.

It is a fully pioneering strategy that's going to change the way businesses measure and optimize advertising investments rather totally.