What’s a cookie worth anyway?

Lots has been written about the impending loss of cookies and how to thrive in a cookie-less future. But how important are cookies to the revenue of publishers? That’s the question addressed by Rene Laub, Klaus Miller and Bernd Skiera (Goethe University Frankfurt) in their presentation ‘The Economic Value of User-Tracking for Publishers’ on June 9^th at the Theory + Practice in Marketing Conference

First, the main findings: publishers only obtain 61.5% of the price of an ad with user-tracking. Interestingly, while 93% of publishers suffer, premium publishers suffer more. Moreover, it is the non-browsing data about the users that create most value for publishers.
 
How did the researchers arrive at this conclusion? They want to estimate the causal effect of taking cookies away but can’t do a controlled experiment. So you aim to find to best counterfactual, i.e., what would have happened if the cookies were not taken away. The difference is the Average Treatment Effect (ATE):

Technically, the researchers use Augmented Inverse Probability Weighting (AIWP), which is double robust (see link for an introduction). Based on the calculation, the average treatment effect is –0.25 Euro, or -38.5% loss for the publisher when cookies are taken away.

But do all publishers experience the same loss? No, regressing the treatment effect on publisher characteristics reveals that premium publishers lose more, while niche publishers lose less. Moreover, higher revenue concentration (only a few advertisers) is associated with lower cookie loss.

So who should be afraid of cookie loss? You are more vulnerable if your revenue is spread out over multiple advertisers, and if you are a premium publisher. 93% of publishers can expect revenue loss though, and should consider strategies to soften the blow.

Advertisement