Ad measurement is becoming more challenging. (You know, signal loss and all that.)
But there was never such a thing as perfect measurement, says Jen Faraci, chief data officer at Publicis-owned Digitas North America, on this week’s episode of AdExchanger Talks.
“No matter how sound your methodology is or how comprehensive your variable set or how contained you want your experimental design to be, there’s always noise,” Faraci says.
Even before Apple released its AppTrackingTransparency framework and Google gave third-party cookies an expiration date, there were external factors that marketers couldn’t control for, from word of mouth to human error.
Which is why the most sensible approach to measurement is to have multiple approaches rather than being a religious devotee of one technique over another. (MTA and MMM, unite!)
A multimethodology approach “gives you a mechanism for validating what you see,” Faraci says, and can help marketers explain away any noise that might have crept into their model.
“You need more than one approach to have a high degree of confidence,” Faraci says.
Also in this episode: Measuring attention, tapping into generative AI, using data to inform creative, why first-party data is “the backbone of the modern marketing engine” and tips for using unusual ingredients in recipes, including Japanese Hakurei turnips.
