Teixeira, Picard, and el Kaliouby (2014) ask how much entertainment a TV ad should carry, and whether that entertainment should come before or after the brand.
Across 82 ads shown to 178 people in homes and workplaces, entertainment has an inverted-U relationship with purchase intent. Entertainment that comes after the brand is positively associated with purchase intent. Entertainment that comes before the brand is not.
Method
A web-based face tracker recorded smile responses, viewing interest, and purchase intent while people watched ads outside a lab. A Bayesian hierarchical model treats viewing interest as endogenous when estimating the purchase effect.
Main findings
- More entertainment is not always better. Purchase intent rises and then falls as entertainment increases.
- Brand before the bit. Entertainment placed after the brand is positively associated with purchase intent.
- Front-loaded entertainment does not pay. Entertainment that arrives before the brand is not positively associated with purchase intent.
Why this is in the Darwin index
This is the dosage paper next to Teixeira, Wedel, and Pieters (2012) on joy and surprise. It is the published argument for putting the brand in before the entertainment peak.
Source of record: doi:10.1287/mksc.2014.0854.