Working Paper: CEPR ID: DP17956
Authors: Andrey Simonov; Tommaso Valletti; Andre Veiga
Abstract: Does online news content facilitate display advertising effectiveness? We conduct an online experiment in which subjects read various articles and are shown (randomized) ads for brands next to these articles. Using non-intrusive eye-tracking technology, we measure the attention each individual pays to each article and ad. Then, respondents are asked which ads they recall seeing, and choose between cash or vouchers for the brands advertised. We show that articles that capture more of readers’ attention increase the amount of attention readers pay to ads on the page. In turn, more attention to ads increases brand recall and purchase probability. Building on the experimental results, we formulate and estimate a stylized model of attention allocation, purchase and recall. The model features spillovers of attention from articles to ads. The type of news content (“hard” versus “soft” news) does not detectably impact ad effectiveness – evidence against the practice of “block lists” of sensitive news topics by advertisers. We discuss the implications of such attention spillovers for firms’ investments in captivating news content.
Keywords: online advertising; online news; experiments; attention; ecommerce
JEL Codes: M37; C91; L86
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| attention paid to articles (Y90) | attention to ads (M37) |
| attention to ads (M37) | brand recall (M37) |
| attention to ads (M37) | probability of choosing brand's voucher (C25) |
| captivating news content (Y60) | effectiveness of display ads (C91) |
| type of news content (hard vs. soft) (L82) | ad effectiveness (M37) |