Media, ad blockers, and COVID-19

How to cite

Veiga, Andre; Valletti, Tommaso. 2020. "Media, ad blockers, and COVID-19." AEA RCT Registry. AEARCTR-0006010. https://doi.org/10.1257/rct.6010-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Andre Veiga* , Tommaso Valletti

* Primary investigator

Registry ID
AEARCTR-0006010
DOI
https://doi.org/10.1257/rct.6010-1.1
First registered
Last updated
Status
Completed
Trial dates
2020-07-24 – 2020-08-06
Countries
United Kingdom of Great Britain and Northern Ireland ()
JEL codes
M37, L86
Registry record
http://www.socialscienceregistry.org/trials/6010

Abstract

The COVID-19 outbreak has resulted in an unprecedented media consumption in households, yet this increase in not reflected in additional revenues for newspapers. On the contrary, digital advertising is reported to have decreased. A large problem is related to advertisers using “ad blockers” that effectively prevent their advertisements from be shown next to COVID-related stories. This project will investigate if there are ways to arrest this decline during the crisis. We will conduct an eye-tracking study of a 1000 UK and US participants. The project’s findings would constitute experimental causal evidence regarding whether ads shown next to COVID-related articles harm the advertisers. These findings will help us understand the impact (or lack thereof) of coronavirus on attention to advertising in newsbrands. This may have a dramatic effect on the revenues of one of the communities.

Experimental design

8 x cells of 125 online representative respondents (a desktop and a mobile cell in each country, each split into a pair of sub cells, n=1000) will be asked to read 12 pages of newsbrand content while having their eyes tracked using a webcam eye tracking software.

The pages will be selected by the researchers, and contain a mixture of ‘hard’ and ‘soft’ news stories, including both COVID- and non-COVID related stories. The sites and pages will be chosen with reference to the sites' daily visitor numbers. The same content and content/advertising combinations will be shown across both mobile and desktop cells. The pages will be served in a random order to remove order effects.

There will be three ad slots on each page
• Desktop 1 x static billboard and 2 static DMPUs (Double Mid Page Units)
• Mobile – 3 MPUs (Mid Page Units)

The number, format and location of ads on each will be held constant to remove the impact of differences in media inventory. All the formats chosen are regularly used in the digital advertising industry and are well understood by publishers and advertisers.

Respondents will be invited to read the pages as they would normally, and then complete a short mental arithmetic exercise, designed to wipe their short term memory and so reduce recency bias

Having done this, they will be asked to answer a questionnaire to assess the primary outcome variables (mentioned above).

A demographic and geographic location questionnaire will have been administered before the test commenced.

Respondents would be fully aware that they are taking part in an eye tracking study and paid for their time. However, we would not reveal to them the objectives of the study before asking them to read the pages and complete the questionnaire.

Design details

Randomization method
Randomization done in office by a computer.
Randomization unit
Individual
Clusters
8 cells of 125 online representative respondents
Observations
1000 respondents
Arms
8 cells of 125 online representative respondents (a desktop and a mobile cell in UK and US , each split into a pair of sub cells, n=1000).
Primary outcomes
• recall of the ads (spontaneous and prompted ad awareness) • attitudes to advertising that appears in newsbrands and/or next to Covid content • willingness to pay/intent to purchase

Cited by

No citation data available for this registration yet.

Keywords