One Additional Minute of Exposure to Display Advertising Can Boost Direct Traffic to a Company’s Website by 10%, According to New Research from NYU Stern
A new digital attribution model shows that advertising effects are amplified up to four times when consumers are targeted earlier and for longer periods of time
This new framework gives advertisers the ability to causally measure the impact of digital advertising on the search and purchase behavior of consumers, which is a key component in a broader marketing campaign.
This first-of-its-kind study uses real-world settings and data to examine how actual viewability of display advertising – and the length of exposure to display ads – impacts consumer behavior from initial exposure to final purchase. This framework is part of a pioneering set of research focused on digital attribution, or determining which display advertising channel deserves the most credit for a consumer’s purchase – a key question among advertising executives who manage campaigns for brands.
Using a large-scale, individual-level data set, which tracked the entire advertising campaign of a U.S.-based retailer, the study reveals:
- On average, 55% of display ads are not viewable – meaning they are not visible on a consumer’s screen area for more than one second.
- Exposure to display ads increases the chances that a consumer will engage in active search and visit a company’s website by more than 36%.
- Additionally, consumers are 25% more likely to click on a search ad served up to them on Google.
- When a consumer is targeted earlier in the purchase funnel, one additional minute of exposure to display advertising can increase the likelihood he/she will visit a company website by almost 10%.
- Consumers are also 7% more likely to complete a purchase after seeing a display ad.
The article, “Towards a Digital Attribution Model: Measuring the Impact of Display Advertising on Online Consumer Behavior,” is forthcoming in MIS Quarterly and was nominated for a Best Paper Award at the 2015 INFORMS Conference on Information Systems and Technology (CIST).