lecture 5 Flashcards
What are display ads
A type of paid online advertising, display media consists of text, image, video, audio and motion ads, which display on websites, apps and social media platforms. These ads can appear in several sizes, including banner and reach users across the web
how does an online ad get to appear
someone clicks on the webpage
The publisher of the page puts up the ad inpression for auction
The ad marketplaec holds an auction among the advertisers competing for the impression
The advertiser willing to bid the most for the impression wins the right to display their ad
The ad is delivered to the prospective customer
customer clicks on the ad and the advertiser converts
Difficulty is not always perceived
Measuring the online sales impact of an online ad or a paid-search campaign in which a company pays to have its link appear at the top of search results-is straightforward; we determine who has viewed the ad, then compare online purchases made by those who have and those who have not seen it
Two alternative sequences of ads have been studies
public service ads (replace firms ads with ads for NGO or charity)
The “second best” ad in the auction
Called “ghost ads” in the literature
Public service ads experiment
Randomised control trial (i.e. an experiment)
Users are split into two groups
Control condition:
Users sees public service announcement ads - ads that help raise social awareness
Treatment 8aka test) condition
users see the companys ad as usual
Public service ads experiment main idea
by serving these real, alternative ads, we obtain the information on which users would have been exposed to the companys ad if the experiment did not allocate them to see PSA ads
Test group with firms ad vs control group with unrelated psa ad
Remark:
To show the PSA alternative, you have to win the auction and choose to serve a different ad
PSA ad sequence treatment vs control
Treatment: Macys - starbucks - macys
Control: Red cross - starbucks - red cross
PSA experiments are
costly
you spend 10$ on treatment users and spend another 10 on control users
ghost ads
Identify ads among control group that would have been treatment ads if user was in treatment group
Treatment group
User who see your companys ad (macys in this example)
Control group
User sees whatever ad platform chooses to deliver
Ghost ads implementation
use ghost ad to compare effectiveness of advertising in control and treatment group
ad log with ghost ad and not ghost ad
Ghost ad impression
a log entry that flags when an experimental ad would have been served to a user in the control group
natural experiment
not controlled by experiment designer
