week 4 in-class activity
Cooper2021
4-2.pptxBusiness Ethics Summer 2022 (1) Week 4, Lecture 2
Chaeyoung Paek
In today’s class…
We’ll look at what Cathy O’Neil says about online advertising; her primary example is online ads for for-profit universities.
There will be no in-class activity for today’s lecture.
What is so wrong with the for-profit university advertising?
“For-profit university”, Vox video: https://youtu.be/wBEZnvU2mz8
ABC news on for-profit university predatory marketing strategy: https://youtu.be/vmmYKlPx3Sw
Frontline on “get to their pain” marketing strategy: https://youtu.be/PClS3W5HzKw
Online Advertising
Arrington’s defense of advertising techniques applies to the “traditional” means of advertising: TV commercials, posters, etc..
In “Propaganda Machine,” Cathy O’Neil points out that online advertising, in its attempt to be as effective as possible, manipulate the audience in a unique way.
All advertisements make use of some information about the audience; what is unique is how online advertising can be personalized for each potential customer & can do so in a great scale.
Targeted Online Advertising
Targeted advertising: a form of online advertising that focuses on specific traits, interests, or preferences of each customer, based on the data on one’s online activities.
(ex) Online ad banners / Facebook ads
Based on her experience as a data scientist, Cathy O’Neil argues that targeted advertising is predatory practice.
Her argument is based on the amount of information targeted online advertising uses & how it uses that information.
Case Study: For-profit Universities
O’Neil’s primary example of targeted online advertising is online ads for for-profit universities.
Here’s how their advertising strategy works:
The marketing team at a for-profit university pays platforms such as Facebook or Google to run its ads to a very specific group of people.
Online platforms gather a lot of personal information just by tracking each user’s online activities; their financial situation, their behavior patterns, their medical/relationship history, etc..
For-profit universities look for people who meet specific conditions: people who live in the poorest zip codes, people who have clicked on ads for payday loans, etc; and online platforms allow them to micro-target the specific audience.
Case Study: For-profit Universities
2) Then they run endless series of competing ads against each other to figure out which one is the most effective.
- This “A/B testing” method is used by other forms of advertisement.
But online advertising makes using this method in a giant scale easier; it also gets feedback much faster than traditional forms of advertisement.
With machine learning, the computer can “learn” the behavior patterns of numerous people based on simple instructions.
Case Study: For-profit Universities
2*) Sometimes for-profit universities use lead generation to get a list of potential students & directly reach out to them.
Lead generation is a method to create a list of potential customers, which could be sold to different firms.
(ex) The ad asking you to provide your personal info for new policy about financial aid for moms
For-profit universities also use lead generation themselves too.
(ex) The College Board is engineered to direct poor students toward for-profit universities
What makes targeted online advertising harmful?
O’Neil argues that the contents of targeted online ads are not that different from the traditional ones; they promise some imaginary benefits to the audience or only provide indirect information in their repetitive ads.
So, the problem of the targeted online ads is not that they often use puffery/indirect information transfer/subliminal advertising technique.
It seems that the problem is that they can reach so many people so easily and so effectively.
What makes targeted online advertising harmful?
How can targeted online ads reach so many people so easily and so effectively?
They use more accurate models of behavior patterns, developed by data scientists and the machine learning process.
Because targeted online ads are better at expecting who would respond to them & responding to them is easier for potential customers, they get better response rates.
(ex1) Newspaper ads for University of Phoenix:
- reach 100,000 people, 1% responds = 1,000 students
(ex2) Targeted online ads for University of Phoenix:
- reach 1,000,000 people, 2% responds = 20,000 students
(O’Neil) Targeted online ads could end up harming so many people, when the advertised product is harmful.
Conclusion
Cathy O’Neil concludes that targeted online advertising (a) is predatory in many cases and (b) has the potential of harming so many people’s lives when the advertised product turns out to be harmful.
She believes that this is a unique problem of online advertising; the traditional advertising could never use the vast amount of data and could never reach so many people.
A Question re: O’Neil
O’Neil does not explicitly argue for this, but it seems that she believes that many targeted online ads are deceptive or manipulative in a unique way.
In order to argue so, it must be shown that targeted online ads undermine one’s autonomy in a way that other forms of advertising do not; then they would be harmful because they are targeted online ads.
This is what Daniel Susser, Beate Roessler, and Helen Nissenbaum do in their article, “Technology, autonomy, and manipulation.”
For the next class…
Read Daniel Susser, Beate Roessler, and Helen Nissenbaum, “Technology, autonomy, and manipulation.”
