Society | DOI: 10.1145/2240236.2240243
Samuel Greengard
Advertising Gets Personal
Online behavioral advertising and sophisticated data aggregation
have changed the face of advertising and put privacy in the crosshairs.
Collusion, a Firefox add-on, lets a person see all the third-party entities tracking his or her movements across the Web.
NOt lOnG aGO, a man walked into the local Target store in Minneapolis and demanded to speak to the manager. He wanted to know why his
then-high school daughter was receiving coupons and promotions for maternity clothing, cribs, and other items
that would indicate she was pregnant.
“Are you trying to encourage her to get
pregnant?” he asked. The store manager examined the stack of coupons
and promptly apologized. He said he
did not have any idea why the girl had
received the coupons. The man then
left for home.
This could have been the end of
the story. But after talking to his
daughter the man discovered she was
pregnant. Target had used sophisticated predictive analytics to determine that her previous buying patterns and behavior had indicated a
high probability of expecting a baby.
In fact, Target and other stores have
become so good at gauging customers’ buying patterns they now disguise customer-specific promotions
by including coupons that are completely irrelevant to the recipient.
Welcome to the new world of adver-
tising. As statisticians, software devel-
opers, and advertising experts mine
and mix growing volumes of online
and offline data and develop increas-
ingly complex algorithms, they are
building new and remarkably sophis-
ticated advertising models designed
to maximize results—and revenues.
“Technology is enabling new—and in
some cases hyper-local and person-
alized—forms of advertising,” states
John Nicholson, counsel for the law
firm Pillsbury Winthrop Shaw Pitt-
man. However, “there’s a fine line be-
tween what’s acceptable and what con-
stitutes an invasion of privacy.”
“Online behavioral advertising
methods are advancing at an incredi-
bly fast pace,” says Lorrie Faith Cranor,
an associate professor of computer sci-
ence, engineering, and public policy
at Carnegie Mellon University. “There
are clearly advantages to receiving rel-
evant ads, but the Internet, combined
with today’s data-collection technol-
ogy, poses serious privacy concerns.
Unfortunately, most consumers feel
as though they have little control over
what happens to their data and how it
is used by advertisers.”
By the numbers
Over the years, advertisers have struggled to better understand the whims of
the marketplace and target consumers
more effectively. Identifying market
niches and customer segments has
been a daunting task and there has
been no easy way to deliver relevant
ads. The result? Most ads target broad
demographic segments through television, radio, newspapers, magazines,
kiosks, billboards, shopping carts, and
other media. In many cases, advertisers simply hope for positive results and
learn by trial and error.
However, the last few years have
brought about a revolution in data
mining particularly as online and conventional methods have allowed advertisers to assemble and reassemble
data in new ways. A growing number
of retailers, including Target, assign
each customer a unique ID number
or guest code. It is associated with
a credit or debit card, and an individual’s purchase history is stored for
analysis. This is separate from a loyalty program, and the only way to avoid
tracking is to pay with cash and avoid
giving a phone number or any other
personal data. But the process does not
stop there. Increasingly, retailers and
others plug in information from third-party sources that track the same individual. This might include the person’s
Web browsing patterns, credit history,
what magazines they read, and even
conversations they have had at social
media sites.
The result is a fairly comprehensive
picture of an individual’s buying hab-