USA: +1(669)289-8683 
Sign in
Predictive Analytics uses data gathered with and without your consent (whether you realize it or not), and it is as easy as the swipe of a membership card at the register. Participants have volunteered (or been paid) for having their retina’s tracked or even worn brain scan monitors. The data gathered is used to influence, persuade, and manipulate others (who have not volunteered) to purchase products &/or services, or respond to advertising/PR messages.
Be sure to read ALL the articles before participating in this discussion, and then reference them to support your posts. Please respond to the following questions in your post:
Is this too much invasion? Have marketers gone too far? Or is all fair in love and war?
How could you use this information and technology in your business today?
Politics
Online Data Helping Campaigns Customize Ads
By TANZINA VEGA Published: February 20, 2012
Political campaigns, which have borrowed tricks from Madison Avenue for decades, are now
fully engaged on the latest technological frontier in advertising: aiming specific ads at potential
supporters based on where they live, the Web sites they visit and their voting records.
Video: http://video.nytimes.com/video/2012/02/21/us/politics/100000001375023/customizedpolitical-advertising.html
One recent ad from the Mitt Romney campaign was geared toward committed party members,
encouraging them to vote.
A second ad, which focused on Mr. Romney as a family man, was shown to undecided voters.
In recent primaries, two kinds of Republican voters have been seeing two different Mitt Romney
video ads pop up on local and national news Web sites. The first, called “It’s Time to Return
American Optimism,” showed the candidate on the campaign trail explaining how this was an
election “to save the soul of America.” It was aimed at committed party members to encourage a
large turnout. The second video ad, geared toward voters who have not yet aligned themselves
with a candidate, focused more on Mr. Romney as a family man. Versions of the two ads were
seen online in Florida, Iowa, New Hampshire and South Carolina.
Kenneth M. Goldstein, the president of the Campaign and Media Analysis Group at Kantar
Media, part of the advertising giant WPP, said Mr. Romney’s directed ads represented a sea
change in political advertising.
“Forty years ago, you’d watch the same evening news ad as your Democratic neighbor,” Mr.
Goldstein said.
The technology that makes such customized advertising possible is called microtargeting, which
is similar to the techniques nonpolitical advertisers use to serve up, for example, hotel ads online
to people who had shopped for vacations recently.
In the last few years, companies that collect data on how consumers behave both online and off
and what charitable donations they make have combined that vast store of information with voter
registration records.
As a result, microtargeting allows campaigns to put specific messages in front of specific voters
— something that has increased in sophistication with the large buckets of data available to
political consultants.
Zac Moffatt, the digital director for Mr. Romney’s campaign, worked with a company called
Targeted Victory for the online ads. “Two people in the same house could get different
messages,” Mr. Moffatt said. “Not only will the message change, the type of content will
change.”
Few campaigns like to talk about this kind of advertising. Representatives from the Obama
campaign and the Gingrich campaign would not confirm whether they were using targeted ads
tied to voter data. Saul Anuzis, chairman of the Republican National Committee on Technology,
said he expected spending on digital political ads to reach 10 to 15 percent of campaign budgets
in the 2012 election season.
Those numbers pale beside what campaigns will spend on television or direct mail. But the chief
benefit of microtargeting is that campaigns can spend their money more efficiently by finding a
particular audience and paying $5 to $9 per thousand displays of an ad, Mr. Anuzis said.
“We can now literally target the household,” Mr. Anuzis said.
Microtargeting is largely done by a handful of campaign consultant groups including Aristotle,
CampaignGrid and Targeted Victory, which collect some of their data from direct marketing
companies like Acxiom and Experian. The companies are reluctant to discuss which candidates
are their clients, but according to a Federal Election Commission filing, CampaignGrid does
work with the Ron Paul “super PAC,” Endorse Liberty.
The process for targeting a user with political messages takes three steps. The first two are
common to any online marketing: a “cookie,” or digital marker, is dropped on a user’s computer
after the user visits a Web site or makes a purchase, and that profile is matched with offline data
like what charities a person supports, what type of credit card a person has and what type of car
he or she drives. The political consultants then take a third step and match that data with voting
records, including party registration and how often the person has voted in past election cycles,
but not whom that person voted for.
Throughout the process, the targeted consumers are tagged with an alphanumeric code, removing
their names and making the data anonymous. So while the campaigns are not aiming at
consumers by name — only by the code — the effect is the same. Campaigns are able to aim at
specific possible voters across the Web. Instead of buying an ad on, say, The Miami Herald Web
site, a campaign can buy an audience.
Another advantage is that these ads can be bought quickly — using an auction process to obtain
ad space — when campaigns need to move rapidly to aim at an audience, for example, to counter
a bad debate performance or an unflattering newspaper article.
“If you can get in front of a news story, if you can help frame the debate rather than respond to
the debate,” Mr. Anuzis said, “that makes a big difference.”
John Simpson, media director at Blue State Digital, which worked with the Obama campaign in
2008, said bidding technology means strategists can “get a campaign up and running very fast
and also potentially pull it down very fast.”
In 2009, Chris Christie, then a candidate for governor in New Jersey, worked with
CampaignGrid to respond to accusations from Gov. Jon S. Corzine that he supported cutting
health care coverage including mammograms. In response, Mr. Christie’s campaign quickly
created a video ad showing him sitting at a kitchen table with his wife and telling the story of his
mother’s struggle with breast cancer.
It was aimed at female Republican voters who were searching for information on breast cancer.
“It’s awful for the governor to try to desperately hold on to power by scaring people,” Mr.
Christie said at the end of the video.
Mike DuHaime, a partner at Mercury Public Affairs who ran Mr. Christie’s campaign, said of the
ad: “I think the biggest thing in politics is just being able to move quickly. I don’t know if it won
us the campaign, but it kept us from losing.”
When Gov. Bobby Jindal of Louisiana was running for re-election last year, his campaign used a
number of ads with different messages. Blaise Hazelwood, the president of Grassroots Targeting,
the company that worked on Mr. Jindal’s campaign, said voter registration data was critical to
the success of the digital campaign.
“We want to hit the people who can actually go out and vote,” Ms. Hazelwood said. The digital
campaign ran in September and October, and the company placed ads online to reach registered
Republicans as well as registered Democrats. There were more registered Democrats in the state,
and early polling had shown that some were “favorable towards Jindal,” Ms. Hazelwood said.
Critics say that the ability to limit political messages to registered voters toes the line of social
discrimination. Daniel Kreiss, an assistant professor at the School of Journalism and Mass
Communication at the University of North Carolina at Chapel Hill, called some of the targeting
techniques a form of political redlining.
“These practices, as they get more sophisticated, leave entire segments of the population out of
the political communication of the campaign,” Professor Kreiss said, adding that “campaigns
aren’t going to spend resources on people who aren’t seen as being important.”
Prof. Joseph Turow of the Annenberg School for Communication at the University of
Pennsylvania said that ads aimed at registered voters, while efficient for the campaign, benefited
the candidate in another way.
“Different people getting different ideas about a candidate maximize the chances that a person
would agree with you.”
A version of this article appeared in print on February 21, 2012, on page A1 of the New York edition with the headline: Online Data Helping
Campaigns Customize Ads.
A New Crystal Ball to Predict Consumer and Investor
Behavior
Agent-based modeling provides insights into the complex behavior of large
groups of people, like consumers, voters and investors
Agent-based modeling turbocharged by cloud computing has a growing role in
finance and regulation. Photo: David Plunkert
By Paul J. Davies June 10, 2019 10:07 pm ET
E.W. Scripps Co. had a big year in 2018 for political advertising on its network
of local TV stations. Its new approach to political-ad sales helped to nearly
double its revenue in the sector versus the previous midterm elections in 2014.
Behind these gains was a form of market analysis called agent-based modeling, which can show how all sorts of
different people or companies or investors interact with each other and might react to things that happen to
them. It helped E.W. Scripps advise political-campaign managers on how to get the best bang for their buck,
and it’s helping a growing number of other companies understand their markets.
“We aim to simulate how consumers make decisions,” says Dejan Duzevik, chief product officer at Concentric
Inc., a Cambridge, Mass.-based modeling company that helped E.W. Scripps. “Lots of companies are doing this
to see if they can bring new consumers into their market.” Concentric’s clients have included Microsoft Corp. ,
Toyota Motor Corp. and Whirlpool Corp. , as well as advertising companies like Havas.
Agent-based modeling also has a growing role in finance and regulation, particularly as a way of understanding
how crises and panics cascade through markets—things that are missed by traditional, simplified economic
models. Traditional economics and market analysis average out the differences among individuals to arrive at a
representative of the group, who is typically expected to act rationally. This creates elegant models that can
produce simple answers for researchers—and simple predictions of what people will do. However, like a map
drawn crudely, traditional models are useful for some purposes but fail to match reality at a finer resolution. In
contrast, agent-based models can represent many different kinds of individuals with a wide variety of desires
and needs at the same time, and offer a wide range of possibilities of how they will act, which is much more
realistic. But with realism comes complexity and predictions that are often less specific.
Crisis testing
Financial regulators are running these sorts of models to see how crises spread—how quickly and how far a
shock in one market or asset class can spread to other parts of the financial system. They try to identify weak
points in the financial system—for example, in banks, clearing houses or money-market funds—and then test
regulations or policies that could bolster those weak points and make the financial system more resilient.
Richard Bookstaber, a financial-risk manager, academic and author who built such a model for the U.S.
Treasury after the 2008 global financial crisis, says agent-based modeling works well for financial crises
because at those times investors and traders become more risk averse, which compels them to follow rules of
behavior that are easier to discern. Scientists call these rules of behavior heuristics. In normal times, investors
and traders often change behavior or disguise preferences when that could help them make money.
“You can’t use agent-based modeling if you don’t know the agents or their heuristics,” Mr. Bookstaber says.
And that usually makes modeling finance harder. “Heuristics and preferences are more stable in the consumer
world,” he says. “You don’t profit from hiding, lying or changing your preferences in unexpected ways,
whereas you do in finance.”
These models aren’t new, but in the past their use was limited by the extremely high number of calculations
they perform to re-create the complex behavior of large groups of people. The computing power needed to
quickly perform those calculations has only become accessible inexpensively in recent years through cloud
computing. The growing amount of data collected also has helped make agent-based modeling a much more
commercially viable tool.
All this is possible because enormous computing power has become widely available with the growth of cloud
computing in recent years. Andrew Skates, founder of Sandtable, a London-based modeling company, says that
when he started in 2009, he would run models on a series of laptops linked together. “Over a weekend, we
would get 50 hours of computer time on 10 cores [processors], which is 500 computer hours,” he says. In a
recent week, “we did a run that was 50,000 computer hours, and that’s not even large for us.”
Political choices
Michael O’Brien, vice president of distribution at E.W. Scripps, says agent-based modeling allows the company
to test the impact on voters of events like news stories, political rallies, security scares or even the weather. The
model is built on multiple sets of data such as audience ratings for TV shows, social-media volumes, public and
private polling and other information to build a useful picture of attitudes among the voters in a particular race.
Mr. O’Brien says the agent-based model produces much better predictions than traditional polling alone about
the most effective ways to encourage voters to back a candidate. “Campaign managers tell us who they want to
target, and we can say where are the best opportunities to target those voters and how to get the best conversion
rates,” he says.
Beyond politics, agent-based models are being more widely used by advertising and consumer-products
companies to generate predictions about what certain advertising messages or pricing strategies will push a
company’s rivals to do, as well as how they will affect opinions and choices among consumers. Another major
role is to test what might happen in the event of some kind of shock to a company’s market. “Almost every
customer has some scenario that checks out the impact of an exogenous shock,” says Concentric’s Mr. Duzevik.
“We have a joke about it: Usually, it’s Amazon.”
One of the strengths of this modeling is that it produces a range of probabilities rather than a simple prescription
for how to proceed. David Teague, portfolio head for data analysis and products at British Broadcasting Corp.,
says agent-based models don’t exactly forecast the future, but they do let you explore what could happen in a
complex system that keeps changing. The BBC, Britain’s taxpayer-funded broadcaster, is developing a model to
test how different mixes of TV programs will appeal to different segments of the audience. Because the
company is publicly funded, it has public-service duties—and has to ensure taxpayers’ money is spent to the
widest possible benefit.
Like all models, Mr. Teague says, agent-based ones can only ever be something that helps inform decision
making rather than making decisions for you. “It’s a balance of ‘This is going to help me’ versus ‘This is going
to answer everything for me,’ ” he says. “With anything, once you think it will answer everything for you, that’s
when you come unstuck.”
Mr. Davies is a Wall Street Journal reporter in London. Email him at paul.davies@wsj.com. • Copyright ©2019 Dow
Jones & Company, Inc. All Rights Reserved.
MEDIA & MARKETING August 4, 2013, 7:31 p.m. ET
Old-School Ad Execs Sweat as Data Geeks Flex Muscle
Madison Avenue Increasingly Values Digital Savvy Over Conventional
Creative Talent
By SUZANNE VRANICA And CHRISTOPHER S. STEWART
When former ad buyer Samantha Giangrande began looking to return to work recently after a two-year
child-rearing break, she noticed an unsettling fact: the advertising world she knew had changed.
“I’m 37 years old and extinct,” she said, lamenting that all the agencies she talked to about a job were
obsessed with something she didn’t possess—digital experience.
Once home to creative types in the mold of Don Draper and “Mad Men,” Madison Avenue is
increasingly a bastion of geeks: computer programmers, data heads and quantitative analysts.
People “with traditional backgrounds aren’t getting hired,” said Amy Hoover, president of Talent Zoo,
an Atlanta-based company that operates an online job site specializing in the ad business.
That shift is causing a bit of existential anxiety. “If there’s one word that applies to this industry, it’s
fear,” said Bill Crandall, a consultant to Della Femina Advertising, where he worked as the chief
marketing officer before his position was eliminated in 2011. “I never fathomed this going into the
industry.”
AMC/Associated Press
Don Draper of ‘Mad Men’ never had to worry about digital upheaval.
He has been trying to land a full-time advertising job and has watched colleagues in the industry
struggle to make a comeback. “It’s like trying to get a seat on a plane that is booked solid,” he said. “The
people who are out of work are on standby.”
Last month’s merger agreement between New York-based Omnicom Group Inc. and Paris-based
Publicis Groupe SA only intensified the nervousness. The two ad giants, whose combined stock-market
value totaled $35.1 billion, touted their marriage as driven in part by the rise of big data and digital
advertising. Some employees are fretting about possible job cuts, though the two companies have said
their deal wasn’t about that.
“This wasn’t done just so we could go in and cut jobs,” said John Wren, Omnicom’s chief executive, in
an interview last week
Even so, with all the talk about the importance of data to Omnicom and Publicis, the “creative people
can’t help but feel left out,” said Mike Sheehan, chairman of Hill Holliday, a Boston-based ad agency
owned by Interpublic Group of Cos.
Mr. Sheehan, whose parent company is itself now the subject of takeover speculation, dislikes the new
emphasis.
“Algorithms cannot give people goose bumps, and algorithms cannot tell a story,” said Mr. Sheehan.
They “cannot give you that nugget of insight that differentiates brands and products,” he added, “That is
only performed by human beings that can tell stories and can understand consumer behavior
instinctively.”
The trend toward digital advertising has been under way for about a decade. This year digital ad
spending in the U.S. is expected to grow nearly 14% to $41.9 billion, or about 25% of total ad spending,
according to eMarketer.
Traditional ad agencies eventually embraced digital media, initially the province of specialist firms,
often by acquiring those firms, hiring digital-savvy executives or training staffers in new techniques. As
the new executives moved in, some people thought to be too invested in the old ways were dismissed.
Times got tougher during the recession. U.S. ad spending fell just over 12% in 2009. Omnicom cut its
workforce to 63,000 by the end of 2009 from 70,000 in 2007, according to the company’s annual report.
Interpublic cut 5,000 jobs between 2008 and 2009, ending that year with 40,000 employees, company
filings show.
While both companies have added back staffers since then—Omnicom is now bigger than it was in
2007—the growth has largely focused on digital advertising, executives have said.
For several years, Publicis USA, an ad agency owned by Publicis, has had a policy of requiring new
hires to have some digital experience, according to people familiar with the matter.
Early on, the creative side of the business was affected, as executives had to learn how to write search
ads, create interactive ads, and craft campaigns that would resonate on social-media sites. At the same
time, media buyers, the executives who buy ad time and space on behalf of marketers, had to figure out
where to find consumers, who were spending more time online.
More recently the media-buying part of the business has undergone another wave of change, thanks to
the digital revolution. More digital ads are bought through automated exchanges, a process known as
“programmatic buying.”
It is a far cry from the days when media schedules were faxed back and forth and relationships between
ad-sales executives and media owners were cemented by long nights of schmoozing over drinks.
So far, automated real-time bidding accounts for a relatively small part of total media spending. But it is
on the rise. According to eMarketer, U.S. advertisers will spend $3.4 billion on real-time bidding for
online display ads in 2013, up 73% from last year.
Although the rise of automated trading has meant adding new types of executives it also will likely have
a downside: job cuts for conventional media buyers.
“If you are on the analytics and digital side of things that is where the bulk of the jobs are,” said Talent
Zoo’s Ms. Hoover. Five years ago traditional media buying and planning job placements accounted for
about 25% of Talent Zoo’s business. Now it is about 5%, she said.
While some ad executives are nervous, for others the massive changes haven’t “sunk in yet,” said Joe
Grimaldi, chief executive officer of Mullen, an ad agency owned by Interpublic. “Looking ahead a few
years, a lot of those roles will not exist,” he added. Someday, he said, automation some of the business
of buying TV ads will be automated.
“Everyone worries about their jobs,” said Paul S. Gumbinner, founder and president of Gumbinner Co.,
a recruiter whose clients include Publicis Kaplan Thaler and Omnicom’s BBDO. “There’s no loyalty.”
Mr. Gumbinner said that the industry’s consolidation has unsettled its ranks. “If you’re a Don Draperlevel person, you’re probably protected by contracts. But for 90% of employees, the smart ones
understand they need a Plan B. It’s a really tough business.”
Even those with jobs in digital advertising are uneasy. “Everyone is overworked and stressed,” said Jon
Bond, co-founder of an ad firm now called Kirshenbaum Bond Senecal + Partners, who currently runs a
digital firm Tomorro LLC, a holding company of digital marketing and media assets.
In the past, Mr. Bond said, an agency on average might have produced three print ads and three
commercials a year for a typical client. The Web, with its constant stream of new content, has changed
the demand. “Now you have to do the equivalent of that every week,” he said.
“Look at ‘Mad Men’ when they could go out drinking and come back three hours later and come up with
an idea,” Mr. Bond said. “Now they go to Starbucks and get caffeine. The double latte is the new double
martini.”
—Ruth Bender contributed to this article. Write to Suzanne Vranica at suzanne.vranica@wsj.com and Christopher S.
Stewart at christopher.stewart@wsj.com A version of this article appeared August 5, 2013, on page B1 in the U.S. edition of
The Wall Street Journal, with the headline: Old-School Ad Execs Sweat As Data Geeks Flex Muscle.
Copyright 2012 Dow Jones & Company, Inc. All Rights Reserved
