college tuition

“When I saw that, I thought, I’m already a member of their loyalty club,” she said. “Why don’t they know that?”

Despite the streams of data Facebook has collected about people like Ms. Williamson, the social network needs to know its users much better if it is going to become, as the company hopes, the Web’s most effective advertising platform. And Facebook is scrambling to do just that.

In shaping its targeted advertising strategy, it is no longer relying solely on what Facebook users reveal about themselves. Instead, it is tapping into outside sources of data to learn even more about them — and to sell ads that are more finely targeted to them. Facebook says that this way, marketers will be able to reach the right audience for the right products, and consumers will see advertisements that are, as the company calls it, “relevant” to them.

In late February, Facebook announced partnerships with four companies that collect lucrative behavioral data, from store loyalty card transactions and customer e-mail lists to divorce and Web browsing records.

They include Acxiom, which aggregates data from a variety of sources, including financial services companies, court records and federal government documents; Datalogix, which claims to have a database on the spending habits of more than 100 million Americans in categories like fine jewelry, cough medicine and college tuition; and Epsilon, which also collects transaction data from retailers.

Acxiom and Datalogix are among nine companies that the Federal Trade Commission is investigating to see how they collect and use consumer data.

Facebook’s fourth partner is BlueKai, based in Cupertino, Calif., which creates tracking cookies for brands to monitor customers who visit their Web sites. That data can be used to show an advertisement when those users log on to Facebook.

“Our goal is to improve the relevance of ads people see on Facebook and the efficacy of marketing campaigns,” Gokul Rajaram, product director for ads at Facebook, said in an interview on Friday.

In announcing the partnerships, Facebook said it would allow, for instance, a carmaker to customize an advertisement to users interested in a new car.

The push to refine targeted advertising reflects the company’s need to increase its revenue. Its shares are worth far less than its ambitious initial public offering price of $38 a share last May, and Wall Street wants to see it take concrete steps to prove to advertisers that it can show the right promotions to the right users and turn them into customers.

The partnerships are part of a continuum of efforts by Facebook to hone targeted advertising. Last fall, it invited potential advertisers to provide the e-mail addresses of their customers; Facebook then found those customers among its users and showed them ads on behalf of the brands.

JackThreads, a members-only online men’s retailer, tried this tactic recently. Of the two million customer e-mails it had on file, Facebook found more than two-thirds of them on the social network, aided in part by the fact that JackThreads allows members to sign in using Facebook login credentials. Facebook then showed those customers ads for the items they had once eyed on the JackThreads site.

The nudge seemed to get people to open up their pocketbooks. Sales increased 26 percent at JackThreads, according to AdParlor, an agency that buys the company’s advertisements on Facebook.

Targeted advertising bears important implications for consumers. It could mean seeing advertisements based not just on what they “like” on Facebook, but on what they eat for breakfast, whether they buy khakis or jeans and whether they are more likely to give their wives roses or tulips on their wedding anniversary. It means that even things people don’t reveal on Facebook may be discovered from their online and offline proclivities.

Facebook says that in devising targeted ads, no identifying information about users is shared with advertisers. E-mail addresses and Facebook user names are encrypted and then matched. Users can opt out of seeing specific brand advertisements on their page, and they can opt out of receiving any targeted messages by visiting each third-party data partner’s Web site.

Article source: http://www.nytimes.com/2013/03/26/technology/facebook-expands-targeted-advertising-through-outside-data-sources.html?partner=rss&emc=rss

My column in the Sunday Review section makes the case for going to college and cites two just-released reports, one by two Georgetown University researchers and the other by two Hamilton Project researchers. Each report has some charts worth reproducing.

The opening chart in the Georgetown paper, by Anthony Carnevale and Stephen Rose, estimates the demand for and the supply of four-year college graduates, both past and future.

Center on Education and the Workforce, Georgetown University

The supply of graduates is easy enough to measure; it is simply the number of graduates. To estimate demand, the two economists look at the wage premium for graduates. When the premium is rising, demand is outstripping supply. When the premium is falling, demand is rising more slowly than supply.

Looking ahead, Mr. Carnevale and Mr. Rose use school enrollment to estimate supply. Future demand is trickier to estimate. Mr. Rose, by e-mail, explains:

Over the last 100 years, the growth in demand for college-educated workers has varied within a narrow range and averaged 2.8% increase per year. Given the current high penetration of computer technologies, this paper takes the conservative position that this growth in demand from 2010 to 2025 will grow by 2 percent per year.

The bottom line is that, unless the country begins producing more graduates, supply is unlikely to catch up to demand — and income inequality is unlikely to fall by much if at all.

The next chart, from Hamilton, estimates the annual return from different investments, including college tuition:

The Hamilton Project

I was surprised that two-year colleges had a higher return than four-year colleges, but Adam Looney, one of the authors, notes that tuition at two-year colleges tends to be very low.

And the fact that two-year colleges have a higher return does not mean it makes sense to stop after two years. The next two years of college still have a huge return. You can see this by comparing the total value of the two degrees, known in economic terms as net present value:

The Hamilton Project

The Hamilton researchers — Michael Greenstone and Mr. Looney — also offer a chart on lifetime earnings, which shows that a big part of college’s value is that it brings much larger raises over someone’s career:

The Hamilton Project

Finally, the Georgetown paper points out that the value of college is not merely that it’s necessary for many good jobs, like doctor, teacher, scientist or corporate executive. A college degree also often lifts people’s earnings in occupations that do not require a degree, like construction worker, day-care worker, plumber and secretary:

Center on Education and the Workforce, Georgetown University

My column and the chart that ran with it have more details on this last point.

Article source: http://feeds.nytimes.com/click.phdo?i=4b129f6a28ec4e4edaa5ebdb6f3397fa

Facebook Expands Targeted Advertising Through Outside Data Sources

Economix: Why College Brings a Huge Return

Recent News

Partners