The Man Who Took on Nielsen

PrecisionDemand's Jon Mandel

It’s about 9 p.m. New York time, and Jon Mandel is in Seattle, on the phone, telling a story about the story the Nielsen ratings weren’t telling.

“It was Danny Abraham,” he says, speaking about billionaire Slim-Fast magnate S. Daniel Abraham. “I was presenting a buy—I’ll never forget this—it was a daytime network buy,” Mandel says, rapidly shifting into high gear. “It was ABC, CBS and NBC. What are they trying to do? They’re trying to reach young women. NBC was more expensive, but it made sense because it was women 18-49. But Danny always looked at the unit cost.”

Mandel pauses for breath. “I wish you could type in one of those New York Jewish accents: ‘There you go, confusing me with that ratings shit,’” Abraham said to Mandel. “‘If the cash register rings, we’ll do more. If it doesn’t ring, we’ll do something else.’ That’s what the business is about, and we’ve cluttered it up with GRPs and all these proxies.”

Mandel, who spent his career as a media buyer before joining forces with Nielsen for a time, is now the CEO of PrecisionDemand, an analytics company that breaks down several different data sets into recommendations on which network to buy ad space, how much of it to buy and for which time slot—all with the goal of making, as Abraham put it, the cash registers ring.

The process, for those unfamiliar with the purchase of TV advertising, goes something like this: Advertisers buy airtime from networks in the form of gross ratings points (GRPs) measured by Nielsen, which break down viewership by age, sex and other broad segments. “Here’s X-million dollars,” you tell the network. “Run my ad on your new show every week at 7 p.m. until it’s been seen by at least 4 percent of women 18-49 watching TV at that hour, since that’s the consumer set market researchers tell me buys my product.”

As with other companies (including TRA, Simulmedia and Invidi), PrecisionDemand gives advertisers a data set to use as an overlay, telling them which of those GRPs is more valuable to the client’s specific needs—because there are enormous differences.

Women watching Bravo, for example, have far more money than women watching soap operas, though they fit into the same Nielsen demographic. If a retailer doesn’t have stores on the West Coast but half of the network’s viewers are in Los Angeles, then clearly half the advertiser’s money is wasted. Without analysis to tell them which networks, programs and time slots are most likely to reach and appeal to a client, then marketers are simply buying points for “women 18-49,” measured with a yardstick Mandel maintains is only “40 to 60 percent effective.”

Industry grousing about Nielsen ratings has gotten louder in recent years as TV viewership has become fragmented and smaller demographic measurements have become more valuable. One Nielsen executive, who asked not to be identified for this story, extols the virtues of the company’s ratings panel, which tracks the behavior of 22,000 households and projects statistics down to the thousands for the 111 million homes with televisions in the U.S.

“The mistake, I think, that’s made is that consistent is accurate,” says the exec, who admits that “as the ratings get more and more dispersed, we need a larger sample size,” adding that Nielsen has several initiatives planned to increase GRP accuracy.

IF YOU CAN'T BEAT ‘EM…

Mandel is familiar with the selling points for Nielsen’s GRPs—after all, he used to sell them.

Long before PrecisionDemand, Mandel made his living off the Nielsen ratings. A longtime national TV media buyer for Grey Advertising, he ultimately became CEO shortly after Grey funneled its media buying and planning business into MediaCom. Over that 30-year span, Mandel acted not merely as the middleman preaching the virtues of Nielsen ratings to clients but as head middleman.

Mandel ultimately grew disenchanted with the inadequacies of TV measurement and in 2006 joined Nielsen, where, he says, he hoped he could effect change at the research giant (Nielsen gathers all manner of data beyond TV ratings; recent reports say it is now incorporating consumer banking data into its mix). There, he headed an internal division, NielsenConnect, designed to generate new and deeper research by fusing data from disparate parts of the company. “They were really just a bunch of fiefdoms that didn’t work together,” Mandel relates. “What they wanted to do was put together the data … I thought they wanted me to save Nielsen, and it gave me a chance to tie the purchase data into the ratings data, to tie it all to the cash register.”

What he found, he says, was a mess. Mandel claims to have identified 14 different healthcare subsidiaries within Nielsen, all operating independently of one another. (A source close to the situation says that number is improbable.) Added to that, the methodology bothered him.

“I’m talking to the consumer side,” he says, “and they’re insisting to me that the use of menopause products among boys 6-11 are spiking, and I said, ‘That’s not right.’ And they said, ‘No, look at the data.’” (A source could not confirm or deny the account but says Mandel may be conflating two separate incidents). Ultimately, the time came to present a report featuring findings about menopausal grade schoolers to one of the company’s higher-ups. Rather than fix the data, Mandel recalls, somebody said, “Just take it off the chart. Don’t show him that.”

“They were doing things garbage-in, garbage-out,” he says. “If it came out of the computer, it was great! I couldn’t live with that. And I realized, wait, I’m here to save Nielsen and God can’t save this place.”

Dave Thomas, a self-described Mandel fan who worked with him at Nielsen, argues that the company’s problems are complicated by clients (read: the networks) that are leery of change. “Everybody wants a bigger sample size, but nobody wants to pay more,” Thomas says. “Everybody wants to go to heaven, but nobody wants to die.”

A BETTER BRAND OF DATA?

Mandel’s beef with Nielsen, one industry vet points out, is something of a self-serving pitch—the old I-was-blind-but-now-I-see gambit. Mandel, onetime champion of Nielsen ratings and a veteran of the advertising world, is here to tell you that he used to sell you snake oil but has discovered something better. In person and over the phone, Mandel does not come off as gleeful or disingenuous, though; he sounds tired, and even a little angry.

Is this all part of a new con? Some may think so. But there’s also the matter of his new product actually working. Adweek called as many PrecisionDemand clients as we could find (many of them in the direct response business), and they generally believe this is the way of the future. “They’re changing the landscape of media buying,” says Jessica Abramson, senior director of consumer marketing at streaming music service Rhapsody. “I wouldn’t doubt that in several years, the entire industry is revamped.”

Abramson says her company was involved with PrecisionDemand back when it was Proceed Media—just a few rocket scientists (literally—we’ll get to that in a moment) crunching some data—and was very happy with it. “It’s a legitimate and trackable channel for us to grow our business,” she explains. “No one really truly knows the effect that TV has, whether it’s moving product or digital subscriptions.”

Indeed, even the prospect of analytics-driven media buying is a great relief to many of Mandel’s clients. “Saying that people who look alike act alike—it’s borderline offensive, right? You people all act alike!” cracks Sean Baenen, CMO of ClearChoice, a company that makes and sells pricey dental implants. Baenen says he’s impressed chiefly because he sees a lot of analytics firms claiming to be tech companies but having the same culture as Madison Avenue when what he really wants is Silicon Valley.

It’s important, he says, for new thinking to come from outside, because too much of the ad business has grown complacent. He points to the music world as an example. “The record industry had to be changed by the tech business,” he points out. “It was never going to be changed by the record industry.”


But what’s so special about the numbers being crunched at this one company? To start, where other companies mine data and then apply it, PrecisionDemand is agnostic. It uses material from everywhere. It has a partnership with data-gathering giant Acxiom and uses information purchased from Experian, Personix and Rentrak, set-top box stats bought from several smaller cable companies, vehicle registration information, loyalty card data—pretty much every source of purchasable information its researchers can get their hands on. How those data sets are weighted depends on which of them best identifies the customer base for a given client. The company also handles media buying for its clients, obviating the need for the industry’s endorsement of its methodology.

Sifting through that data is punishing, but Mandel’s team (“my nerds,” he calls them) is holed up in Seattle, spreadsheeting away.

“I [had] a guy who works for NASA—that’s one I always work into a new business pitch,” he says. The NASA guy’s commitment to precision stands in contrast to how derelict in their duties TV researchers have become, Mandel says. “You Snoopy-dance if you’re within 10 percent in this business,” he gripes. “If my guy was off by one-tenth of 1 percent [when he was working on the Mars Rover project], he’d have blown right by Mars.”

WHAT BROUGHT HIM BACK

Interestingly, Mandel says he didn’t even want to join PrecisionDemand. He had quit Nielsen in a rage. (Here’s the quote he gave the press when he left: “I don’t want to say anything negative about the Nielsen Company, but I never understood the humor in Dilbert cartoons before, and now I think they’re the funniest things I’ve ever seen.”) Instead, he embarked on something like a permanent fishing trip on his beloved boat off Long Island, away from people who annoy him and stuff he didn’t want to deal with.

Eventually, his pal Doug McCormick, former CEO of Lifetime and NBCU’s iVillage, called asking for a read on a company in which his firm was looking to invest. That company was PrecisionDemand (McCormick sits on its board). After he signed off on the methodology, Mandel was asked to be CEO.

It was his daughter who talked him into it. “You’re an old fart and you’ve been talking about this for years,” Mandel recalls her saying. “You get a cocktail in you and you talk about GRPs and all this stuff. What have you done? Nobody’s going to remember you. You bought media; now you want a legacy.”

So this, Mandel decided, would be it.

The jury’s still out on whether Mandel’s so-called legacy is in fact the Holy Grail of media research.