Advertising Psychology—or an Advertising Psychologist?

A funny thing happened during the Super Bowl. No, not that thing. Yes, the 34-minute power outage brought a lot of issues into focus—issues linked to the "Sandy Studies" Lee proposed in the hurricane's wake (and is now teaching). But I have a different sense of power in mind—the power of (academic) science in the marketplace.

On Sunday, Prudential aired a thirty-second spot (that could've cost up to $4,000,000!) advertising its retirement planning services. The ad featured a "real-life experiment" (their term, from the video's description) emceed by a real-life academic scientist: social psychologist, bestselling author, and Harvard Professor Daniel T. Gilbert.

Details of the proceedings are fleshed out in both the one-minute and behind-the-scenes expansions of the original spot. Some folks—though surprisingly few on Twitter, and comments are disabled for the video—might be curious why Gilbert got involved. This is certainly the first academic I've seen in a pretty straightforward network advertisement (and a Super Bowl one, no less).* 

But for those of us interested in the authority (and marketability) of science, the "why" question matters more with respect to Prudential than Gilbert. If you're a company, why run such an "experiment"? Why this "experiment"? And why film it, rather than use it behind the scenes?

Because it's not an experiment—at least not in the sense it seems to be. 

So what's going on here? In the voiceover for the "behind the scenes" version of the ad, the company gives an account of the motivation and setup for the conceit: 
Today's generation is living seven years longer than the previous one. So how do you get people to start thinking about living longer, and what it means for their retirement? Well, we built an eleven-hundred square foot wall and conducted an experiment. With the help of Professor Daniel Gilbert, and the city of Austin, Texas, we asked people to place a sticker next to the age of the oldest person they've known. And they did. 
When all the stickers were up, participants were asked how they felt about "this longer life that lay ahead of them." The responses were predictable. "It's really really true: people can expect to live to be a hundred," said one, while another felt the display gave him "every reason to believe" he'd hit 100 himself. "I need to start thinking about retirement like, now," said another.

And that was the hook: "We're living longer than ever," the ad concluded, "and yet the official retirement age hasn't changed. This is the challenge—and opportunity—of living longer, and this is where Prudential can help." But was all of this demonstrated ("brought to life," as they put it) by the "experiment" Gilbert and Prudential conducted?

The easy answer is: "No."

The experiment itself provides no evidence that "we're living longer than ever," nor does it furnish the sort of data about average lifespan or life expectancy with which we might reasonably make decisions about our own savings and investments. Would the results of such a survey have been any different a century ago, when the average lifespan was considerably lower? Maybe, maybe not—we have no idea.

Rather than excluding outliers, Gilbert's setup celebrates them. And that's the point. Prudential isn't interested in furnishing data for making informed decisions. As they put it themselves, they just want "people to start thinking about living longer, and what it means for their retirement." If you think you might live to a hundred, you'll sock more away (with Prudential) today. And remembering Grandma's 93rd birthday puts us in that frame of mind. 

That's how Gilbert sees it, too. Here's how he described what they were up to in his plug for the video on Twitter:
Temporal discounting is what accounts for our failure to invest in Prudential save for retirement. It's something academics like Gilbert know a good deal about—and something companies like Prudential are increasingly interested in. This is where we start to understand how these two parties came together in a Super Bowl ad—to put scientific ideas about discounting to work in a very real sense.

What's at issue is the fact that we don't discount the future in the simple, exponential manner that rationality seems to dictate and economists used to assume. Rather, we behave irrationally—and, since the 1970s, economists, decision theorists, and cognitive scientists have been coming together under the banner of "behavioral economics" (and, more recently, "neuroeconomics") to figure out why. (Here's a great, recent overview of the field.)

Since Daniel Kahneman and Amos Tversky published the work that eventually won a Nobel Prize, scholars have continued to show that we tend to value (and undervalue) things in ways that are both irrational and—and this is the key—predictable. This is the idea behind behavioral economist Dan Ariely's bestselling book, Predictably Irrational, and it's the sort of reasoning that gives the counterintuitive punch to Freakonomics et al.

What does all this have to do with Gilbert and the Super Bowl? Well, this is the world Gilbert lives in. His bestselling book, Stumbling on Happiness, provides similarly fascinating stories about how we so often fail to understand what makes us happy (and behave in ways you wouldn't expect). And its based on the well-regarded research on "affective forecasting" that got him a job at Harvard.

And, as it turns out, there's an increasing market for that kind of knowledge. Perhaps the most obvious way to sell it is in book-form, which is what "the Gladwell industry" I wrote about a few months ago (in connection with Jonah Lehrer) is all about. And who blurbed Gilbert's book? Who do you think? There's a lot of money—and lecturing—in behavioral economics. And a lot of fun!

Another interested group includes "non-profits, governments and international agencies," and some academics—like Harvard's Sendhil Mullainathan and his "ideas42" project—are providing them with expertise on how psychology and economics interact. The policy relevance of behavioral economics (in the form of Nudge) was another one of the topics I covered when I wrote about Jonah Lehrer. 

The last (and to some, most pernicious) market is the business community. When Kahneman gave a two-day "Master Class" in Napa on the topic of "Thinking about Thinking," the list of participants was a who's-who like no other—"a microcosm of the recently dominant sector of American business," as the organizer billed it. Predicting how people will spend their money is, it turns out, of great interest to the founders of Amazon and Google.
Notable Kahneman fans (Source:
None of this is new. A 1996 article by our friend Malcolm Gladwell—tellingly titled "The Science of Shopping"—makes that clear: "The practice of prying into the minds and habits of American consumers," he wrote, "is now a multibillion-dollar business." Behavioral economics has never been hotter, and that extends to all the markets mentioned above. 

But putting an academic in an advertisement – that seems new. So why do it? This is where it matters that the experiment in the ad wasn't what it seemed. Prudential doesn't care how old people's grandparents are, and, if they did, they wouldn't hire Gilbert—a psychologist—to figure it out. What they care about is what people think about getting older—and how we can make them think about it in such a way as to invest more in preparation for it. 

And that, as it turns out, is something Gilbert's an expert at. As he told Business Insider last month: "The single best way to make predictions about what you're going to want in the future isn't to imagine yourself in the future," says Gilbert. "It's to look at other people who are in the very future you're imagining." So an ad that celebrates anomalously old people might steer us toward imagining ourselves in their shoes—and wanting to be ready for it. 

But still: why do it as an experiment? This is pure speculation, but—since we can assume every part of that ad was market-tested and focus-grouped—it seems to say that experiments sell. And not just as data: it seems like the process itself—design, setup, participation, analysis—is a marketable move (maybe as long as it features a diverse group of attractive people in a sunny environment?). Prudential's experiment, then, was an experiment in marketing.

Is this a new window onto the relationship between (academic) science and capitalism (as we've discussed here before)? I'm not sure. Is it a good sign (or a bad one, for that matter?) for the marketability and authority of academic science more generally? Again, no idea. But it seems worthy of note that behavioral economics and social psychology—long a staple of both bestseller lists and corporate board-rooms—had a prominent (if brief) methodological moment in the sun. 

*Can anyone think of other examples, either product endorsements or primetime promotions? 


This is fascinating Hank. I missed the Superbowl commercials here in Berlin.

I'll offer one interesting fact: in the 1920 and 30s, as the question of life extension took on some real weight, many people did in fact doubt that even the outliers lived to be in the 90s to 100s range. Because the birth registrations systems were ridiculously porous and incomplete, it was often hard to trust people's birth days and thus easy to ascribe anything past the biblical three score and ten to a vital statistics problem. I'd say there is a big difference now, when we can see so many outliers (which makes each less of an outlier) in a range that was previously very hard to believe.

That makes sense to me! I'm wondering how the logic here would be different for a life insurance company. I take it that MetLife wouldn't want Gilbert convincing potential customers they might live to be a hundred—and that, instead, they'd want him spinning stories about the youngest adult people have known to die. But is that wrong?

I think we can both agree that Prudential isn't *using* this data. They're trying to convince people—with the visual gap between the retirement age and the normal distribution around 90—that they'll live a long time. Is the aggregation of "oldest people known" a historically significant way of either estimating life or prudentially planning? It definitely strikes a chord—which is precisely what they're going for, for which the trappings of "experiment" has been chosen as the marketing vehicle.

Prudential is/was just as much a life insurer as Met. Throughout the twentieth century, nearly all life insurers became group, health, and pension providers, a transition fueled partly by regulatory changes (pensions were not so well regulated as life insurance) but also an awful lot by the massive demographic transition that began in the nineteenth century and led to much longer lives in the twentieth, even setting aside the largest contributor to longer average life expectancy (much, much lower infant mortality).

But yeah: Pru is not doing this for the data it provides. The purpose is not to produce a nice skew curve---although it does that. It's to get people thinking about old people they know. And yet: all those Superbowl viewers don't get forced to think about old people they know---they get forced to look at a skew curve. That suggests that Pru or the advertisers also trust that people can learn something from a graph...

Good first point – I guess I meant that the way you sell Life and Retirement has got to be different, and that they'd be different sorts of clients for behavioral economics. Even if they're the same company! But that's obvious.

Depends what you mean by "learn." As Gilbert gestures, you imagine living through each successive stack of stickers—even though each represents a death. I guess we still have trouble grappling with chance. Hacking was right.

Glad someone else realized the "experiment" was misleading and did not test Prudential’s hypothesis -- seems especially disingenuous given a scientist was attached to the spot. It makes scientists in general look bad and this is the one of the reasons there is so much controversy over potentially political items like global warming. When people see that a scientist at a respected institution (Harvard) is willing to do this bad-science advertisement other respected scientist’s reputations become diminished in their eyes. We don’t know who to believe, and we begin to suspect scientists of having an agenda and not searching for the truth.

The last commenter has hit the nail on the head. The real story here is that Gilbert and prudential have made a mockery of science and the profession of academic education. They redefine the calling as manipulation of the public into doing what they "know" is best for them, rather than informing in a disinterested manner. These people are involved in an elitist pursuit - the company they keep is thus no surprise. What is interesting about this is merely how brazen and shameless the shilling is becoming. In the long run, this is a good thing, as nothing good will come out of our current system of deceptive and self serving institutions until their collapse is completely apparent.

Gilbert is an airhead. He is riding a wave of popular psychology that is almost entirely unrelated to science . He is in love with his image and his popularity. He is a moron and his theories are trash. The experiment in this ad is about the on par for him in quality and legitimacy.

Was it Will Rogers who said about life insurance that They are betting I will live, and I am betting I will die, and I am paying them for it.? I have never seen such a bogus use of statistics. There was always the story of the guy who lost a silver dollar. When asked why was he looking here for it said, "While I think I lost it over there, but the light is better here." Duh...

Note: Only a member of this blog may post a comment.

back to top