“A lost wallet lies on a Manhattan street, stuffed with cash. A white middle-income male New Yorker, between 30 and 44, picks it up. Will he look for the rightful owner, or pocket the cash? Who knows? But if George Costanza, the white middle-income male New Yorker between 30 and 44 from Seinfeld, picks it up, everyone knows exactly what he’ll do. He’ll almost certainly keep the money, yapping endless self-justification to his friends at the coffee shop to conceal his feelings of guilt.”

This brilliant opening to Elizabeth Gardner’s Internet Retailer article, “Persona-lizing a Site,” shows exactly how personas [define] can lead to actionable insights that demographic and psychographic data alone can’t provide. But for me, her next example illustrates something far more important: How and why most of what passes for “personas” falls short. Here’s what she writes:

“It’s hard to target a message to a generic 35-year-old middle-class working mother of two. It’s much easier to target a message to Jennifer, who has two children under four, works as a paralegal, and is always looking for quick but healthy dinners and ways to spend more time with her kids and less time on housework.”

Of course, Elizabeth is just sketching her example persona to save space and to illustrate a broader concept, but you’d be surprised how many companies purchase (or create) “personas” that are little more than this kind of demographic segment with a name and photo attached.

Compare “Jennifer” from the last example to George Costanza, or any favorite fictional character, like Carrie Bradshaw or Atticus Finch. Can you imagine Jennifer speaking to you as easily as you could these other characters? Can you hear her express her likes and dislikes?

Of course not. Jennifer isn’t a living, breathing persona — she’s just a cardboard cutout, masquerading as a persona. She’s effectively mute, and that makes Jennifer all but useless.

If your marketing personas won’t talk to you the way your favorite fictional characters will, how can you possibly use them to plan conversations? How can you create persuasive content/copy for your personas if you can’t imagine their reactions as easily as you can simulate Costanza’s rationalizations?

Quite simply, you can’t — a point that Jason at 37 Signals drives home in his post on Personas. Notice that first on his list of things personas don’t do is “talk back.” Well, he’s right. Far too many so-called “personas” don’t talk back. And he’s also quite right in concluding that non-talking personas are all but useless.

But he’s stunningly wrong in two areas:

1. (Real) personas DO talk back (and answer questions, have opinions, etc.). Three-dimensional characters, such as Costanza, do talk back. And while personas aren’t a replacement for real life testing and results, the insights and answers they provide can guide and direct testing far more effectively and profitably than any other method currently used.
2. Personas are meant to facilitate persuasion, not usability. Good usability is generally a human thing. Effective design impacts all humans/users (within a given culture) similarly, whereas persuasion is very much an individual thing. What’s convincing to one person can actually be offensive to another — which can make relying on one’s self as a model for persuasion a particularly bad idea.

Without personas, we naturally tend to craft messages that are persuasive to ourselves; messages that often miscommunicate to our core audience. Even when we consciously try to account for the psychology of others, our abilities as intuitive psychologists often fail us; something that’s been well documented by social psychologists.

We’re all prone to misjudging others, especially those who are different; hence the need for psychological aids for creating empathy and motivational insight. And while personality theory and psychographics can help — as testified by their effective use by salespeople, marketers, politicians — theory has limitations when it comes to generating real, usable empathy.

Empathy created from less-than-conscious processing is different from perspective gained through conscious, cognitive analysis. In fact, according to research highlighted in Dan & Chip Heath’s book, Made to Stick, analytical thinking can actually impair empathy. This is why imagined simulation is a more effective vehicle for achieving empathy than conscious analysis.

So the “Costanza test” is actually a good litmus test for actionable personas.

For real people, talk may be cheap. But for personas, talking is really all they’re good for. And it’s these imaginatively constructed conversations that lead to persuasive messaging and improved marketing ROI. It’s talking that matters.

So if your personas don’t talk to you, fire them!

And then call Future Now