Persuasive Technology -- Not

May 12, 2003 (Computerworld)

B.J. Fogg has many rules of thumb about what works and what doesn't in persuading users through technology.
As director of the Stanford University Persuasive Technology Lab, he consults with large companies and directs laboratory tests. Sometimes it's instructive to know what doesn't work at persuasion, he says.
One thing that doesn't seem to work, or at least is risky, with current technology is screen avatars, those partly animated faces that pop up and offer to be a customer service agent. "Having a screen avatar doesn't help much," he says. "It's a risk to use one because if I have a bad product to sell and use an avatar to sell it, the avatar acts like I'm turning up the volume. If buyers hate the product, they hate it more with an avatar."
Fogg, in his book Persuasive Technology: Using Computers to Change What We Think and Do (Morgan Kaufman, 2002), says that using avatars or computers as social actors works best when selling leisure, entertainment or educational products, but is less appropriate when the point of the technology is to improve efficiency.
He offers the example of buying gas at a pump with a credit card when the customer doesn't want to deal with a cashier or have a social experience. When people buy certain things, it's more of a job and not a social event, and adding in social cues at such times would be "distracting, annoying or both."