We recently reported on important new neuroeconomics research in Brain Scans Predict Buying Behavior. This study is the first that attempts to correlate fMRI brain scan data with actual purchasing behavior. George Loewenstein of Carnegie Mellon University, Brian Knutson at Stanford, and other researchers presented subjects with cash, put them in an fMRI machine to record their brain activity, and then presented them with different offers of items and prices. Some of the offers were overpriced, and others were good value. The subjects then actually chose to buy items with their money, or keep the money. The researchers compared self-reporting of purchase intentions by the subjects, brain scan data, and actual purchases. (You can read the full paper in Neuron.) We interviewed Loewenstein after publishing that article, and are happy to share a few highlights here.

According to Loewenstein, one significant aspect of the findings is that the brain scans predicted buying behavior almost as well as the self-reported intentions of the subjects. In other words, absent any knowledge of what the subject intended to do, viewing the brain scan was nearly as predictive as asking the subject what he would do. Loewenstein noted that in this experiment, the questions about the intentions of the subject were quite straightforward and would be expected to be good predictors of actual behavior.

From a neuromarketing standpoint, we think this finding is particularly important. While simply asking people questions is a lot less expensive than sticking them in fMRI machines, there are many situations where the individual may not respond with complete candor, either intentionally or unintentionally. The fact that the scans do almost as well as self-reporting in a situation with no emotional loading suggests that they might do better than conventional methods when user self reporting is likely to be less reliable. (Indeed, that’s the point that advocates of fMRI lie detection make, though certainly using scans as legal evidence of truth or falsehood is certainly not imminent.) Lowenstein makes the point that the fMRI data enhances the self-reporting data.

The “negative” activation produced by cost is relative, according to Lowenstein. That is, it isn’t just the dollar amount, it’s the context of the tranasaction. Thus, people can spend hundreds of dollars on accessories when buying a car with little pain, while a vending machine that takes 75 cents and produces nothing is very aggravating. Auto luxury bundles are designed to minimize negative activation because their price tag covers multiple luxury items. The consumer can’t relate a specific dollar amount to a particular item, e.g., $1000 for leather seats, and hence can’t easily evaluate the fairness of the deal or whether the utility of the accessory is worth the price.

Cost isn’t the only variable that causes “pain” – it’s really the perceived fairness or unfairness of the deal that creates the reaction. Other parts of an offer that caused it to appear unfair would presumably cause a similar reaction as a too-high price.

Overall, Lowenstein isn’t interested in using this work for neuromarketing purposes. He points out that for many years credit card companies have prospered while encouraging consumers to spend too much by exploiting the principles he’s now uncovering his research. In essence, for a large number of consumers, the credit card takes the pain (quite literally, from the standpoint of the customer’s brain) out of purchasing. Pulling cash out of one’s wallet causes one to evaluate the purchase more carefully.

We think this makes a lot of sense, and is entirely consistent with real-world behavior. A credit card reduces the pain level by transferring the cost to a future period where it may be paid in small increments. Hence, not only does a credit card enable a consumer to buy something without actually having the money, but it also puts a finger on the scale as one’s brain weighs the pain vs. the benefit of the purchase. This can be a bad combination for individuals lacking financial savvy.

Lowenstein didn’t analyze Super Bowl ads, and he thinks that we aren’t yet at the point where marketers can do meaningful branding studies with fMRI scans. Lowenstein is part of the faculty at CMU’s Social and Decision Sciences Department, an interdisciplinary program that studies a wide range of topics, including the psychology of decision making, innovation and entrepreneurship.

email