How Velvety is Your Wine?: How Asking the Right Question Can Help Shape Consumer Perceptions

January 1, 2011

It seems too good to be a true: a costless way to get consumers to like your product more, to pay more for it, and to consume more of it that involves no changes to the product itself. However, a new working paper from Jong Ming Kim and Professor Nathan Novemsky of the Yale School of Management suggests that, at least for certain categories of products, this effect can be achieved just by having consumers rate a product on dimensions that are described using elegant language. For example, when people were asked to rate how velvety a wine was, they liked it considerably more than when they were asked to rate how smooth a wine was, even though both groups rated the same wine.

However, lest we see a rush of customer satisfaction surveys asking consumers to rate the velvetiness of a whole range of goods, Kim and Novemsky determined that this effect is only present when the product fits with the elegant wording. So while using fancy language in surveys influences consumers favorably for wine, chocolate, and coffee, it backfires with cheap liquor. Why? Because consumers don’t respond well when the language doesn’t fit with the product in question. When the language doesn’t fit the product (for example, asking consumers how “honeyed” they find low-quality liquor), people actually like the product less than if they had been asked to evaluate it using plain language.

Even when your product warrants elegant descriptors, fancy language won’t have the same impact when used in advertising that it does in the survey context. Consumers are alert to attempts to influence their behavior and compensate when they think someone is trying to persuade them. A commercial touting a wine as “very velvety” will be recognized by a consumer as a persuasion attempt and discounted. In contrast, filling out a survey doesn’t trigger the same attempt to avoid persuasion because consumers don’t instantly recognize as an attempt to change their opinions. Accordingly, as long as the consumer isn’t bothered by a disconnect between the language and the product that appear in the survey, the survey’s word choice can impact the consumer by slipping “under the radar” of psychological defenses to influence perceptions and behavior.

Takeaway #1: You are better off asking consumers to evaluate your product using language that fits with the product. For example, Kim and Novemsky highlight that the relative brand positioning of Starbucks and Dunkin Donuts suggests that while Starbucks would benefit from using elegant language when asking ratings questions, Dunkin Donuts should stick to plainer attribute descriptions when surveying customers.

Takeaway #2: When consumers are primed to be suspicious of claims (when viewing ads, for example), elegant language won’t help sell an elegant product. However, the language that appears in surveys doesn’t automatically raise the same consumer suspicions and, when it is in accord with the product itself, has the power to change consumers’ feelings and decisions.