The following article, written by Ira N. Bachrach, president of NameLab Inc., is an excerpt from a chapter in his proposed “how to” book titled Making Names, which is due for publication in 1990.
When product names are assessed by the usual techniques (attribute panels and focus group discussions), results are often taken with a grain of salt by sophisticated new product marketers. As one such manager put it: “It looks like this study would have rated Lite ‘n Natural over Jell-O and Strong’ n Gentle over Bufferin when those products were named, and that makes me suspicious.”
At NameLab, because our only business is making names, we haven’t run name tests in-house since 1985. But in the course of numerous name development projects, we have watched various clients conduct nearly 100 name evaluation studies. As a portion of our fee depends upon our client’s decision to register and use a name we have proposed, it is inappropriate for us to criticize the design or the results of a study involving our candidate names.
So we hold our tongues, unless asked by the marketer or marketing researcher to suggest a test design or comment on a muddy result. But over the years, we have observed what seems to be the root of the name testing problem and an innovative test method which seems to avoid it.
Name function tests aren’t really about names. The marketers couldn’t care less whether their new shampoo is called Sweet ‘n Sassy or Salamander Sweat. All they care about is the effect of the name on the consumer’s expectations of the product.
The difficulty with most name tests is that any consumer whose IQ exceeds their waist size realizes that the interviewer or moderator is asking about names. As names are words rather than shampoos, the consumer tells the interviewer or moderator whatever that particular word means or suggests, including (in one focus group tape we saw) that “It reminds me of my grandmother cursing in Polish,” and (in another tape) that “It has too many n’s. Wouldn’t it be better to use a lot of different letters?”
The problem is rarely so humorous or obvious, but it affects most name studies, although this is somewhat reduced by monadic structure. (In a monadic test, the consumer sees only one name.) After an investment of time and money in name testing, the marketers are presented with reports that focus on the relative characteristics of words rather than how a candidate name as an attribute of a product would affect expectations of that product (and presumably purchase intent).
A small number of packaged goods companies employ a technique called adjectival analysis to compare consumers’ expectations of a product with one name versus expectations of the same product with another name. Here’s how it works.
In mall intercepts (or whatever acquisition technique makes sense for the product being researched), appropriate consumers are shown what appears to be a full-page magazine advertisement for the product.
Each consumer is shown only one ad (the test is monadic), although separate ads are prepared for each of the names being tested. The ads are realistic but simple: a clear photo of the package, a simple headline like “A New Shampoo,” and the product name in large, uncomplicated type at the bottom. Only the product name on the package and at the bottom of the layout changes from one ad to the next. After seeing the ad, the consumer is asked the least directive question possible. (The question we like best is “What do you think?”) His or her answers are tape recorded.
Typing off the tapes yields a stack of verbatim responses to the product with the name Sweet ‘n Sassy and a second stack of responses to the product with the name Salamander Sweat. Obviously you can evaluate as many names in such a study as your budget and patience permit.
An analyst goes through each stack of responses, crossing out all words that are not adjectives, leaving one list of adjectives used in response to the product when named Sweet ‘n Sassy and a second list of adjectives used in response to the product when named Salamander Sweat.
On examining the two lists of adjectives, the analyst will observe that the majority are the same in both lists, reflecting reactions to common elements (like package design) in the ads and to the generic idea of shampoo. Deleting those adjectives largely common to both lists leaves clusters of adjectives on each list that elucidate the differences in consumers’ expectations of the same product with another name. And that is usually what the marketers want to know.
In the English language, qualities of objects or experiences are largely communicated in adjectives, such as hot, rich, exciting, or gentle. By stripping away other semantic elements of language (verbs, nouns, and so forth), you reveal what you really want to know: Which candidate name creates product expectations that most closely match the marketers’ goals, without the obfuscating fog of respondents’ varying abilities to express ideas in complex semantic constructions and analysts’ varying abilities to comprehend those constructions?
Most importantly, adjectival analysis deals with consumers’ expectations of products rather than their ideas about the words we propose as candidate names. It’s simple to put together, reasonable in cost, and produces a report which is understandable and defensible.
We at NameLab are linguists, not market researchers. While we did invent the linguistic structure of adjectival analysis, much of the practical organization of the method described here stems from trial-and-error refinement of our idea by researchers at a few packaged goods companies over the years.