Editor's note: Robert A. Roy is vice president, client service, with the Chicago office of Total Research Corporation. This article is the latest installment in his series on the research efforts of the mythical Second City Bus Service. Previous installments have appeared in the March 1991 and October 1990 issues of QMRR.

Author's note: A smile, a smirk, and especially a grin should never be mixed with marketing research. After all, it is very serious! I have, however, an excuse. My mind has been clouded by Lamont Cranston. (Who, you ask? Only The Shadow knows!) Hereafter, I'll only write in the most leaden tone. Yes, I will - Not!

Ned Dweeb and Sally Dumluck continue to labor as research analysts in the third basement of the Second City Bus Service. Their corporate superior - and mental and moral inferior - Ed Goodride, enjoys his top floor corner office and the title of director.

The bi-annual meeting of the board of directors is in progress as John Gottrocks III starts to pontificate, "When I was at the Hahvahd B School, it was ahxiomatic to think it terms of mahket segmentation. Just today, while wahlking across my yahd to the pahked cah it came to me! Of course! Even the transportation marhket in this mistaake on a laake must be segmented! Don't you agree, Goodride?"

"Oh, rolling wheels! Did he just say 'Goodride'?" Ed could think fast and lie faster. In that spirit, he responded, "I sure do, Mr. Gottrocks. In fact, I'm planning to announce at the next board meeting how the market is segmented, and how can we increase ridership."

Ed could feel the sweat that had nothing to do with temperature roll down his back. Slipping out the door, ignoring the elevator, he ran to the third basement looking for Ned Dweeb.

Dweeb's dilemma

"Hey, Dweeb," Ed panted, "D'ya remember that segmentation study I told you to do? Well, how's it going?"

"Well Mr. Goodride, uh, I...did you say segmentation study?" Dweeb squeaked.

"Sure, you know. The way they do it at the @#%& Harvard B School! I'm going to present it at the next board meeting. Don't let me down. I'm outta here, Dweeb." With that, Ed slithered out the door.

"Sally," gasped Ned, "what will I do? What will I do?"

"Do about what, Neddie? Goodness, you look terrible."

"Oh, Sally, Mr. Goodride just laid his 'I told you' act on me. This time I'm supposed to have a segmentation study done by the next board meeting. Here's what I think. I'll measure the importance of, say, thirty attributes of transportation selection. Also, I'll be sure to measure relative importance.

"Let's see, you said I could have the attributes either rank-ordered, or use a rating scale that's anchored. If I use an anchored rating scale, a person will first identify the most important attribute and place that on, say, the number ten. The least important will be placed on the number one, and then all of the remaining attributes are rated. Am I getting this?"

"You got it, doll!"

"Then I'll use the Howard-Harris clustering program to find groups of people (segments) that agree about the importance of the attributes. Good plan, eh?"

Sally's gentle nudge

"Neddie, you're on the right track. I especially like the part about measuring the relative importance of attributes, but let me show you a possible problem. Using the Howard-Harris program, you might get results like these:

Exhibit A
Benefit Differentiation

Segment

Importance of Attribute A

Segment I

9

Segment II

22

Segment III

45

Segment IV

57

Segment V

34

Exhibit B
Benefit Differentiation

Segment

Importance of Attribute B

Segment I

38

Segment II

49

Segment III

9

Segment IV

24

Segment V

35

"Just as you said Ned, you can find groups of people who value the attributes differently."

"Then you agree, Sally!" Ned positively beamed at the compliment.

"Almost, my pocket protector pal. At the same time you find segments that are different with regard to the importance they place on the attributes, they may not be different in terms of their behavior. Let me show you:

Exhibit C
Conventional Segmentation

Segment

% Taking a Camel

Segment I

6

Segment II

6

Segment III

5

Segment IV

6

Segment V

6

"See, Neddie? In my example, the five market segments all have about the same percentage who usually take a camel to work! "

Dweeb scratched his head, not sure what she was getting at, and replied, "Is that bad, Sally?"

"It's absolutely dreadful, Neddie-pie. By the way, do you have plans tonight? Hold on that. Let me continue. You see, if you find market segments that regard attributes differently, but behave the same, that's not very useful, right?"

"It sure isn't, you valedictorian wizard, you!"

"Neddie, you sure have a way with words. Here's an example of the kind of results we'd like to see:

Exhibit D
Predictive Segmentation

Segment

% Taking a Camel

Segment I

3

Segment II

14

Segment III

0

Segment IV

5

Segment V

6

"If you got results like this, you would have found segments that are different both in terms of attribute importance and in behavior. In this example, more than twice the number of people in Segment II (14%) ride a camel to work as those in Segment V (6%). I call this 'improved segmentation."'

"Great, Sally, but why do you call it improved segmentation?"

"Because, Ned, we can find segments that value attributes differently, while at the same time predict their behavior!"

"How can I find out how to do it?" Ned pleaded.

"There's the rub, Neddie. Different analysts use different approaches. For example, one talks about finding a 'common space of attributes and measures of behavior.' This is an application of multidimensional scaling. Another practitioner advocates using a clustering program where the variables are both attributes and purchases. Critics, and there is no shortage, question how to weight the two sets of variables. Critics aside, however, there is no doubt that the goal of finding market segments that differ both in terms of attributes desired (benefits) and critical marketing behavior, such as purchases, is correct. Dweeb, remember Spacey Sam the Statistician we had lunch with yesterday?"

"Yes, Sally."

"If you tell that space cadet what you want, I'll be he can dig deep into his bag of tricks and do it. Oh, and one more thing, my cuddly, bespectacled boy: never forget to pay close attention to back issues of that learned journal, The Marketing News."

"Sally, how can I ever repay you? You've saved me once again. You're quite a gal! I'll improve my segmentation, get Mr. Goodride off my back, and read even more of those fine articles in The Marketing News."

Back in the board room

Dweeb came through. Goodride took the credit and made his presentation. John Gottrocks III continued to talk about the Hahvard B School. And Sally had her way with Ned.