1. Would you recommend that Roland accept the YKG Group proposal?

2. If yes, what conclusions can be drawn from the data in Phase 3 of the research?

3. If the proposal is not accepted, what alternative designs should be considered?

Roland Development was a leading builder of homes in the western United States. Its emphasis was on condominiums and ownhouses, which were forecast to have an attractive future in these markets. These housing types lent themselves to standardization and cost-reduction possibilities. Further, rising land costs were causing the share of single-family detached houses to decline significantly. Meanwhile, the share of market for single-family attached houses (houses with common walls, floors, or roofs) was expected to double in the next five years.

Roland was well positioned to exploit these trends by following a strategy that differed from competition in three areas:

• Market segmentation. Roland typically segmented the market more finely than other home builders and then designed homes to meet the specific needs of these groups.

• Direct selling. Shoppers in some department stores could find full-scale, fully furnished Roland homes on display.

• Low prices for a complete housing package (including all the furnishings and necessary financing).

The company had begun to expand its limited line of condominiums and townhouses to provide design and squarefootage combinations that would appeal to higher-income households. The management was especially pleased with the elegance, convenience, and durability of the four new models they were planning to launch. Several problems remained to be solved. The first was the identification of a creative strategy that would position the new models and attract the largest number of purchases. That is, the company wanted to know what main ideas and themes should be used in the advertising of the new models. Another problem was to identify those segments of the market with the highest probability of purchasing the new models. The company asked the YKG Group, a large national research firm, to submit a written proposal for research which would provide Roland’s management with information useful in solving these two problems. Their proposal is summarized below.

Research Proposal

The recommended research design would use a consumer panel and employ both telephone interviews and mail questionnaires. The research firm felt that the needed information could be obtained only from that very small proportion of the population who might buy such a home. Each of several different market segments would be studied to determine how they positioned the new models in relation to competing homes already on the market. The likelihood of purchasing a Roland model would also be determined during the study for each of the three market segments and also for each of several different advertising themes. This information would help Roland identify the most promising market segments for the new models, as well as the creative advertising strategy that would most appeal to them. The proposed research design consisted of three phases:

(1) The members of a large consumer mail panel would be screened to locate qualified prospects for the new models; (2) a relatively small sample of qualified prospects would be interviewed “in depth” to identify possible advertising themes; and (3) a large sample of qualified respondents would be surveyed by mail to test their response to alternative creative strategies.

Phase 1. The YKG Group maintained a bank of over 200,000 families who agreed to cooperate in research projects undertaken by the firm. Considerable information existed about each family, including geographic location, occupation and age of male and female heads of family, total family income, and presence and age of children. Roland managers felt that the four new models would most likely appeal to middle- and upper-income families of size two, three, or four, with a household head 30 years of age or over. For this reason, the first phase of the proposed research involved mailing a short questionnaire to all panel members with those characteristics. 

The questionnaire asked panel members to indicate the likelihood of their purchasing a home in the next two to three years and also to report their attitude toward buying a townhouse or condominium. It was expected that this screening process would locate some 3,000–5,000 families who would be prospects to buy the new models over the next few years. To be considered a prospect, a family had to report being likely to purchase a home in the next two to three years, as well as having a favorable attitude toward a condominium or townhouse. Among these prospects, three market segments would be identified. A highincome family would be a “very good” prospect if it was “very likely” to buy a home; a medium-income family would be a “good” prospect if it was “very likely” to buy a new home; and a high-income family would be a “fair” prospect if it was “somewhat” likely to buy a new home. All other responses were considered to indicate nonprospects.

Phase 2. In this phase about 200 qualified prospects would be interviewed using a combination of telephone and mail. These families would be mailed pictures, specifications, and line drawings of the company’s new models of condos and townhouses, although they would not be identified by the Roland name. The line drawings would include front and rear views of each unit’s exterior as well as sketches of each room. The specifications would include the number of square feet, wall thickness, heating and cooling equipment capacities, appliance brands and models, slab thickness, type of roof covering, and other features.

After reviewing these materials, respondents’ reactions and impressions would be obtained through telephone interviews using open-ended questions. Interviewers would be told that the objective was to obtain qualitative data useful for ascertaining how potential buyers perceived the new models with respect to appearance, comfort, elegance, convenience, durability, ease and economy of maintenance, and other criteria. Interviewers would be instructed to record verbatim responses and were told that it was very important to do so because none of the responses would be tabulated or analyzed statistically.

Responses to the open questions would then be studied to identify four or five ideas of themes that might be considered for use as creative strategies in advertising the new models.

Phase 3. This phase would be undertaken after four of the best advertising themes had been identified. Some 2,400 families would be selected from the list of prospects obtained from Phase 1—approximately 800 “very good” prospects, 800 “good” prospects, and 800 “fair” prospects.

All of the families in each of the three market segments would be sent pictures, line drawings, and specifications (including prices) of the new Roland models as well as those of major competing models, all identified by brand name. Each of these three groups of prospects would then be randomly divided into four subsamples of 200, each of which would receive one, and only one, of the four advertising themes identified for the new models. Thus, the study design would consist of three samples of 800 families each. In turn, each sample would be broken into four subsamples, each of which would receive a different advertising theme.

Analysis. The effect of each advertising theme on each prospect segment would be evaluated on three measures: the degree to which it (1) resulted in the new line being rated as “most appealing,” (2) led respondents to request further information about the company’s products, and (3) led respondents to indicate that they would be most likely to select one of the company’s homes if they were to make such a purchase in the near future. For each advertising theme-prospect segment combination, the research would yield three percentages.

For example, for theme #1 and the “very good” prospects, the research might show that 38 percent of the respondents found a model in the new line “most appealing” among all the models reviewed; that 26 percent requested further information about the Roland models; and that 17 percent indicated that they “most likely would purchase” one of the new Roland models.

By comparing these three percentages for each advertising theme-prospect segment combination, it would be possible to identify the most promising combinations. These results could be weighed by the relative size of each prospect segment to decide which creative strategy would be most effective in generating sales interest in the new models.