MKR 7000 Module 2 A Three-Stage Research Design Example

Explore, Describe, Then Test: Designing the Meal-Subscription Study

Student Name

Master of Business Administration, Thomas Edison State University

MKR-7000: Marketing Research

Mentor Name

Month Day, Year

Explore, Describe, Then Test: Designing the Meal-Subscription Study

Purpose

This proposal designs the research approved in principle for Hearthstone Market, a composite 22-store Wisconsin grocer weighing a prepared-meal subscription. Babin (2019) describes three basic research designs, exploratory, descriptive and causal, each suited to a different kind of question. The proposal uses all three in sequence, because the decision needs insight into what customers want, a measure of how many want it, and evidence of what they actually do when offered it.

Stage 1: Exploratory Research

Exploratory research clarifies problems and generates ideas when little is known. Two sources will be used in the first three weeks. Secondary data from the loyalty program will show which households already buy prepared foods regularly, how often and at which stores, at no added cost. Three focus groups of eight loyalty members each, one with busy families, one with older couples and one with single professionals, will explore how people decide what to eat on weeknights, what they dislike about meal kits and what a grocer's subscription would need to offer. Krueger and Casey (2015) recommend recruiting participants who share key characteristics within each group, which encourages open discussion, and holding several groups per segment when budgets allow; the budget here allows one per segment, so findings will be treated as directional. The output is a list of features, price ranges and concerns to measure in the survey.

Stage 2: Descriptive Research

Descriptive research measures the size and characteristics of a market. An online survey of loyalty members, sampled by store, will estimate the share of households likely to subscribe at three price points, rate the importance of the features identified in the focus groups and measure expected duration of subscription. The survey design, measures and sample are the subject of the next assignment. Its weakness is that it records what people say rather than what they do, which is why a causal stage follows.

Stage 3: Causal Research

Causal research tests whether one variable changes another. A four-week field experiment will offer the subscription in four stores while four matched stores, similar in sales, demographics and prepared foods volume, continue as normal. The main outcome is sign-ups per thousand loyalty households, with four-week retention as a second outcome and prepared foods counter sales as a check on cannibalization.

Comparing the change in test stores with the change in control stores, a difference-in-differences approach, separates the effect of the offer from seasonal trends or chain-wide promotions. Field experiments of this kind often surprise managers: Lodish et al. (1995), analyzing hundreds of split-market advertising tests, found that many changes expected to raise sales did not, which is the reason to test before committing $1.2 million.

Threats to Validity

Shadish et al. (2002) describe threats to internal validity that could bias the result. History is a risk if a local event, such as a competitor's promotion, affects one group of stores; matched pairs in the same market areas reduce it. Selection is a risk if test stores differ systematically; matching and the pre-period comparison address it. External validity, whether results generalize to all 22 stores, is limited by testing in only four, so the test stores should span urban and suburban locations. A four-week test also cannot show long-run retention, so the launch decision should include a review point after six months.

Timeline and Budget

The research will take ten weeks: three for exploratory work, three for the survey including fielding and analysis, and four for the test, with the pre-period overlapping the survey. Estimated costs are $14,000 for focus groups, $16,000 for survey programming, incentives and analysis, and $55,000 for the test, mostly meal production, packaging and staff training, for a total of $85,000.

Ethics

Focus group and survey participants will give informed consent, be told how their responses will be used, and have their data reported only in aggregate. Loyalty data will be analyzed without names. In the field test, customers will receive the advertised service at the advertised price; if the service is withdrawn after the test, subscribers will be notified in advance and refunded for any prepaid meals.

How the Stages Inform the Decision

Each stage feeds the next and the final decision. If the focus groups reveal that busy families care most about portion size and weeknight timing, the survey will measure those features at three price points; if the survey shows interest concentrated among households near certain stores, those stores become candidates for the test. The launch decision then rests on criteria agreed in the first assignment, such as a minimum sign-up rate in test stores relative to controls and at least 60 percent of subscribers renewing after four weeks. If results fall short, the research still has value: it shows which features or price points might make a smaller pilot viable, or confirms that the money is better spent expanding the prepared foods counters.

Conclusion

Sequencing exploratory, descriptive and causal research gives Hearthstone ideas, estimates and evidence of behavior before it commits $1.2 million. Of the three stages, the store test carries the most weight, since it records actual sign-ups instead of stated intentions.

References

Babin, B. J. (2019). Essentials of marketing research (7th ed.). Cengage.

Krueger, R. A., & Casey, M. A. (2015). Focus groups: A practical guide for applied research (5th ed.). SAGE.

Lodish, L. M., Abraham, M., Kalmenson, S., Livelsberger, J., Lubetkin, B., Richardson, B., & Stevens, M. E. (1995). How T.V. advertising works: A meta-analysis of 389 real world split cable T.V. advertising experiments. Journal of Marketing Research, 32(2), 125-139. https://doi.org/10.1177/002224379503200201

Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi-experimental designs for generalized causal inference. Houghton Mifflin.