BUSI 453 Week 8 QMR Report: Final Example

On-Demand Classes at Riverside Cycle and Strength: A Quantitative Marketing Research Report

Executive Summary

The owner of Riverside Cycle and Strength wants to know whether 640 members would pay an extra $15 each month for recorded classes, a library that would cost roughly $18,000 to produce in its first year. A survey of all members produced 237 usable responses. Members who see on-demand classes as useful, who already use workout videos and whose travel or shift work keeps them from the studio are all more interested in subscribing. About 37 percent would pay $15 or more, rising to 54 percent among schedule-constrained members. After adjusting stated intentions downward, the report forecasts about 45 subscribers in the first year, or roughly $8,100 in revenue, which does not cover the full build. Counting the memberships the library could save from cancellation, however, the project could break even. The report recommends a 90-day pilot costing about $4,500, offered first to schedule-constrained members, with a target of 60 paid subscribers before the full library is built. The pilot limits the studio's risk to a quarter of the full cost while producing the evidence a survey cannot: what members actually pay for and whether subscribers cancel less often.

Revisions From the Drafts

This report combines the three drafts and revises each in response to the instructor's comments. The approach now states the analytical model in words and ties each construct to a measured variable. The research design section explains why an online survey was chosen over other survey methods and lists the pretest changes with the reason for each. The analysis adds a multiple regression, a sensitivity table for the forecast and a plain-language interpretation of every result. The forecast and recommendation sections are new, completing the sixth step of the marketing research process, report preparation and presentation (Malhotra, 2015).

Problem Definition

The studio's owner faces a declining attendance trend: average weekly visits per member fell from 2.4 to 1.9 between 2022 and the start of this study, and 31 percent of members who canceled in the past year cited schedule conflicts, chiefly business travel and shift work. An on-demand library filmed with the studio's own instructors could keep those members exercising with the studio when they cannot attend in person, and could earn added revenue. Two larger gyms in the area already bundle app-based workouts into their memberships, though with generic instructors rather than familiar ones. The owner's question is therefore not whether on-demand fitness has a market, which national brands have shown, but whether Riverside's own members value it enough, at a price, to justify the studio producing it. The decision matters because $18,000 is roughly two months of the studio's profit, and a library that few members use would also take instructors' time away from live classes.

In the language Malhotra (2015) uses, the management decision problem is whether the studio should launch the $15 monthly on-demand subscription. The information problem behind it is to estimate adoption of a recorded library among members, the price they would accept and the traits that mark the likeliest adopters. The research was also expected to inform how any library would be marketed, since knowing why members want it shapes what the studio says about it. Four specific components guided the study: the share of members likely to subscribe at $15, the class types they want, whether schedule-constrained members are more interested, and whether members who already use workout videos are more interested.

Approach

The study rests on the technology acceptance model, which traces a person's plan to adopt a new tool mostly to two judgments, how much it would help and how easy it would be (Davis, 1989). A later synthesis of eight adoption models added the influence of other people and the practical support available, and showed that experience changes how strongly some of these beliefs matter (Venkatesh et al., 2003). In this study, perceived usefulness is measured directly; prior use of online workout videos represents experience and ease of use; and schedule constraint represents the situation that makes the library useful.

Three hypotheses follow:

• H1: The more useful a member judges recorded classes to be, the stronger that member's intention to subscribe.
• H2: Intention to subscribe is higher among members who streamed or watched any workout video during the half year before the survey.
• H3: Intention to subscribe is higher among members whose jobs require monthly travel or nonstandard shifts.

A fourth question, asked without a formal hypothesis, was whether each of these factors predicts intention when the others are held constant.

Answering these questions required seven survey measures, covering usefulness, intention, top acceptable price, wanted class types, home video habits and work schedules, plus tenure and attendance drawn from the booking system. In response to feedback on Draft 1, the analytical model was stated more precisely: usefulness is expected to drive intention, and video use and schedule constraint are expected to raise intention both directly and through usefulness.

Research Design

Design Type

Booking records had already pointed to the likely market: about a third of members, 212 in all, had a month with three or fewer visits, and cancellations clustered in the summer and December travel seasons. Those records located the members most likely to benefit but could not reveal interest or price, so primary data were needed. The study used a descriptive, single cross-sectional design (Malhotra, 2015). Focus groups would have suggested content ideas but could not estimate adoption, and a priced trial, a causal design, would have required building the library before knowing whether it was worth building. Among survey methods, an online questionnaire delivered through the studio's app and email was chosen over telephone or paper surveys because it was inexpensive, let members respond at a convenient time, and reached the infrequent visitors the study most needed to hear from, who would never see a paper survey at the front desk.

Measurement

Measures were built from defined constructs and multiple items, following the measure-development paradigm Churchill (1979) proposed for marketing. Usefulness was captured with four agreement items reworded for a fitness setting from the published scale of Davis (1989), and intention with three agreement items about subscribing at the $15 price. Prior video use and schedule constraint were yes-or-no items, the latter with examples of irregular work. Willingness to pay was an open dollar question, and class preferences were ranked. Summed Likert scales were analyzed as interval data, a treatment parametric tests handle well (Norman, 2010).

Hypothetical willingness-to-pay questions tend to overstate what consumers will actually pay (Miller et al., 2011), so stated prices were treated as ceilings. The questionnaire ran 18 items in about five minutes, short enough to limit satisficing, in which tired respondents give acceptable rather than accurate answers (Krosnick, 1991).

Pretest

Eight members of varied ages completed the questionnaire aloud with the researcher. The pretest led to three repairs: one item that asked about convenience and cost together became two items, the work-schedule question gained concrete examples because members had read it in different ways, and the offer description gained a sample list of classes so members could judge the price against real content.

Sampling

Because the studio could reach every member by app and email at no cost, all 640 members were invited, a census attempt rather than a sample. The survey's opening screen explained its purpose, its length and how answers would be used, stated that participation was voluntary, and promised that results would be reported only in summary form. Responses were matched to booking records by a code rather than a name. The principal risk was therefore nonresponse error, not sampling error (Malhotra, 2015). Response rates are only a weak guide to nonresponse bias, which depends on whether the decision to respond is related to the variables measured (Groves & Peytcheva, 2008), so the analysis tested for bias directly.

Fieldwork and Data Preparation

Fieldwork ran for two weeks in March with a single reminder at the halfway point, and every completed survey earned a smoothie credit worth $10. In all, 251 members responded, 39.2 percent of the membership. Eleven responses missing a hypothesis variable were removed. Two further screens, both recommended for catching careless online answers (Meade & Craig, 2012), looked for surveys finished in a minute and a half or less and for surveys that gave one identical answer to every agreement item. Three responses failed both screens and were removed; four that failed only one were kept after inspection. The final data set held 237 responses, 37.0 percent of members. Coding followed the plan in Draft 2, with agreement items scored one to five, yes answers scored one and prices kept in dollars.

Respondents had been members for an average of 26 months, close to the studio average of 24. Because late respondents tend to resemble people who never respond (Armstrong & Overton, 1977), the 65 surveys received after the reminder were set against the 172 received before it. Average intention was 2.86 in the late group against 2.97 in the early one, a gap too small to be significant, which suggests but does not prove that nonrespondents were not sharply different. The slightly lower score among late respondents points in the expected direction, however, and the forecast below assumes nonrespondents are less interested than respondents rather than equally so.

Analysis and Results

Reliability and Descriptive Results

Both multi-item scales were reliable: Cronbach's alpha was .88 for perceived usefulness and .91 for intention, above the .70 standard and below the very high values that suggest redundant items (Tavakol & Dennick, 2011). On the five-point scales, the usefulness mean was 3.62 with a standard deviation of 0.91, and the intention mean was 2.94 with a standard deviation of 1.08. Just over half the sample, 129 members, reported recent home video workouts, and 71, or three in ten, reported travel or shift schedules. Strength training was the most wanted on-demand class type, ranked first by 38 percent of respondents, followed by cycling at 27 percent and yoga and mobility at 19 percent. Far fewer put interval training or core work first. Because many members do not own a stationary bike, the strong showing for cycling suggests either that a meaningful group trains at home on their own bikes or that some members ranked their favorite live class rather than what they would actually do at home, a question the pilot can answer.

Hypothesis Tests

Usefulness and intention rose together. Intention averaged 3.21 among members who watch workout videos and 2.62 among those who do not, and 3.38 among members with travel or shift work compared with 2.75 among everyone else, with standard deviations near 1.0 in every group. Judged by the benchmarks Cohen (1992) proposed, both group differences are medium-sized.

Multiple Regression

The hypothesis tests examine each factor alone, but the factors overlap; video users, for instance, may also be the members who find the library useful. Malhotra (2015) presents multiple regression as the tool for isolating what each predictor adds once the remaining predictors are fixed. Intention was regressed on perceived usefulness, video use (1 = yes) and schedule constraint (1 = yes).

The model explained 45 percent of the variance in intention, R squared = .45, F(3, 233) = 63.55, p < .001. All three predictors remained significant. Perceived usefulness was by far the strongest: each one-point rise on the usefulness scale went with a 0.64-point rise in intention, holding the other factors constant. Schedule-constrained members scored 0.41 points higher in intention than otherwise similar members, so the studio's most promising buyers can be identified from its own records. Video use added a smaller but real increase. The regression also clarifies the earlier t-tests: part of the difference between video users and non-users reflects the fact that video users find the library more useful, so once usefulness is held constant, the video-use effect shrinks from about 0.59 points to 0.27. Schedule constraint keeps more of its effect, which suggests that members who cannot attend want the library for practical reasons that go beyond liking the idea.

Willingness to Pay by Schedule

Slightly more than a third of the sample, 88 members, named a price of $15 or higher. That share climbed past half, 38 of 71, for members with travel or shift work, against 50 of 166 for the rest, a gap the chi-square statistic of 11.67 shows to be far larger than chance would produce (p < .001).

Forecast

Stated intentions predict purchases imperfectly and usually overstate them, but they are informative when adjusted and when the people measured are close to the decision (Morwitz et al., 2007). The forecast therefore takes three cautious steps:

• Count only firm prospects, the 49 members, about a fifth of the sample, whose intention score reached 4 and whose price reached $15.
• Assume the 403 nonrespondents are half as likely to be firm prospects, 10.3 percent, which adds about 42 members, for about 91 firm prospects in all.
• Assume half of firm prospects actually subscribe, which gives about 45 subscribers.

At $15 a month, 45 subscribers would bring in about $8,100 a year, well short of the $18,000 first-year cost.

The conversion assumption carries the most uncertainty, so the forecast was repeated at three rates:

Even at the most optimistic rate, subscriptions alone fall short of the full first-year cost. The decision therefore turns on two things the survey cannot measure: how many members would actually pay once the library exists, and whether it would keep members from canceling.

The library has a second source of value, however. If it keeps 10 schedule-constrained members a year from canceling, a modest share of the roughly 60 who cancel for schedule reasons each year, the studio keeps 10 memberships worth $89 a month, about $10,680 a year. Subscription revenue and retained memberships together, about $18,780, would roughly cover the first-year cost, but only if the retention effect is real, which this survey cannot show.

Conclusions and Recommendations

The findings fit the theory the study began with. The technology acceptance model predicts that perceived usefulness is the main driver of intention to adopt a new technology (Davis, 1989), and in this study it was by far the strongest predictor. The unified theory's emphasis on experience and facilitating conditions (Venkatesh et al., 2003) also fits: prior video use, a form of experience, and a schedule that makes home training practical both raised intention. A small local service behaved much as the broader research predicts.

The research supports three conclusions. First, interest exists and is concentrated: members whose schedules keep them away, and those already comfortable with workout videos, are the natural market. Second, perceived usefulness drives intention more than any other factor, so the offer must make plain how the library helps members keep training when they cannot attend. Third, subscriptions alone will not pay for a full library in the first year; the case depends on retention, which must be tested.

The report recommends a staged decision:

• Run a 90-day pilot with about 30 recorded classes, emphasizing strength and cycling, filmed with existing equipment at a cost of about $4,500.
• Offer it first, at $15 a month, to schedule-constrained members and video users, identified from the booking records and the survey.
• Build the full library only if at least 60 members, about 9 percent of the membership, are paying by the end of the pilot.
• Track cancellations among pilot subscribers against similar non-subscribers to measure the retention effect directly.

The findings also shape how the offer should be promoted. Because perceived usefulness is the strongest driver of intention, messages should show concretely how the library helps, for example by naming the travel weeks and early shifts it covers, rather than describing production quality. Members who already use workout videos need little explanation, while those who do not may need a free first week to experience the format. The stated prices support $15 as a ceiling rather than a starting point; if the pilot falls short of 60 subscribers, a $12 price or a bundle with membership upgrades should be tested before the idea is dropped.

Limitations

The study measured stated intentions and prices, not behavior, and the forecast's conversion assumptions are judgments, not findings. The early-late comparison reduces but does not remove concern about nonresponse. Perceived usefulness and intention were measured in the same survey, which can inflate their correlation. The regression's explained variance of 45 percent should be read with that in mind, since part of it may reflect members answering related questions consistently rather than a true connection between the two. Finally, the study describes one studio's members at one time; the results may not hold at other studios or after a competitor launches its own service. The class preference rankings, in particular, reflect what members imagine they would do at home, which the pilot's viewing data will test against what they actually watch.

A Biblical Perspective

Proverbs 14:15 sets the gullible person, who swallows whatever is said, beside the shrewd one, who watches where the next step will land (English Standard Version Bible, 2001/2016). Prudence in this project meant not believing every stated intention, but discounting it, testing it and planning a step that can be reversed. Proverbs 16:9 adds that people may map out a route while God directs where their feet actually land. Research does not remove uncertainty; it helps a business plan humbly within it. A pilot that commits a modest sum, measures real behavior and then decides honors both proverbs: it plans carefully and holds the outcome loosely. The research also served the members themselves, since it asked what would help them keep training when life pulls them away, and a decision built on their answers treats them as people to serve rather than revenue to capture.

References

Armstrong, J. S., & Overton, T. S. (1977). Estimating nonresponse bias in mail surveys. Journal of Marketing Research, 14(3), 396-402. https://doi.org/10.1177/002224377701400320

Churchill, G. A., Jr. (1979). A paradigm for developing better measures of marketing constructs. Journal of Marketing Research, 16(1), 64-73. https://doi.org/10.1177/002224377901600110

Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155-159. https://doi.org/10.1037/0033-2909.112.1.155

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. https://doi.org/10.2307/249008

English Standard Version Bible. (2016). Crossway. (Original work published 2001)

Groves, R. M., & Peytcheva, E. (2008). The impact of nonresponse rates on nonresponse bias: A meta-analysis. Public Opinion Quarterly, 72(2), 167-189. https://doi.org/10.1093/poq/nfn011

Krosnick, J. A. (1991). Response strategies for coping with the cognitive demands of attitude measures in surveys. Applied Cognitive Psychology, 5(3), 213-236. https://doi.org/10.1002/acp.2350050305

Malhotra, N. K. (2015). Essentials of marketing research: A hands-on orientation. Pearson.

Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in survey data. Psychological Methods, 17(3), 437-455. https://doi.org/10.1037/a0028085

Miller, K. M., Hofstetter, R., Krohmer, H., & Zhang, Z. J. (2011). How should consumers' willingness to pay be measured? An empirical comparison of state-of-the-art approaches. Journal of Marketing Research, 48(1), 172-184. https://doi.org/10.1509/jmkr.48.1.172

Morwitz, V. G., Steckel, J. H., & Gupta, A. (2007). When do purchase intentions predict sales? International Journal of Forecasting, 23(3), 347-364. https://doi.org/10.1016/j.ijforecast.2007.05.015

Norman, G. (2010). Likert scales, levels of measurement and the "laws" of statistics. Advances in Health Sciences Education, 15(5), 625-632. https://doi.org/10.1007/s10459-010-9222-y

Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach's alpha. International Journal of Medical Education, 2, 53-55. https://doi.org/10.5116/ijme.4dfb.8dfd

Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425-478. https://doi.org/10.2307/30036540