Some marketing science case studies

Some marketing science case studies

Analytics is not software. Software is not analytics. Software and analytics do not make management decisions. They are tools that help managers make decisions.

Many analytic tools are now available in user-friendly software packages. User friendly doesn’t mean easy to use skillfully, however. That requires formal education in research methods and statistics and a lot of experience applying this classroom learning to real-world problems.

Of course, technical proficiency by itself will not make one a competent marketing scientist. Marketing science is social science. It calls upon diverse skills, such as being able to partner with marketers to help them see the big picture and anticipate key decisions they’ll have to make.

Successful digital transformation is a matter of know how and access to the best talent. We connect you to both.Click for more.

Let’s briefly sum up my philosophy. The purpose of marketing science is to enhance decisions. We first need to make sure we’re tackling the right marketing issues. Research design, data and analytic methods must be aligned with these concerns. Finally, analysis requires thought, imagination and judgment as well as attention to detail. It is not a series of assembly line tasks.

Marketing science is a way of thinking.

Here are a few case studies as a way of illustrating my views on this.

Consumer Segmentation

Challenges: A major manufacturer of personal care products felt their product portfolio needed to be revamped. Some brands in their lineup overlapped and their respective positionings were unclear. There also had been introductions by competitors and the category was extending in new directions. Hence the client required insights for R&D, in addition.

Solutions: Consumer Segmentation was proposed as one element of a research program, which included qualitative phases (undertaken by another agency) before and subsequent to a large scale quantitative phase. The client’s previous experience with consumer segmentation had not been positive, however. While fairly clear attitudinal segments had emerged these segments did not differentiate target demographic groups well, and brand usage was similar across the segments. Reflecting these concerns, full-profile segmentation was used for this project.

Outcomes: Six segments were identified. The segments were well defined in terms of benefits sought, demographics and brand usage. Based on the results of the research, the decision was made to de-list one brand and slightly reposition the client’s flagship brand. Several new opportunities for new product introduction have been explored.

Market Response Modeling

Challenges: Businesses have often relied on gut feel and historical precedent when developing marketing plans. Many companies, however, are now beginning to incorporate Market Response Modeling (MRM) into these decisions. MRM uses historical marketing and sales data to measure sales “bang” for marketing “buck”. One such company, a multinational FMCG client, lacked the internal resources to conduct MRM and contacted a marketing research agency for assistance. The agency, in turn, subcontracted the project to Cannon Gray.

Solutions: The client provided four years of sales and marketing figures. Cannon Gray used these data plus economic and other “macro” data in econometric time series analysis to measure the impact of marketing inputs on sales. The modeling included competitor activity. Based on the results, “What if?” simulations were also conducted under various marketing scenarios.

Outcomes: Our findings indicated the client had been allocating too much budget to TV advertising and not enough to in-store promotional activities. They also suggested the client should follow a slightly more aggressive pricing strategy and rationalize their social media marketing. One of the “What if?” scenarios was later used as the basis for a new marketing plan.

Successful digital transformation is a matter of knowledge and access to the best talent. We connect you to both.Click for more.

Data Mining and Predictive Analytics

Challenges: A Financial Services company planned to diversify and expand its menu of retirement savings vehicles. It also wished to better understand how to effectively market new products to current and potential customers. Previous research had shed some light on these issues but specific direction on next steps was lacking. The company retained Cannon Gray as an external consultant.

Solutions: Extensive analyses of customer records was performed with the assistance of Cannon Gray and revealed a number of associations between customer characteristics and their historical savings and investment behavior. It was suspected, however, that some of these patterns resulted from the company’s own marketing activity (and perhaps that of competitors).

Many questions remained as to why customers were “doing what they were doing” and what marketing strategies would be most effective with which kinds of customers. In-depth interviews and focus group discussions conducted by a marketing research agency provided further insights and a follow-up quantitative phase was undertaken. Both qualitative and quantitative phases included consumers who were not current customers of the client. For the customers we interviewed data from customer records were fused with their survey responses and all data used in a variety of advanced analytics.

Outcomes: One key finding was that risk acceptance/aversion, general financial sophistication and other attitudes predicted interest in new investment vehicle concepts beyond that explained by life stage, demographics and historical behavior. In addition, predictive models and marketing communication tailored to individual consumers were developed based on their predicted interest in different concepts. One new product has been successfully launched and several others are under development.

Customer Satisfaction

Challenges: With the assistance of a marketing research agency, a bank had been conducting satisfaction (CS) tracking surveys of its retail customers for several years. The survey covered the main consumer channels (e.g., Call Centers, Internet Banking) for their largest branches. Customers were asked to rate the bank and its staff on approximately two dozen attributes such as courtesy, speed of response and ease of transactions. The data had not been analyzed beyond the descriptive level, however, and management wanted to identify the main drivers of overall satisfaction with the banking experience.

Solutions: The agency hired Cannon Gray to conduct Key Driver Analysis. CS data from the past year were used and Structural Equation Modeling (SEM) employed to derive the relative importance of the attributes. Because the nature of customers’ interaction with the bank was different for each channel several versions of the questionnaire had been used. Accordingly, separate driver models were required.

Outcomes: The SEM modeling helped management better understand the priorities of their customers.  We matched customers’ priorities against their satisfaction with the bank’s performance to see where gaps and opportunities lay. Based on the results of this research, modifications in staff training were made and some transaction procedures streamlined.

Successful digital transformation is a matter of know how and access to the best talent. We connect you to both.Click for more.

Pricing

Challenges: Faced with downward price pressure in one of its largest markets, a global manufacturer of personal care products decided it needed to rethink its pricing strategy. It put out an RFP on pricing research to several marketing research agencies. Cannon Gray was asked by one of these agencies to assist with proposal development and, if the bid was won, perform the necessary analyses. We recommended Discrete Choice Modeling as the core analytic method.

Solutions: The project was commissioned and 500 current users of the category interviewed. The centrepiece of the interview was a choice experiment in which respondents were shown existing brands at various price levels and asked which, if any, they would purchase. The results were statistically analyzed with a method known as Hierarchical Bayes to obtain the relative preferences of the brands and their price elasticities. “Utilities” (importance scores) for brand and price were calculated for each respondent individually and various “What if?” simulations were conducted among respondent subgroups of interest.

Outcomes: Results of the modeling and simulations provided guidance on how far the client would be able to reduce prices before profitability would begin to erode. Importantly, brand utilities and price elasticities were found to vary by consumer type, in some cases markedly. Another benefit of the research was that it helped the client refine their general thinking about the category.

These are just a few examples, with client-sensitive details removed.

Browse

Article by channel:

Read more articles tagged: Featured, Marketing Analytics

Analytics