Marketing Principle #3 All Competitors React Managing Relationship-based Sustainable Competitive Advantage Agenda
Download 1.79 Mb.
|
- Navigate this page:
- Key Relationship Dimensions and Example Measures
Measuring Relationship Equity
Key Relationship Dimensions and Example MeasuresMultivariate Regression Analysis Multivariate regression is a statistical approach used to quantify the sign and magnitude of the relationship between a focal dependent variable (marketing outcome in our context) and several independent variables (e.g., marketing efforts). DAT 7.1 Description
When to Use It How it Works The purpose of multivariate regression is to capture the statistical association between a focal marketing outcome of interest (e.g., sales, loyalty, CLV, profitability) and several marketing interventions that simultaneously may affect the focal outcome (e.g., relationship marketing efforts, marketing mix). Performing a multivariate regression enables five important discoveries. First, we can discern whether a particular marketing intervention truly influences a marketing outcome. That is, a multivariate regression can provide statistical validation of the significance of the impact of a certain marketing intervention. Second, we learn the sign of the relationship between a marketing invention and a marketing outcome. In some cases, the sign is well-known a priori (e.g., as the price increases, sales decrease), but in others, it remains unclear. For example, a firm may not know whether a financially oriented relationship marketing program offering free shipping ultimately increases CLV. The regression can help the firm verify the sign of the relationship. Third, a multivariate regression helps researchers compare the relative strength of multiple marketing interventions. For example, a firm may need to know which of its social, structural, or financial relationship marketing efforts are most and least influential, and this determination is enabled by a regression. Fourth, with a multivariate regression, we can control for confounds while gauging the relationship between marketing interventions and marketing outcomes. For example, while trying to understand the relationship between financial relationship marketing efforts by a supplier firm devoted to a buyer firm and marketing outcomes earned from this buyer firm, we might control for the buyer firm’s size, because larger firms typically buy more, regardless of whether they receive marketing interventions. Fifth and finally, multivariate regression enables predictions of the marketing outcomes following from various scenarios of marketing interventions, which is useful in scenario analysis. If the marketing outcome is given by Y, and we have three marketing interventions (X1, X2, and X3), and two confounds (Z1, and Z2), the formula is given by: where to are the coefficients (or weights) that capture the sign and strength of the relationship between the marketing interventions and the marketing outcome, and is a random error term. In most cases, we would have data about past outcomes and marketing inventions/confounds, then rely on software such as SAS or SPSS to provide the sign, strength, and statistical significance of the coefficients. A B2B supplier of electrical equipment is going through a redesign of its relationship marketing efforts directed at buyers, and it seeks to ensure that it is investing in RM efforts that boost the CLV of each of its buyers. Currently, the supplier is investing in three kinds of RM efforts: social (e.g., meals, sporting events), structural (e.g., customized packaging), and financial (e.g., free giveaways of small electrical parts that are part of the electrical installation service it provides). To perform this exercise, the supplier created a database of the CLV of its 3,500 buyers, as well its investments in social RM (SRM), structural RM (StRM), and financial RM (FRM) for each of these buyers. It also collected data on the buyers’ location (east or west coast), the number of employees in the buyer firm, and the firm industry type (corporate or government). The results showed:
Based on the analysis, the supplier also launched another study, to understand why its structural RM efforts were not successful. Download 1.79 Mb. Share with your friends:
|