Lainez, J. M., Reklaitis, G. V.,

From Digital Culture & Society

<Linking marketing and supply chain models for improved business strategic decision support:

Laínez, J. M., Reklaitis, G. V., & Puigjaner, L. (2010). Linking marketing and supply chain models for improved business strategic decision support. Computers And Chemical Engineering, 34 (12), (10th International Symposium on Process Systems Engineering, Salvador, Bahia, Brasil, 16-20 August 2009), 2107-2117.

To operate successfully, marketing activities must be coordinated with other corporate functional areas. Developing better business policies that increase efficiency and reduce production costs increase profit margins, increases an organization’s competitive edge in the global market. A supply and demand chain model which incorporates strategic marketing decisions is crucial for large multinationals looking to increase their competitive advantage. Specifically, marketing professionals should evaluate the trade-off between promotion and supply chain decisions to increase overall sales and reduce costs. Recently, prominent multinational corporations have made significant progress in developing machine learning softwares and powerful supercomputers that can model quantitatively based marketing decisions.

In their article Linking marketing and supply chain models for improved business strategic decision support, authors Laínez, Reklaitis, and Puigjaner, build on these developments to formulate a mathematical model that accounts for the main relevant business functionalities. The authors create a mathematical equation to model all relevant variables and support managers in the decision making associated with supply chain and marketing decisions. To assist marketing managers, the model provides recommendations managers regarding product pricing, the amount of advertising expenditure, and its most appropriate allocation. Further assisting marketing managers, the equation models the relationship between pricing, advertising, and demand, in order to accurately forecast how equipment capacity should be optimally expanded according to supply chain demand. Additionally, production and distribution planning is performed by the equation more efficiently than by its human counterpart, leading to more accurate production output relevant to market demand. The equation is designed so that all core business decisions are anticipated according to a predefined planning horizon, and all decisions yield the ‘best’ possible results, which is interpreted as maximizing supply chain efficiency and return on promotional marketing material. The parameters of the equation define what the machine will learn to consider ‘best’. The result is a mixed integer nonlinear programming formulation which optimizes the supply chain and marketing strategic decisions in an integrated fashion.

The traditional business strategy of a large multinational corporation is modeled as a hierarchical process where functional strategies, operations, logistics, marketing, and finance are driven by higher level strategy. “A key element of the strategic framework involves coordinating functional level plans to work in concert so as to achieve the overall business strategy rather than to locally optimize outcomes for individual functions, business units, plants, or distribution centers.” When problems relating to all previously stated elements of business management are formulated as a mixed integer nonlinear program (MINLP), these variables and constraints can be classified into four groups: the formulation of marketing activities, supply chain design and operations, constraints related to the financial formulation, and the integration of operations and marketing decisions.

According to the authors, marketing models can be seen as an opportunity achieve improvement in the business bottom line. Marketing now resembles design engineering because of its capacity to mine data, build models, conduct real time analyses, and facilitate computer simulations to design effective marketing plans. Access to personal information and sensitive customer data used to create psychographic profiles is now easily obtained, and large multinationals often find they have more data than they are able to process. Implementing a marketing decision making model named BRANDAID, the MINLP searches through backlogs of customer data and returns actionable intelligence marketers can use to create more effective messaging, relevant to the target customer.

BRANDAID is a model for assembling the elements of the marketing mix to describe the market and evaluate multiple strategies simultaneously. The BRANDAID model is modular, so marketers can determine which sub-models to take into account according to their specific needs. This flexible model is comprised of three sub-models: the advertising sub-model, the pricing sub-model, and the sales force sub-model.

The Advertising sub-model assumes that there is a fixed advertising rate which maintains predetermined sales target. When initializing the model, If a brand starts out with its sales rate at its target and marketing conditions reporting less that optimal values, then the advertising rate will maintain sales at the reference values. This advertising will be designated as the maintenance or baseline advertising rate. If advertising is less than reference, the sales rate will presumably sag, and, after a while, level off at a new lower value. Price is a sensitive control variable frequently referenced by Marketing managers. The price under consideration is the basic price per unit charged by the firm, accounting for all costs associated with production and distribution. Price elasticity of demand is considered for this sub-model. Sales force is another sensitive control variable frequently referenced by Marketing Managers. The size, compensation model, and resources allocated for sales accounts for all costs associated with the facilitation and administration of sales related activities. Other business parameters taken into account by the general model include; Pricing, Supply Chain Design Planning, Mass Balance Constraints, Production Sites, Distribution Centres, Marketplaces, Capacity and Facilities Location Restraints, Capacity Expansion, Capacity Utilization, Supplier Limitations, Financial Formulation, and Integration Among Models.

Laínez, Reklaitis, and Puigjaner, present a MINLP model which simultaneously considers supply chain design and retrofitting, and financial and marketing decisions, and demonstrate the economic benefits that this new holistic approach may provide. The authors conclude by bringing attention to the potential of response surface and data mining techniques in this field. By taking advantage of these new data mining techniques, descriptive data driven models for the marketing activities can be obtained and recycled later. Further work should focus on applying decomposition techniques so that more complicated marketing models may be computed in more acceptable computation times in the future.

Warren Buzanko 5750021 >