You are on page 1of 90

Marketing Optimization Running heading: MARKETING OPTIMIZATION

Marketing Plan Optimization: A Quantitative Approach to Resource Allocation

Scott D. Graffius Cardinal Stritch University

A Capstone Research Product submitted in partial fulfillment of the requirements for the degree of Master of Business Administration March, 2009

Marketing Optimization Approval Page

I approve the Capstone Research Product of Scott D. Graffius:

Tom Bramorski, Ph.D, Instructor

Date

Marketing Optimization Abstract While a growing number of companies are embracing analytical metrics to measure the performance of their critical business processes, many neglect to apply the

same mathematical rigor to their marketing efforts. This paper examines how to apply a disciplined quantitative approach to resource allocation within the context of marketing. Managements objective is to construct a marketing plan that will maximize both total revenue as well as revenue from a targeted geographic region, while working within the budgetary and other constraints. Operations Research offers sophisticated tools for translating complex business situations into mathematical expressions. These expressions may then be evaluated to identify one or more optimal solutions. Multivariate Testing is used to quantify all the relevant variables for the mathematical model. This statistical tool allows the researcher to simultaneously test the effect of various input factors on the response variable, and also test for any interactions between factors. These interactions are often significant in predicting the response outcome. Using indexed utility values produces a blended model that can simultaneously satisfy both objectives. This blended model results in a marketing plan that will increase total revenue by more than five million dollars, a 57% gain. At the same time, the plan will generate an 18% increase in targeted revenue, positioning the company for future growth. The project shows that companies can, and should, insist on the same degree of financial accountability from their marketing investments that they would expect from any other capital or operational expenditures.

Marketing Optimization Table of Contents

Page Approval Page...................................................................................................................2 Abstract...............................................................................................................................3 Table of Contents ..............................................................................................................4 List of Tables ......................................................................................................................7 List of Figures (when applicable) ...................................................................................8 Chapter 1 Introduction .............................................................................................9 Overview of the Research Project ................................................................9 Organizational and Environmental Profiles ...............................................9 Description of the Organization ......................................................9 Description of the Unit ..................................................................10 Research Goals and Objectives ..................................................................11 Background of the Business Problem.............................................11 Business Problem and Opportunity Statement.............................13 Scope and Limitations....................................................................13 Conclusion .................................................................................................13 Chapter 2 Literature Review....................................................................................15 Competitive Strategy & Game Theory ......................................................15 Statistics & Business Intelligence .............................................................17 Marketing Analytics..................................................................................19

Marketing Optimization Operations Research & Utility Theory......................................................21 Conclusion .................................................................................................23 Chapter 3 Method .....................................................................................................25 Statement of Purpose .................................................................................25 Research Question .....................................................................................25 Alternatives....................................................................................25 Criteria...........................................................................................27 Data Collection ..........................................................................................29 Data Collection Procedures ...........................................................30 Analytical Tools and Tests ............................................................31 Conclusion .................................................................................................33 Chapter 4 Results and Analysis...............................................................................34 Findings .....................................................................................................34 Implication for the Research Question ......................................................43 Chapter 5 Conclusion ..............................................................................................47 Recommendations .....................................................................................47 Implementation .........................................................................................49 References .......................................................................................................................51 Appendix A Historical Lead Generation Values .....................................................53 Appendix B Complete Statistical Software Output for Total Leads.....................57 Appendix C Complete Statistical Software Output for Targeted Leads ..............58 Appendix D Table of Discreet Factor Coefficients...................................................59

Marketing Optimization Appendix E Multivariate Testing Results for Total Lead Generation..................61 Appendix F Multivariate Testing Results for Targeted Lead Generation ...........65 Appendix G Indexed Values for Total Lead Generation ........................................69 Appendix H Indexed Values for Targeted Lead Generation .................................73 Appendix I Blended Final Utility Values ................................................................77 Appendix J Saturation Point Analysis .....................................................................81 Appendix K Linear Programming Constraint Values.............................................82 Appendix L Marketing Plan Summary Tables for All Four Models ....................84 Appendix M Reference Table of Constraint Satisfaction Values............................86 Appendix N Table of Costs, Revenues and Efficiencies for All Four Models......88

Marketing Optimization List of Tables

Table 1 2 3 4 5 6

Title Generator Products by Size and List Price, 2009 Vertical Market Sectors with NAICS Codes, 2009 Advertising Channels and Cost Per Unit, 2009 List of Product, Market and Channel Mix Constraints, 2009 List of Saturation Constraints, 2009 Final Marketing Plan from the Blended Utility Model, 2009

Page 27 27 28 41 43 46

Marketing Optimization List of Figures Figure 1 2 Title Statistical Software Output for Multivariate Testing, 2009 Saturation Point Examples, 2009 Page 36 42

Marketing Optimization Marketing Plan Optimization: A Quantitative Approach to Resource Allocation Chapter 1: Introduction Overview of the Research Project Mainstream business thinking has come to embrace the vital importance of

analytics, often referred to as "business intelligence" (Williams & Williams, 2007, pg. 97). Yet many organizations have not applied the same scientific approach to their marketing plans. This research paper will attempt to show how businesses that utilize quantitative analytical techniques in developing their marketing strategies will be able to achieve a competitive advantage over those that simply rely on anecdotal evidence and speculation. These heuristic methods of decision-making yield sub-optimal results, because the most advantageous solutions are often counterintuitive (particularly in marketing applications). This project is an extensive case study of how a business can use techniques from the field of Operations Research to optimize a marketing plan, and measure its performance against predetermined strategic goals. Organizational and Environmental Profiles The marketing strategies and tactics vary significantly between companies who sell directly to consumers and those who market their products to other companies, to institutions, or the government (Kotler & Keller, 2006, pg. 210). Here the author has chosen to illustrate quantitative marketing decision techniques in a business-to-business setting, but these same techniques can be applied to any marketing model. Description of the Organization. This project will examine a company that

Marketing Optimization manufactures and sells electric power generation components. This firms executive management believes that developing nations in Asia have a particularly strong need

10

for power generation equipment as they build and expand new infrastructure at a rapid pace. In less than 20 years, rich industrial countries will be overshadowed by developing countries in their produced share of the worlds output (Pearce & Robinson, 2009, pg. 130). The vast majority of sales currently come from North America, but the company believes that global diversification will help mitigate the risk of regional economic shifts and position the company for future growth. With this in mind, one of the goals for the firms marketing plan is to maximize sales in this targeted geographic region, while at the same time maintaining strong total revenue performance. Description of the Unit. Marketing departments are continually challenged with allocating limited financial resources in order to maximize revenue. Surprisingly, many companies fail to measure the efficacy of these investments, and fewer still use quantitative criteria for deciding how to allocate their marketing budgets. This firms Director of Marketing is concerned with promoting five different generator products, each with its own unique features and applications. The Director of Marketing has identified five specific vertical markets that have strong historical demand for these generator products, as well as a positive outlook for continued growth. Within each market, the firm must choose from among five different channels, the term marketers use to describe the promotional vehicles used to communicate with potential customers (Kotler & Keller, 2006, pg. 472). Each of the five markets has a

Marketing Optimization

11

variety of targeted publications, tradeshows, and other industry-related media channels that provide opportunities to advertise the firms products. Research Goals and Objectives The firms previous marketing plans were based purely on anecdotal evidence and speculation. Although they produced adequately successful results, executive management wants to determine if a more scientific approach will yield higher levels of total revenue as well as targeted revenue growth. To that end, this paper will draw on techniques from the field of Applied Mathematics. While these tools have allowed the business community to achieve breakthrough improvements in operational efficiency, applying them within the context marketing is a relatively new concept (Burnham, 2004, pg. 32). Background of the Business Problem. Executive management has determined that its primary objective is to maximize the total revenue generated by the marketing plan. At the same time, however, management has a secondary goal it wants to take into consideration: targeted revenue from a specific geographic region (namely, Asia). These two goals are conflicted; a marketing plan optimized solely for revenue maximization would be sub-optimal in satisfying the geographical objective. Conversely, a marketing plan optimized for the targeted region would not yield maximum total revenue. The marketing team has been tasked with the challenge of finding a way to quantify, and then simultaneously address, both of these objectives. Business Problem or Opportunity Statement. In preparing a marketing plan for the coming year, the Director of Marketing must decide where to allocate limited financial

Marketing Optimization resources and which marketing opportunities to pursue. Each marketing option has a specific utility value related to both of the strategic goals, and a cost to execute that option. Executive management has set a total budget limitation, and there are a finite number of possible options from which to choose. Ultimately, the question facing the

12

Director of Marketing is, What marketing plan will generate the best possible combination of both strategic goals, while working within the overall budget constraint? The first of three stages in finding an optimal solution to this problem will utilize a complex statistical tool called multivariate testing (also know as MVT). MVT offers the ability to test combinations of factors instead of the traditional one-factor-at-a-time method of experimenting (Burnham, 2004, pg. 34). These traditional methods ignore the interactions between factors, which are often vital to the response. This procedure is relatively new to the field of marketing, but has already yielded significant results. MVT allows a business to understand, and even predict, the response to any possible combination of products, markets and channels employed in the campaign. This insight alone is incredibly powerful, but becomes even more so when used together with tools from stages two and three. After MVT has been used to determine coefficients for each variable, the second stage of the problem will then require the use of a indexed decision table from the field of Utility Theory (Lawrence & Pasternack, 2002, pg. 363). In this case, executive management has two unequal simultaneous objectives: maximization of total revenue and maximization of targeted revenue. The MVT analysis will produce two sets of coefficients (one for each of the two objectives). These two coefficients will be combined

Marketing Optimization into one numeric value by indexing them according to their combined relative efficiency, resulting in a single utility value for each marketing option. The values are

13

indexed in order to make relative comparisons among a diverse group of values (Ling, Marchal, & Wathen, 2005, pg. 622). Now, with only one single coefficient of utility for each possible marketing option, the problem may now be solved using Combinatorial Optimization. Combinatorial Optimization is a branch of Applied Mathematics concerned with finding the best possible solution from among a set of discrete feasible alternatives (Lawrence & Pasternack, 2002, pg. 57). The solution itself is achieved by modeling the problem as an integer linear programming problem to allocate the budget for maximum desirability within the given constraints. The optimized solution will then be examined for its total expected revenue, as well as its expected revenue from the targeted region. Scope and Limitations. While this project addresses the allocation of marketing budgets, the same quantitative approach can be used more broadly for other budgeting and resource allocation problems. Traditionally, however, many organizations already utilize sophisticated techniques for other capital and operational budgeting decisions, and the marketing department is one of the last to specifically measure the financial performance of its investments. Conclusion While most large organizations use metrics to measure and analyze business performance, few have applied the same quantitative approach to their marketing efforts. This paper will illustrate the use of techniques from the field of Applied

Marketing Optimization Mathematics (such as multivariate testing, indexed decision tables and Combinatorial Optimization) to achieve the best possible marketing plan. The marketing plan will specifically address where and how to allocate financial resources to achieve the best possible results. Executive management has set an arbitrary constraint on the total marketing budget, as well as specific minimum requirements to maintain a relatively

14

balanced product/market offering. This paper will also explore methods for calculating saturation points and using them as additional constraints in the linear model. The final solution will be measured against the previous heuristic approach for total revenue and targeted revenue generation.

Marketing Optimization Chapter 2: Literature Review Competitive Strategy & Game Theory

15

The modern study of business management assumes an economic environment of capitalism, which is built on the premise of competition. Thus, executive managers must have a keen awareness of competitive strategy, because it provides a framework for thinking proactively about threats to a firms business model, new business opportunities, and about the future reconfigurations of the firms resources, capabilities, and core competencies (McGuigan, Moyer, & Harris, 2005, pg. 426). In 1980, Michael Porter published his now classic work on the subject of competitive strategy, becoming among the first visionaries to provide a comprehensive analytical basis for managers to utilize. Porter describes the competitive rivalry that inevitably occurs as each firm seeks to maximize its own returns at the expense of others. In most industries, competitive moves by one firm have noticeable effects on its competitors and thus may incite retaliation or efforts to counter the move, that is, firms are mutually dependent (Porter, 1980, pg. 17). This behavior is especially typical of markets that are dominated by only a few large players, which Economists refer to as Oligopolies (McGuigan, Moyer, & Harris, 2005, pg. 507). In addition to the competitive forces of familiar rivals, business leaders must also be on constant alert for what Clayton Christensen refers to as disruptive technologies. Christensen further extended the body of research by attempting to explain why firms that could be esteemed as aggressive, innovative, customer-sensitive organizations could rapidly fail by missing critical new developments in their industries.

Marketing Optimization While Porter and Christensen have provided a qualitative understanding of competitive strategy, a more quantitative approach has developed in a branch of Applied Mathematics known as Game Theory. Pioneered in the 1940s and 50s by the

16

likes of Jon Von Neumann and John Forbes Nash, Game Theory studies the behavior of individuals and organizations by relating real situations from life and business to simple, mathematical games (McGuigan, Moyer, & Harris, 2005, pg. 549). This allows researchers to simulate and evaluate the various alternative strategies available to each player. Just as in a real business environment, players must make every move with an awareness of the potential reactions by competitors. The most successful players in the competitive game of business will be able to predict the moves of opponents, according to their most rational options, and to develop effective counter-moves (Fundenberg, & Triole, 1991, pg. 8). In the field of Game theory, the function that we are interested in is the payoff (or utility) function of a player. The independent variable in such a function is the players strategy. Often we look for the best such strategy; that is, we look for the strategy choice that makes the value of the payoff function greatest (Dutta, 1999, pg. 409). This paper explores the development of competitive business strategy, specifically in the context of marketing. The word marketing, however, is generally understood in very broad terms, encompassing any activities that create demand for the firms products. This may include product positioning, creating advertising/branding

Marketing Optimization

17

campaigns, developing pricing policies, assessing customer needs, and identifying new opportunities for business development (Kotler & Keller, 2006, pg. 5). Differences in competitive structure and in customer preferences between [market] segments results in different key success factors (Grant, 2002, pg. 122). Central to the success of any business is a clear understanding of its customers. In recent years, the business community has moved beyond the limited mentality of transactions, and instead begun to focus on strengthening customer relationships. If a customer develops a relationship with an enterprise, the customer tends more and more to trust the enterprise to act in the customers own interest (Peppers & Rogers, 2004, pg. 37). This deepening loyalty is a source of tremendous potential growth through repeat business. Each loyal customer typically reduces a companys customer acquisition costs (Reichheld, 2003, pg. 53). Reichheld defines this loyalty as more than repeat business alone, but a willingness to recommend the companys products and services to a friend or colleague. The very survival of any business necessitates an obvious need for customers, so a great many companies now include statements about the importance of customer needs right in their mission statement. A focus on customer satisfaction causes managers to realize the importance of providing quality customer service (Pearce & Robinson, 2009, pg. 35). Statistics & Business Intelligence Many organizations are beginning to employ advanced quantitative analytics to build predictive models that drive their strategy development and operational decisions. We have found that a growing number of companies that have recognized

Marketing Optimization the power of leveraging data-driven insights through the use of business intelligence (Davenport & Harris, 2007, pg. 32). When implemented successfully, these tools can benefit both the supplier and the customer, allowing the firm to adjust business

18

processes to provide high levels of service to [the] most profitable customers to ensure retention, thereby increasing profitability (Williams & Williams, 2007, pg. 6). One statistical tool that is beginning to deliver tremendous value is multivariate testing (also know as MVT). MVT offers the ability to test combinations of factors instead of the traditional one-factor-at-a-time method of experimenting (Burnham, 2004, pg. 34). This insight adds a significant amount of depth and increased accuracy to the models that predict response. The MVT process can be used to improve almost anything. The objective of a project may be to increase sales, increase production, raise service levels, or improve product performance (Holland, 2005, pg. 57). These exercises in data analysis involve a series of highly complex mathematical computations. Fortunately, this type of problem can be quickly solved using one of the many statistical software packages and spreadsheet packages available for the computer (Lind, Marchal, & Wathen, 2005, pg. 518). For this reason, the widespread use of statistical techniques in business applications has been relatively recent, accompanying the advent of personal computers. Business executives are increasingly utilizing statistics because they believe that it can improve decision-making within their organization. A robust decision is the best possible choice, one found by eliminating all the uncertainty possible within available resources, and then choosing with known and acceptable levels of satisfaction and risk

Marketing Optimization (Ullman, 2006, pg. 35). According to authors Donald Cooper and Pamela Schindler, executive managers have increasingly high expectations of their business research

19

efforts. The role of business intelligence has expanded to include providing actionable insights, reducing risk in marketplace actions, and improving return on investment (Coop & Schindler, 2006, pg. 13). Marketing Analytics While quantitative analysis provides powerful insights for improving all manner of operational processes, the firms marketing decisions are of primary interest in this paper. Just as in other functional areas of the company, marketing managers are discovering the significant benefits offered by a more scientific approach to decisionmaking (Bradley, 2007, pg. 277). By measuring and analyzing data, marketing leaders are able to understand which systems, projects, staff, and processes are performing well and which need improvement. Just as the dashboard of a car provides critical information about the performance of the vehicle, marketing dashboards can provide real-time metrics on where the marketing investments are paying off and where they arent (LaPointe, 2005, pg. 17). This new emphasis on data-driven decisions is a welcome change for many marketers. Traditionally, the field of marketing was considered to be more of an art than a science, and decisions were made based solely on guesswork and subjective speculation. Though CEOs would agree that marketing plays a role in the companys success, they just dont know how to quantify that role (LaPointe, 2005, pg. 18). For

Marketing Optimization this reason, marketers have struggled to justify their investments, particularly in weakening economic conditions.

20

Quantifiably proving this link between marketing investments and hard financial returns for the company remains one of the greatest challenges for most marketing departments, and demonstrating how to achieve it is precisely the purpose of this project. Recent research shows that marketing ROI [return on investment] is both a priority and a challenge for most companies (Lenskold, 2003, pg. 6). According to author Patrick LaPointe, there are a broadening number of brilliant academics who are now focusing exclusively on the marketing disciplineadvancing mathematical science in marketing in some extremely innovative ways (LaPointe, 2005, pg. 24). At the same time, these efforts are also bringing renewed financial discipline into the field of marketing. New models for marketing communications are being deployed not only to engage more elusive customers, but also to deliver more fiscal accountability required from boards (Young & Aitken, 2007, pg. 14). Customers are continually faced with a discrete choice between similar products from a variety of competing companies. Very often the possible options are mutually exclusive; the customer will only purchase one item (rejecting the other alternatives). The prevalence of this discrete choice has given rise to extensive study of the perceptions that influence those decisions (Mazanec & Strasser, 2000, pg. 11). This discipline of perceptions-based market segmentation provides business leaders with simultaneous market segmentation and product positioning framework (Buchta, Dolnicar, & Reutterer, 2000, pg. 19). If firms know the distribution from which these taste

Marketing Optimization parameters are drawn, they can forecast demand using a discrete choice model of consumer behavior (Anderson, de Palma, & Thisse, 1992, pg. 3). Operations Research & Utility Theory

21

The real value of using statistical methods to quantify and measure the effect of various marketing factors is that these factors can then be optimized in order to maximize a desired outcome (such as expected revenue, market share, or lead generation). This step takes the role of marketing analysis even beyond predicting customer behavior; it allows the firm to essentially control it by optimizing the relevant input parameters. Marketing Analyst Patrick LaPoint promotes strongly the use of optimization within the context of marketing: Once the discipline of financial assessment is adopted across most individual marketing initiatives, the entire portfolio of possible initiatives competes for scarce budget dollars on the basis of forecast returnsOptimization techniques are used to solve for the highest possible return in terms of media mix, segment emphasis, and channel management (LaPointe, 2005, pg. 37). The principle of optimization comes from the field of Operations Research, a discipline dedicated to using mathematical modeling and statistical methods to help people analyze complicated business processes and make good decisions (Crowder, 2006, pg. 2). According to Harlan Crowder, Operations Research Consultant, these tools and techniques are especially useful for helping understand and deal with

Marketing Optimization business complexity and uncertainty (Crowder, 2006, pg. 2). Just as new products typically undergo rigorous testing prior to their release, no business executive

22

should dream of making a strategic decision with major impact on the corporation until that decision has been modeled and validated using operations research technology (Crowder, 2006, pg. 8). Operations Research utilizes many different types of mathematical models, but perhaps the most prevalent are linear programming models (Lawrence & Pasternack, 2002, pg. 48). The expressions in these models contain only linear terms, that is, variables that are raised to the first power. Many problems naturally lend themselves to a linear programming formulation, and many other problems can be closely approximated by models with this structure (Lawrence & Pasternack, 2002, pg. 48). Quite often these problems deal with finding an optimal combination of variables to produce a desired outcome. This subset of problems is aptly named combinatorial optimization, and they frequently occur in transportation, inventory, production, resource allocation, scheduling, packing, and assignment situations, to name but a few of the possible applications. A number of well-documented algorithms exist for efficiently solving linear programming problems (Schrijver, 2008, pg. 102). One of the earliest and still among the most commonly used of these is the Simplex algorithm, developed in the 1950s by George Dantzig (Dantzig, 1963, pg. 94). Richard Bellman, meanwhile, pioneered a variety of techniques for solving nonlinear problems; those containing exponential variables (Bellman, 1957, pg. 7). The model in this paper is, by definition, a non-linear

Marketing Optimization problem initially because it seeks the simultaneous satisfaction of multiple objectives. However, the model can be translated into a linear problem by using the technique of indexing from the field of Utility Theory.

23

Utility theory attempts to quantify the factors that lead to decision outcomes, and provide a framework for rational decision-making. All possible alternatives are listed in a payoff table and assigned specific utility values as they relate to the decision criterion (Lawrence & Pasternack, 2002, pg. 329). If there are multiple criteria (which is very often the case), those criteria are weighted according to the priorities of the decision maker. The utility values are first multiplied by their respective criteria weights and then summed and evaluated objectively against all other alternatives. The alternative with highest score is chosen for execution. Decision theory is a general approach to decision making when the outcomes associated with alternatives are often in doubt (Krajewski, Ritzman, & Malhorta, 2007, pg. 30). Conclusion This project advances existing literature by combining methods from several different mathematical disciplines and applying them in the context of marketing. Multivariate Testing will be used to determine exact parameter values for each of the possible marketing options. Next, the two coefficients will be indexed by their relative correlation to the objective functions, and then blended into a single utility value. With all relevant data now in place, the final step will be to employ combinatorial optimization in order to find the best possible allocation of the marketing budget, subject to the predefined constraints. The result will be a more scientific approach to the

Marketing Optimization

24

marketing plan, yielding a higher return on the marketing investment and, ultimately, give the company a stronger competitive position in the marketplace.

Marketing Optimization Chapter 3: Method Statement of Purpose Marketing departments are continually challenged with allocating limited

25

financial resources in order to maximize revenue. Surprisingly, many companies fail to measure the efficacy of these investments, and fewer still use quantitative criteria for deciding how to allocate their marketing budgets. Indeed, most marketing plans are based purely on anecdotal evidence and speculation. Although these efforts may produce moderately successful results, a more scientific approach will, in many cases, yield significantly greater returns on marketing investment. Research Question In preparing a marketing plan for the coming year, the Director of Marketing must decide where to allocate limited financial resources and which marketing opportunities to pursue. Each marketing option (J) has a utility value (UJ) and a cost (KJ) associated with it. Executive management has set a total budget limitation (B), and there are (N) possible options from which to choose. The objective is to construct a marketing plan within the budget constraint that will generate the highest possible utility. Let QJ = the quantity of each option (J) chosen for the final marketing plan. The question can now be formulated as a simple integer linear programming problem in standard form: MAXIMIZE SUBJECT TO UJQJ DJQJ B Q0 Q= (Total utility) (Budget constraint) (Nonnegativity) (Integers)

Alternatives. For the purposes of this study, each generator product has its own unique size (in kilowatts), features and applications, such that only one product will be

Marketing Optimization featured at a time (Table 1). Table 1 Generator Products by Size and List Price, 2009

26

Label P1 P2 P3 P4 P5

Name Raven Generator Falcon Generator Hawk Generator Osprey Generator Eagle Generator

Size 750kW 1000kW 1250kW 1500kW 1750kW

List Price $25,540 $38,325 $49,940 $65,515 $77,075

The Director of Marketing has identified five specific vertical markets that have strong historical demand for these generator products, as well as a positive outlook for continued growth. The five markets are found in Table 2 by their NAICS code (North American Industry Classification System, as defined by the U.S. Department of Commerce). Table 2 Vertical Market Sectors with NAICS Codes, 2009

Label M1 M2 M3 M4 M5

Name Highway Construction Marine Shipping Airport Infrastructure Data Processing Centers Hospitals/Medical Centers

NAICS Code 2373 4831 4881 5182 6221

Each of these five markets has a variety of targeted publications, tradeshows, and

Marketing Optimization other industry-related media that provide opportunities to advertise the firms

27

products. The Director of Marketing has determined that the five advertising channels listed in Table 3 will best serve the companys marketing needs. The cost per unit for each channel (including production costs) is consistent within each of the markets. Table 3 Advertising Channels with Cost Per Unit, 2009

Label C1 C2 C3 C4 C5

Channel Print Ad in Trade Magazine Tradeshow Exhibit Direct Mailing Targeted Email Blast Search Engine Campaign

Cost Per Unit $11,000 $19,000 $9,500 $7,250 $8,000

Criteria. The final indexed utility value is designed to facilitate the simultaneous satisfaction of multiple criteria. In this case, executive management has determined that the primary objective is to maximize the total revenue of the marketing plan. At the same time, however, management wants to take a secondary goal into consideration: targeted revenue from a specific geographic region. These two goals are conflicted; a marketing plan optimized solely for revenue maximization would be sub-optimal in satisfying the geographic objective. Conversely, a marketing plan optimized for the targeted region would probably not yield maximum total revenue. Taking the average coefficients of the two goals, the utility value provides a single quantitative value for the final linear programming problem.

Marketing Optimization

28

Each marketing option (J) represents a specific combination of a featured product (P), a targeted vertical market (M), and an advertising channel (C). In this scenario, the Director of Marketing is considering five different products to feature, promoted in five different vertical markets, each offering five different marketing channels (as previously described). Thus, there are three discrete factors within each marketing option (J). Each factor has five levels, and the total number of possible combinations (N) is 53 = 125. In order to preserve a relatively balanced mix of products and markets, executive management has set arbitrary minimum requirements (TP, TM). These must be added to the model as constraints: QP TP ; QM TM . Linear programming models (by definition) necessitate an assumption of proportionality (Lawrence & Pasternack, 2002, pg. 49). This requires that all variables produce constant returns to scale. In reality, however, marketing efforts frequently violate this assumption. Marketing response behavior is often more accurately depicted as a curve (similar to the curves of supply and demand used in Economics). In the early stage, a marketing campaign will typically exhibit steady response in direct proportion to the quantity or volume of effort. Eventually, response will peak and then subside as the product or brand suffers from overexposure. The peak of this curve is known in Marketing as the saturation point. Although this curve violates the assumption of proportionality, the returns up to the saturation point can be assumed to be (approximately) constant. The point of saturation (S) must be determined for each market, and for each channel within each market (V), and then entered as constraints in the model: QM SM ; QC VC (for each market M).

Marketing Optimization With these constraints added, the final form of the model is now: MAXIMIZE SUBJECT TO: UJQJ KJQJ B QP TP QM TM QM SM QC UC Q0 Q= (Total utility) (Budget constraint) (Product mix requirement) (Market mix requirement) (Market saturation constraint) (Channel saturation constraint) (Nonnegativity) (Integers)

29

For each M,

Where:

J = Each marketing option N = Total number of possible options Q = Quantity of each option J U = Utility value of each option J K = Cost of each option J B = Total budget limitation P = Featured product for each option J M = Targeted vertical market for each option J C = Advertising channel for each option J T = Minimum requirement for each market M and product P S = Total saturation point for each market M V = Saturation point for each channel C (within each market M)

Data Collection To derive the indexed utility value, the expected total revenue (R) and expected targeted revenue (G) must be determined for each option (J). Multivariate testing methods will be used to find the lead generation coefficients for each product (P), market (M), and channel (C). The total number of expected leads generated for each option (J) will be used to determine expected revenue (R), while only those leads originating from the specific geographic region will be counted toward the targeted revenue objective. The predicted number of leads generated (LR and LG) is then multiplied by the historical conversion rate (VP) of the sales team. The result is the expected number of products sold, and this can, in turn, be multiplied by the list price

Marketing Optimization (W) for that product to find to total expected revenue and targeted revenue for that

30

option. Thus, for each option (J): RJ = LRVPWP and GJ = LGVPWP . These values RJ and GJ are then divided by the cost (KJ) of executing the respective marketing option (J) to find an efficiency rating (FR and FG) for that option relative to both goals. Next, these efficiencies are indexed (XR and XG) by dividing each individual value by the average of all values in order to provide a level comparison. Finally, both indexed efficiency ratings are averaged together to reach a final utility value (U) for each option (J). Data Collection Procedures. Many industrial companies unfortunately do not track the marketing performance data required for the analysis in this project. It is the authors intent to show that doing so is not only financially beneficial, but also vital to securing a long-term competitive advantage. Those companies that do measure and analyze quantitative data on their marketing efforts typically keep such information extremely confidential and proprietary. This is done for good reason, since detailed information regarding the performance of specific products, market segments and promotional channels constitutes the very essence of an organizations competitive strategy. If a rival corporation were to possess this information, they would have an immediate and unfair advantage, because strengths and weaknesses could be readily evaluated and dominant counter-strategies would be deployed. Such a breach of data security would quite likely violate anti-competitive legal regulations. The author does not advocate sharing or reporting any information that might put an organization at risk. Instead, the purpose here is to encourage companies to carefully measure marketing performance metrics for internal use only. Because this

Marketing Optimization

31

data is not available in many companies, and to avoid potentially exposing confidential information, much of the lead generation, product pricing and channel cost data used in this paper was obtained through Monte Carlo simulation performed in Microsoft Excel. This data was reviewed and verified by an independent Marketing Data Analyst to ensure authenticity. The techniques used and conclusions drawn are sound, and these exercises can be duplicated in most industrial marketing environment, and, to a broader extent, in any resource allocation problem. Analytical Tools and Tests. Researchers and statisticians frequently use simple linear regression to test the relationship between one input factor and one response factor. While this tool is useful in some situations, it is much more common in marketing settings to have multiple factors potentially influencing the response outcome. In these circumstances it is far more appropriate to use multiple regression analysis, which can be used either as a descriptive or as an inferential technique (Lind, Marchal & Wathen, 2005, pg. 475). This technique, also known as Multivariate Testing, is particularly valuable because it not only tests the impact of each factor on the response variable, but also tests for any interactions between factors (Burnham, 2004, pg. 34). The effect of these interactions is entirely missed if simple linear regression is used to test multiple factors separately. This project involves three discrete factors (product, market and channel) that potentially impact a response variable (leads generated). The Multivariate Testing will reveal what (if any) effect each factor has on lead generation. These are known as main effects. Additionally, the testing will show if there are any interactions affecting the

Marketing Optimization

32

response outcome. For example, a given product should have a consistent quantifiable impact on lead generation, independent of which market or channel is used; this is the main effect for that product. However, it may also be found that certain products perform better in some markets than others. This situation is an example of an interaction between two factors (product and market) impacting the response variable (lead generation). These interactions are often critical to the final results of the research (Burnham, 2004, pg. 34). Of interest to the marketing analyst is the extent to which each independent variable or variable interaction can explain or predict the dependent variable (Lind, Marchal & Wathen, 2005, pg. 475). The multiple regression analysis will be performed using JMP statistical software from SAS, and will produce a coefficient value for each factor and interaction effect. The regression coefficient is the precise numeric impact of each variable on the response, within the error margin of the overall model (Lind, Marchal & Wathen, 2005, pg. 449). This error margin is called the coefficient of multiple determination, expressed as the percent of the total variation explained by the regression (Lind, Marchal & Wathen, 2005, pg. 484). Mathematical modeling is a process that translates observed or desired phenomena into mathematical expressions (Lawrence & Pasternack, 2002, pg. 7). These expressions can then be quantitatively evaluated, usually with the goal of finding an optimal solution to the problem. The solution identified in the mathematical model is then taken back into the real world environment for verification and implementation. Because these problems can involve extremely complex computations and

Marketing Optimization

33

millions of feasible solutions from which to choose, computers are usually employed to find a solution. To explore the solution space, trying thousands of possible combinations in search of the optimal one, computers execute various algorithms, most notably the Simplex Algorithm (Dantzig, 1963, pg. 94). The optimization for this project will be performed in Microsoft Excel, using the built-in solver tool. Conclusion In preparing a marketing plan for the coming year, the Director of Marketing must decide where to allocate limited financial resources and which marketing opportunities to pursue. Each marketing option has a utility value and a cost associated with it. Executive management has set a total budget limitation, and there are additional constraints intended to preserve a relatively balanced product offering and a diversified mix of markets and channels. The objective is to construct a marketing plan that will generate the highest possible utility, while satisfying all constraints. To derive the utility values, the expected total revenue and expected targeted revenue must be determined for each option. Multivariate testing methods will be used to determine the regression coefficients for each product, market and channel. Additionally, any interactions between these factors will be evaluated for their effect on the response variable (lead generation). Once all numeric values have been calculated, the problem can be formulated and solved as an integer linear programming model.

Marketing Optimization Chapter 4: Results and Analysis Findings

34

Multivariate Testing is used to determine the precise impact of each factor on the response outcome. This MVT model is a full factorial standard least squares regression, because the historical lead generation data includes one row of data for every possible treatment combination, or 125 records (Appendix A). There two separate response outcomes: total lead generation and targeted lead generation (leads originating from the targeted geographical region). The input factors were tested separately for their correlation to the two response outcomes. All three main effects were included in the model, as well as the interactions between market and the other two factors. The interaction between product and channel was tested but then removed from the model because it was found not to be significant. This indicates that each channel generally performs equally well, independent of which product is featured. Thus, there is no interaction between product and channel, which makes sense intuitively. The Multivariate Testing was performed using JMP software from SAS. Figure 1 shows a screen of the softwares output relating to the statistical significance for both total leads and targeted leads (for complete statistical output, see Appendices B & C). The first thing the researcher must examine in the statistical output is the Analysis of Variance (ANOVA) table. The probability of the value of F indicates the level of overall significance for the model. The decision rule for this value is traditionally to reject a model with probability greater than 0.05. In this case the value is less than 0.0001 for both responses, so the overall model is indeed significant.

Marketing Optimization Figure 1 Statistical Software Output for Multivariate Testing, 2009

35

As previously noted, the coefficient of multiple determination (r2) indicates how much of the total variation is explained by the model. Higher values here indicate a stronger ability to predict the response outcome. Figure 2 shows that r2 = 0.76 for total leads, and 0.73 for targeted leads. Both of these values are deemed acceptable, confirming the general accuracy and validity of the statistical model. Once the models overall significance has been verified, the researcher proceeds

Marketing Optimization to examine the factor coefficients (Appendix D). Based on the historical data, which includes a record for every possible treatment combination of factors, the statistical software has calculated a precise coefficient value for each factor (and interaction of

36

factors). These coefficients indicate exactly what impact each factor has on the response variable. For example, the Raven generator has the greatest impact on total lead generation of any product, while the Eagle has the lowest. This is consistent with the historical sales data, since the Eagle is the largest, most expensive product that appeals to a more select customer base. However, it should be noted that this does not apply in the targeted region of Asia, where the Hawk has the highest coefficient and the Falcon has the lowest. Data Centers is the strongest market coefficient for total lead generation, while Airports is the lowest. For targeted leads, however, Airports is the strongest market while Hospitals is the weakest. Of the advertising channels, Trade Shows has the highest coefficient for total lead generation, and Print Ads has the lowest. The weakest channel for targeted leads is Search Engines, while Email is the strongest. Equally important to the response outcomes are the interactions between factors (also shown in Appendix D). The interactions between product and market were significant, as were the interactions between channel and market. For example, the Raven generator product clearly performs strongest in the Construction market, for both total leads as well as targeted leads. The Eagle product, on the other hand, appeals strongly to Hospitals for both response outcomes. For total lead generation, Print Ads perform well when targeted to the Data Centers market, but poorly in the Airports market. To

Marketing Optimization

37

generate leads from Asia, the best choices are Print Ads for Hospitals, or Search Engine campaigns aimed at Airports. These insights, gained from analyzing the factor interactions, have obvious value for the Director of Marketing, and also for the Executive Management. Yet this is exactly the type of intelligence many companies are lacking by not utilizing statistical tools like MVT on their marketing data. Factor coefficients have a purpose beyond individual analysis; they are used to create a prediction equation. Starting with the intercept value and then adding the respective coefficient values for the main effects and interactions, the lead generation response (L) can be accurately predicted for any given treatment combination of product, market and channel. This predicted value is then compared against the actual (historic) value as a reality check for how effective the prediction equation is. The difference between the predicted value and the actual value is known as a residual (Appendices E & F). While some random variation is to be expected, the residual columns for this model show that the equation is reasonably accurate in predicting both total lead generation and targeted lead generation. However, the reader will recall that maximizing lead generation is not the objective function of this problem. Indeed, it would produce sub-optimal results for both total revenue and targeted revenue because thus far the analysis has ignored the cost to execute each option. Predicted values for total lead generation (LR) and targeted lead generation (LG) must first be translated into expected revenues (R), and then be measured against their cost (K) in order to properly evaluate them for their financial efficiency (F). This is yet another step alluding many companies, who fail to measure

Marketing Optimization the financial strength of their marketing investments.

38

To arrive at an efficiency rating for each marketing option, the predicted number of total and targeted leads for each option is multiplied by a respective conversion rate (V), which varies by product and geographic location. The conversion rate is the average percentage of leads that the sales team can expect to convert into actual sales, based on historical performance data. Thus each marketing option has an expected number of product sales, and, when multiplied by the list price for the given product, the result is an expected total revenue (R) and targeted revenue (G) for that option. The next step is to divide the expected revenue for each option by its cost to find its efficiency rating. Now there are two efficiency ratings for each marketing option: one for the total revenue goal and another for the targeted revenue goal. These two values must be consolidated into one to meet the requirements of a linear programming model for optimal resource allocation. To do this, the efficiency ratings must first be indexed (divided by the average of all efficiencies for that goal, Appendices G & H) and then averaged together. These computations have taken each marketing options ratings for two conflicted goals and effectively blended them into a single utility value. Now each option has one value indicating its capacity to simultaneously satisfy both goals (Appendix I). The values for all variables are now in place, so the problem can be modeled using integer linear programming. There are 125 discrete choices from which to build the marketing plan. If the total budget of $250K is the only constraint on the problem,

Marketing Optimization

39

and assuming all variables are nonnegative integer values, there are more than 6.94x1025 possible ways to allocate the marketing budget. This is found using the formula: (n + r 1)! r!(n-1)! This is irrelevant, however, because the optimal solution in such an unconstrained environment would be to simply identify the single product, market and channel with the highest utility value and spend the entire budget on that one option. Instead, management is taking the more realistic approach of imposing (somewhat) arbitrary constraints on the solution to help ensure a relatively balanced mix of products, markets and channels. To guarantee that each of the five products receives some promotion, the marketing plan must include at least four options from each product. Management is also committed to maintaining a diversified cross-section of vertical market presence. Therefore at least two options must be chosen from each of the five markets. The Director of Management wants the marketing plan to include a variety of advertising channels, so each channel must be represented at least four times. The exception to this rule is Trade Shows, which were assigned a minimum value of three, because they are less frequent and more expensive. Table 4 presents the list of mix constraints.

Marketing Optimization Table 4 List of Product, Market & Channel Mix Constraints, 2009

40

Category Product Mix Requirement Product Mix Requirement Product Mix Requirement Product Mix Requirement Product Mix Requirement Market Mix Requirement Market Mix Requirement Market Mix Requirement Market Mix Requirement Market Mix Requirement Channel Mix Requirement Channel Mix Requirement Channel Mix Requirement Channel Mix Requirement Channel Mix Requirement

Item Raven Falcon Hawk Osprey Eagle Construction Shipping Airports Data Centers Hospitals Print Ad Trade Show Direct Mail Email Campaign Search Engine

Relationship

Limit 4 4 4 4 4 2 2 2 2 2 4 3 4 4 4

Another constraint unique to the field of Marketing is the concept of saturation. In the early stage, a marketing campaign will typically exhibit constant returns to scale. Eventually, however, response will peak and then subside as the product or brand suffers from overexposure. The peak of this curve is known in Marketing as the saturation point. All saturation points must be estimated using historical data, and then entered as constraints in the model (Appendix J). In this problem, the Director of Marketing must contend with two different saturation points. First, a constraint must be placed on the number of promotions within any given market sector. Figure 2 illustrates how this value is derived. Note that in some cases, the saturation point is not simply the quantity that produces the highest

Marketing Optimization

41

response. In the channel saturation analysis for P2, for example, the saturation point was determined to be two trade shows within the same market. Even though a third trade show would generate additional leads, the efficiency of this channel will have significantly eroded, and the money would be much more effective elsewhere in the marketing plan. Figure 2 Saturation Point Examples, 2009

Marketing Optimization Yet even within the bounds of the total market constraint, overexposure in any single advertising channel would likely produce a saturation effect. Thus, a second

42

saturation point must be determined for each channel within the same market. Table 5 lists all saturation constraints. Table 5 List of Saturation Constraints, 2009

Category Total Market Saturation Total Market Saturation Total Market Saturation Total Market Saturation Total Market Saturation Within Each Market Within Each Market Within Each Market Within Each Market Within Each Market

Item Construction Shipping Airports Data Centers Hospitals Print Ads Trade Shows Direct Mail Email Campaigns Search Engine

Relationship

Limit 9 9 9 9 9 8 2 4 4 3

Using the same formula as before, the additional constraints have reduced the total number of feasible solutions to approximately 49.7 trillion. Although this number is considerably lower, the problem remains highly complex. Finding the optimal solution from among all feasible solutions will only be possible with the aid of a computer. Further, the probability that the optimal solution could be found by guessing is extremely remote. Thus, with all constraint values now ascertained (Appendix K), the problem can at last be solved using an integer linear programming model. All variables in the mathematical model have been assigned values. The problem, as originally

Marketing Optimization stated, is: MAXIMIZE SUBJECT TO: UJQJ KJQJ B QP TP QM TM QM SM QC UC Q0 Q= (Total utility) (Budget constraint) (Product mix requirement) (Market mix requirement) (Market saturation constraint) (Channel saturation constraint) (Nonnegativity) (Integers)

43

For each M,

Where:

J = Each marketing option N = Total number of possible options Q = Quantity of each option J U = Utility value of each option J K = Cost of each option J B = Total budget limitation P = Featured product for each option J M = Targeted vertical market for each option J C = Advertising channel for each option J T = Minimum requirement for each market M and product P S = Total saturation point for each market M V = Saturation point for each channel C (within each market M)

Implication for the Research Question In preparing a marketing plan for the coming year, the Director of Marketing must decide where to allocate limited financial resources and which marketing opportunities to pursue. Each marketing option has a utility value and a cost associated with it. Executive management has set a total budget limitation of $250,000, and there are 125 possible options from which to choose. The objective is to construct a marketing plan within the budget constraint that will generate the highest possible utility. In actuality, there are four possible marketing plans for management to consider. The previous marketing plan was entirely heuristic, yet it satisfies all of the same constraint requirements as the optimized models. In this way, fair comparisons can be

Marketing Optimization made between the heuristic plan and the optimized plans. For the benefit of managements consideration, the researcher has prepared three different marketing plans. Using the expected revenue values in the linear programming model, management may consider a marketing plan optimized for total revenue, another optimized for targeted revenue from Asia, and the final blended marketing plan to simultaneously satisfy both goals. The specific options prescribed by each of the four marketing plans may be found in Appendix L. Each of the four marketing plans should be checked to ensure satisfaction of all

44

constraints (Appendix M). They may then be evaluated for their financial efficiency and their utility relative to both of the strategic goals of management. The heuristic model has an expected total revenue of $9,692,610.70, and a total cost of $245,500. An efficiency rating can be found by dividing the expected revenue by the total cost. The heuristic model has an total revenue efficiency of 39.97. As for targeted revenue (Asia), the heuristic model has an expected revenue of $2,204, 365.71 and an efficiency rating of 9.09. As previously stated, this method has produced adequate results in the past, but management would like to ascertain if a more scientific method of resource allocation would produce better results. Were management solely concerned with total revenue, the highest possible plan optimized for that goal would produce an expected revenue of $16,495,690.32. However, this plan sacrifices some revenue originating from the targeted region of Asia, when compared against the heuristic model. Alternatively, a model optimized solely for the maximization of targeted revenue would result in $2,807,533.18 in revenue

Marketing Optimization originating from Asia, while at the same time enjoying a considerable increase in total revenue when compared against the previous heuristic model. Employing the indexed utility value, however, produces a blended model that can simultaneously satisfy both objectives. The blended model results in a total expected revenue of $15,173,454.14, as well as an expected targeted revenue of

45

$2,608,981.81. This represents an impressive 57% increase in total revenue, and an 18% increase in targeted revenue. The blended model results in a total revenue efficiency of 62.31, meaning that every dollar of the marketing budget returns $62.31 in total revenue. The blended marketing plan is outlined in Table 6. Table 6 Final Marketing Plan from the Blended Utility Model, 2009

Product Raven Falcon Hawk Hawk Hawk Osprey Osprey Osprey Eagle

Market Construction Shipping Airports Airports Airports Data Centers Data Centers Data Centers Hospitals

Channel Print Ad Direct Mail Trade Show Email Search Engine Trade Show Email Search Engine Email

Quantity 4 4 1 2 3 2 4 1 4 TOTAL

Cost $44,000 $38,000 $19,000 $14,500 $24,000 $38,000 $29,000 $8,000 $29,000 $243,500

Sensitivity analysis offers insights into the impact of each constraint, allowing management to reevaluate them (Lawrence & Pasternack, 2002, pg. 63). The Director of Marketing can identify which constraints are limiting the overall objective and by how

Marketing Optimization

46

much. Email is by far the most efficient channel for promotion. Adding one additional email campaign to the blended marketing plan would result in a total cost of $250,750. Technically, this violates the budget constraint, but in practice this overrun of only 0.3% would generate more than $600,000 in additional revenue (depending on which market and product were chosen). The single option with both the highest utility and highest efficiency is P5M5C4 (Eagle, Hospitals, Email). The blended model fully exploits this option up to the channel saturation constraint for Email in the Hospital market. Print Ads and Trade Shows are much less efficient means of promotion than other channels, yet the channel mix constraint required that they both be included in the final plan at minimum quantities. The Director of Marketing may wish to reconsider this requirement and instead allocate that money to more efficient options.

Marketing Optimization Chapter 5: Conclusion The blended utility model results in a strong marketing plan that can simultaneously satisfy both of managements objectives. The plan will produce a total expected revenue of $15,173,454.14, as well as an expected targeted revenue of

47

$2,608,981.81. This represents an impressive 57% increase in total revenue, and an 18% increase in targeted revenue, when compared against the previous heuristic plan. The blended model results in a total revenue efficiency of 62.31, meaning that every dollar of the marketing budget returns $62.31 in total revenue. Recommendations While the scope of this project was limited to identifying an optimal marketing plan for the coming year, the research has illuminated additional opportunities for improving the process of decision-making within the organization. The author strongly recommends the following actions: 1. Provide greater visibility for marketing data. In many companies, historical marketing performance values are often difficult to find and/or incomplete. Dashboards can provide management with real-time information about critical business processes, revealing potential problems and opportunities. 2. Employ Multivariate Testing methods. Statistical techniques such as MVT can provide accurate and reliable insights about the factors that have the greatest impact on business results. Despite the widespread availability of statistical and spreadsheet software, many business

Marketing Optimization decisions are made without the benefit of these powerful tools. Marketing plans developed heuristically are really not plans at all; they are expensive gambles with company resources. 3. Incorporate strategic goals into the marketing plan. Rather than simply maximizing total revenue, this project identified a solution that will position the company for long-term growth while still demanding strong top-line performance. Executive managements long-term and

48

short-term priorities should both inform the decisions of the marketing team. As these priorities continue to evolve, it is imperative that they be communicated to the marketing organization on a continual basis to ensure alignment. 4. Utilize Linear Programming. The field of Operations Research has developed invaluable tools to solve real business problems and aid in decision-making. Companies that utilize these quantitative analytical techniques will be able to achieve a competitive advantage over those that simply rely on anecdotal evidence and speculation. These heuristic methods of decision-making yield sub-optimal results, because the most advantageous solutions are often counterintuitive (particularly in marketing applications). 5. Additional Research. While this study has shown how the allocation of the overall marketing budget can be optimized, Linear Programming techniques have been equally effective when applied to the execution of

Marketing Optimization a marketing plan, such as campaign scheduling, customer selection,

49

identification of cross-sell opportunities, calculating optimal discounts and designing various promotional tactics. Despite the complexities and challenges of the marketing function, Operations Research can yield tremendous value by reducing uncertainty and risk to manageable levels. Implementation Traditionally, the field of marketing has been viewed as intangible and subjective. Despite this perception, a scientific, quantitative approach to resource allocation can be effectively applied in a marketing environment. Companies should insist on the same degree of financial accountability from their marketing investments that they would expect from any other capital or operational expenditures. Techniques from the field of Applied Mathematics (such as multivariate testing, indexed decision tables and Combinatorial Optimization) can be used to maximize the marketing plans return on investment. Executive management has set an arbitrary constraint on the total marketing budget, as well as specific minimum requirements to maintain a relatively balanced product/market offering. Saturation points must be determined using historical marketing data, and then entered as additional constraints in the linear model. The objective function is to produce strong total revenue, while simultaneously positioning the company for future growth in Asia.

Marketing Optimization This project advances existing literature by combining methods from several different mathematical disciplines and applying them in the context of marketing. Multivariate Testing is used to determine exact parameter values for each of the possible marketing options. The final blended model employs combinatorial optimization in order to find the best possible allocation of the marketing budget, subject to the predefined constraints. The result is a more scientific approach to the

50

marketing plan, yielding a higher return on the marketing investment and, ultimately, gives the company a stronger competitive position in the marketplace.

Marketing Optimization References Anderson, S.P., de Palma, A., & Thisse, J. (1992). Discrete Choice Theory of Product Differentiation. Cambridge, MA: MIT Press. Bellman, R. (2003). Dynamic Programming. Mineola, NY: Dover Publications. (Original work published 1957). Bradley, N. (2007). Marketing Research: Tools and Techniques. Oxford, UK: Oxford University Press.

51

Buchta, C., Dolnicar, S., & Reutterer, T. (2000). A Nonparametric Approach to PerceptionsBased Market Segmentation: Applications. New York, NY: Springer-Verlag Wien. Burnham, R.A. (2004, March). That Voodoo We Do: Marketers Are Embracing Statistical Design of Experiments. Direct Marketing Association. 32-37. Christensen, C.M. (1997). The Inventor's Dilemma: When New Technologies Cause Great Firms to Fail. Boston, MA: Harvard Business Press. Cooper, D.R., & Schindler, P.S. (2006). Business Research Methods (9th ed.). New York, NY: McGraw-Hill/Irwin. Crowder, H.P. (2006). The Science of Better Decisions: Making Operations Research a Core Business Competency. Hanover, MD: Institute for Operations Research and The Management Sciences. Dantzig, G. (1998). Linear Programming and Extensions. Princeton, NJ: Princeton University Press. (Original work published 1963). Davenport, T.H. & Harris, J.G. (2007). Competing on Analytics: The New Science of Winning. Boston, MA: Harvard Business Press.

Marketing Optimization

52

Dutta, P.K. (1999). Strategies and Games: Theory and Practice. Cambridge, MA: MIT Press. Fundenberg, D., & Triole, J. (1991). Game Theory. Cambridge, MA: MIT Press. Grant, R.M. (2002). Contemporary Strategy Analysis: Concepts, Techniques, Applications (4th ed.). Malden, MA: Blackwell Publishing. Holland, C.W. (2005). Breakthrough Business Results with MVT: A Fast, Cost-Free "Secret Weapon" for Boosting Sales, Cutting Expenses, and Improving Any Business Process. Hoboken, NJ: John Wiley & Sons, Inc. Kotler, P., & Keller, K.L. (2006). Marketing Management (12th ed.). Upper Saddle River, NJ: Pearson Prentice Hall. Krajewski, L.J., Ritzman, L.P., & Malhotra, M.K. (2007). Operations Management Processes and Value Chains. Upper Saddle River, NJ: Pearson Prentice Hall. LaPointe, P. (2005). Marketing by the Dashboard Light: How to Get More Insight, Foresight, and Accountability from Your Marketing Investments. Princeton, NJ: MarketingNPV. Lawrence, J.A. Jr., & Pasternack, B.A. (2002). Applied Management Science: Modeling, Spreadsheet Analysis, and Communication for Decision Making (2nd ed.). Hoboken, NJ: John Wiley & Sons, Inc. Lenskold, J.D. (2003). Marketing ROI: The Path to Campaign, Customer, and Corporate Profitability. New York, NY: McGraw-Hill Professional. Lind, D.A., Marchal, W.G., & Wathen, S.A. (2005). Statistical Techniques in Business & Economics (12th ed.). New York, NY: McGraw-Hill/Irwin. Mazanec, J.A., & Strasser, H. (2000). A Nonparametric Approach to Perceptions-Based Market Segmentation: Foundations. New York, NY: Springer-Verlag Wien.

Marketing Optimization

53

McGuigan, J.R., Moyer, R.C., & Harris, F.H.B. (2005). Managerial Economics: Applications, Strategy, and Tactics (10th ed.). Maston, OH: Thomson South-Western. Pearce, J.A. & Robinson, R.B. Jr. (2009). Strategic Management: Formulation, Implementation, and Control (11th ed.). New York, NY: McGraw-Hill/Irwin. Peppers, D., & Rogers, M. (2004). Managing Customer Relationships: A Strategic Framework. Hoboken, NJ: John Wiley & Sons, Inc. Porter, M.E. (1980). Competitive Strategy: Techniques for Analyzing Industries and Competitors. New York, NY: Free Press. Reichheld, F.F. (2003, December). The One Number You Need to Grow. Harvard Business Review. 52-61. Schrijver, A. (2008). A Course in Combinatorial Optimization. Amsterdam, NL: (Author). Ullman, D.G. (2006). Making Robust Decisions. Oxford, UK: Trafford Publishing. Van Horne, J.C. (2002). Financial Management and Policy (12th ed.). Upper Saddle River, NJ: Prentice Hall. Williams, S., & Williams, N. (2007). The Profit Impact of Business Intelligence. San Francisco, CA: Morgan Kaufmann. Young, A., & Aitken, L. (2007). Profitable Marketing Communications: A Guide to Marketing Return on Investment. London, UK: Kogan Page.

Marketing Optimization Appendix A Historical Lead Generation Values


Historical Value Total Leads 86 110 98 89 76 60 76 73 70 67 54 63 53 44 58 65 92 58 82 64 52 77 69 66 47 84 86 48 72 58 65 85 50 59 76

54

J 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

M C

Product

Market Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports Data Centers Data Centers Data Centers Data Centers Data Centers Hospitals Hospitals Hospitals Hospitals Hospitals Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping

Channel Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Historical Value Targeted Leads 11 13 11 10 8 12 20 16 16 12 10 19 13 17 16 10 12 10 15 10 15 9 9 14 11 11 12 7 8 7 12 13 12 13 11

P1 M1 C1 Raven P1 M1 C2 Raven P1 M1 C3 Raven P1 M1 C4 Raven P1 M1 C5 Raven P1 M2 C1 Raven P1 M2 C2 Raven P1 M2 C3 Raven P1 M2 C4 Raven P1 M2 C5 Raven P1 M3 C1 Raven P1 M3 C2 Raven P1 M3 C3 Raven P1 M3 C4 Raven P1 M3 C5 Raven P1 M4 C1 Raven P1 M4 C2 Raven P1 M4 C3 Raven P1 M4 C4 Raven P1 M4 C5 Raven P1 M5 C1 Raven P1 M5 C2 Raven P1 M5 C3 Raven P1 M5 C4 Raven P1 M5 C5 Raven P2 M1 C1 Falcon P2 M1 C2 Falcon P2 M1 C3 Falcon P2 M1 C4 Falcon P2 M1 C5 Falcon P2 M2 C1 Falcon P2 M2 C2 Falcon P2 M2 C3 Falcon P2 M2 C4 Falcon P2 M2 C5 Falcon

Marketing Optimization
Historical Value Total Leads 35 90 64 30 63 84 73 52 55 54 62 74 57 61 63 40 73 56 55 55 46 78 68 57 49 43 48 47 49 72 67 85 67 74 73 46 64 63 74

55

J 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74

M C

Product

Market Airports Airports Airports Airports Airports Data Centers Data Centers Data Centers Data Centers Data Centers Hospitals Hospitals Hospitals Hospitals Hospitals Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports Data Centers Data Centers Data Centers Data Centers Data Centers Hospitals Hospitals Hospitals Hospitals

Channel Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email

Historical Value Targeted Leads 12 16 11 18 18 9 14 10 11 9 9 10 12 9 7 15 14 15 14 13 18 11 14 14 8 11 19 14 17 17 11 13 11 12 11 14 8 8 17

P2 M3 C1 Falcon P2 M3 C2 Falcon P2 M3 C3 Falcon P2 M3 C4 Falcon P2 M3 C5 Falcon P2 M4 C1 Falcon P2 M4 C2 Falcon P2 M4 C3 Falcon P2 M4 C4 Falcon P2 M4 C5 Falcon P2 M5 C1 Falcon P2 M5 C2 Falcon P2 M5 C3 Falcon P2 M5 C4 Falcon P2 M5 C5 Falcon P3 M1 C1 Hawk P3 M1 C2 Hawk P3 M1 C3 Hawk P3 M1 C4 Hawk P3 M1 C5 Hawk P3 M2 C1 Hawk P3 M2 C2 Hawk P3 M2 C3 Hawk P3 M2 C4 Hawk P3 M2 C5 Hawk P3 M3 C1 Hawk P3 M3 C2 Hawk P3 M3 C3 Hawk P3 M3 C4 Hawk P3 M3 C5 Hawk P3 M4 C1 Hawk P3 M4 C2 Hawk P3 M4 C3 Hawk P3 M4 C4 Hawk P3 M4 C5 Hawk P3 M5 C1 Hawk P3 M5 C2 Hawk P3 M5 C3 Hawk P3 M5 C4 Hawk

Marketing Optimization
Historical Value Total Leads 42 35 81 62 32 45 40 50 65 37 41 32 81 34 56 62 49 82 52 52 68 50 91 51 60 42 39 61 47 32 38 33 51 33 34 42 35 47 54

56

J 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

M C

Product

Market Hospitals Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports Data Centers Data Centers Data Centers Data Centers Data Centers Hospitals Hospitals Hospitals Hospitals Hospitals Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Airports Airports Airports

Channel Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail

Historical Value Targeted Leads 6 13 10 11 11 11 17 12 16 11 11 8 15 11 13 11 10 16 13 12 13 11 9 9 11 7 15 13 10 11 13 15 17 13 15 10 9 13 16

P3 M5 C5 Hawk P4 M1 C1 Osprey P4 M1 C2 Osprey P4 M1 C3 Osprey P4 M1 C4 Osprey P4 M1 C5 Osprey P4 M2 C1 Osprey P4 M2 C2 Osprey P4 M2 C3 Osprey P4 M2 C4 Osprey P4 M2 C5 Osprey P4 M3 C1 Osprey P4 M3 C2 Osprey P4 M3 C3 Osprey P4 M3 C4 Osprey P4 M3 C5 Osprey P4 M4 C1 Osprey P4 M4 C2 Osprey P4 M4 C3 Osprey P4 M4 C4 Osprey P4 M4 C5 Osprey P4 M5 C1 Osprey P4 M5 C2 Osprey P4 M5 C3 Osprey P4 M5 C4 Osprey

100 P4 M5 C5 Osprey 101 P5 M1 C1 Eagle 102 P5 M1 C2 Eagle 103 P5 M1 C3 Eagle 104 P5 M1 C4 Eagle 105 P5 M1 C5 Eagle 106 P5 M2 C1 Eagle 107 P5 M2 C2 Eagle 108 P5 M2 C3 Eagle 109 P5 M2 C4 Eagle 110 P5 M2 C5 Eagle 111 P5 M3 C1 Eagle 112 P5 M3 C2 Eagle 113 P5 M3 C3 Eagle

Marketing Optimization
Historical Value Total Leads 30 47 39 73 48 48 49 47 99 68 62 33

57

M C

Product

Market Airports Airports Data Centers Data Centers Data Centers Data Centers Data Centers Hospitals Hospitals Hospitals Hospitals Hospitals

Channel Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Historical Value Targeted Leads 19 12 10 12 13 12 10 15 11 13 15 8

114 P5 M3 C4 Eagle 115 P5 M3 C5 Eagle 116 P5 M4 C1 Eagle 117 P5 M4 C2 Eagle 118 P5 M4 C3 Eagle 119 P5 M4 C4 Eagle 120 P5 M4 C5 Eagle 121 P5 M5 C1 Eagle 122 P5 M5 C2 Eagle 123 P5 M5 C3 Eagle 124 P5 M5 C4 Eagle 125 P5 M5 C5 Eagle

Marketing Optimization Appendix B Complete Statistical Software Output for Total Leads

58

Marketing Optimization Appendix C Complete Statistical Software Output for Targeted Leads

59

Marketing Optimization Appendix D Table of Discrete Factor Coefficients


Total Revenue Coefficients 59.032 10.928 4.968 0.608 -5.032 -11.472 3.288 -2.832 -7.392 5.168 1.768 -7.512 16.568 -1.552 -3.832 -3.672 Construction Shipping Airports Data Centers Hospitals Construction Shipping Airports Data Centers Hospitals Construction Shipping Airports Data Centers Hospitals Construction Shipping 18.552 2.072 -8.168 -2.928 -9.528 2.312 5.832 -0.208 -5.568 -2.368 -7.128 2.792 -0.448 8.392 -3.608 -6.288 -4.568

60

Intercept P1 P2 P3 P4 P5 Main Effects M1 M2 M3 M4 M5 C1 C2 C3 C4 C5 P1 M1 P1 M2 P1 M3 P1 M4 P1 M5 P2 M1 Interactions P2 M2 P2 M3 P2 M4 P2 M5 P3 M1 P3 M2 P3 M3 P3 M4 P3 M5 P4 M1 P4 M2 Raven Falcon Hawk Osprey Eagle Construction Shipping Airports Data Centers Hospitals Print Ad Trade Show Direct Mail Email Search Engine Raven Raven Raven Raven Raven Falcon Falcon Falcon Falcon Falcon Hawk Hawk Hawk Hawk Hawk Osprey Osprey

Targeted Revenue Coefficients 12.296 0.464 -1.056 0.704 -0.616 0.504 -0.816 1.264 1.904 -0.736 -1.616 -0.176 0.944 -0.376 1.104 -1.496 -1.344 1.176 0.336 -0.624 0.456 -1.424 -0.304 1.856 0.096 -0.224 2.016 -1.264 0.696 -0.664 -0.784 0.336 0.456

Marketing Optimization
Total Revenue Coefficients P4 M3 P4 M4 P4 M5 P5 M1 P5 M2 P5 M3 P5 M4 P5 M5 C1 M1 C1 M2 C1 M3 C1 M4 C1 M5 C2 M1 C2 M2 C2 M3 C2 M4 C2 M5 C3 M1 C3 M2 C3 M3 C3 M4 C3 M5 C4 M1 C4 M2 C4 M3 C4 M4 C4 M5 C5 M1 C5 M2 C5 M3 C5 M4 C5 M5 Osprey Osprey Osprey Eagle Eagle Eagle Eagle Eagle Print Ad Print Ad Print Ad Print Ad Print Ad Trade Show Trade Show Trade Show Trade Show Trade Show Direct Mail Direct Mail Direct Mail Direct Mail Direct Mail Email Email Email Email Email Search Engine Search Engine Search Engine Search Engine Search Engine Airports Data Centers Hospitals Construction Shipping Airports Data Centers Hospitals Construction Shipping Airports Data Centers Hospitals Construction Shipping Airports Data Centers Hospitals Construction Shipping Airports Data Centers Hospitals Construction Shipping Airports Data Centers Hospitals Construction Shipping Airports Data Centers Hospitals 6.392 1.432 3.032 -7.448 -6.128 2.432 -1.328 12.472 1.992 0.112 -4.328 4.112 -1.888 3.312 -4.768 -2.408 0.232 3.632 1.432 3.152 0.312 -7.248 2.352 -2.488 -0.968 -6.008 1.832 7.632 -4.248 2.472 12.432 1.072 -11.728 Targeted Revenue Coefficients

61

-1.984 1.856 -0.664 0.416 -0.064 -0.904 -0.664 1.216 1.696 1.416 -4.024 -1.384 2.296 -0.024 0.096 1.256 0.896 -2.224 -0.304 1.016 -0.824 0.216 -0.104 -1.784 -0.864 1.496 -0.264 1.416 0.416 -1.664 2.096 0.536 -1.384

Marketing Optimization Appendix E Multivariate Testing Results for Total Lead Generation

62

J P M C Product

Market

Channel

Intercept P

PM

Historical Residuals Predicted Value (Actual) (PredictedCM (P+M+C+PM+CM) Value Actual) 86.3 111.7 91.7 85.5 83.9 61.8 81.0 70.8 64.4 68.0 42.6 68.6 53.2 44.6 63.2 68.8 89.0 63.4 70.2 69.6 52.8 82.4 63.0 66.0 46.8 64.1 89.5 69.5 63.3 61.7 59.6 78.8 68.6 62.2 86 110 98 89 76 60 76 73 70 67 54 63 53 44 58 65 92 58 82 64 52 77 69 66 47 84 86 48 72 58 65 85 50 59 0.28 1.68 -6.32 -3.52 7.88 1.8 5 -2.2 -5.6 1 -11.44 5.56 0.16 0.56 5.16 3.8 -3 5.4 -11.8 5.6 0.8 5.4 -6 0 -0.2 -19.92 3.48 21.48 -8.72 3.68 -5.4 -6.2 18.6 3.2

1 P1 M1 C1 Raven Construction Print Ad 2 P1 M1 C2 Raven Construction Trade Show 3 P1 M1 C3 Raven Construction Direct Mail 4 P1 M1 C4 Raven Construction Email 5 P1 M1 C5 Raven Construction Search Engine 6 P1 M2 C1 Raven Shipping 7 P1 M2 C2 Raven Shipping 8 P1 M2 C3 Raven Shipping 9 P1 M2 C4 Raven Shipping 10 P1 M2 C5 Raven Shipping 11 P1 M3 C1 Raven Airports 12 P1 M3 C2 Raven Airports 13 P1 M3 C3 Raven Airports 14 P1 M3 C4 Raven Airports 15 P1 M3 C5 Raven Airports Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

59.03 10.93 3.29 -7.5118.55 1.99 59.03 10.93 3.2916.5718.55 3.31 59.03 10.93 3.29 -1.5518.55 1.43 59.03 10.93 3.29 -3.8318.55 -2.49 59.03 10.93 3.29 -3.6718.55 -4.25 59.03 10.93 -2.83 -7.51 2.07 0.11 59.03 10.93 -2.8316.57 2.07 -4.77 59.03 10.93 -2.83 -1.55 2.07 3.15 59.03 10.93 -2.83 -3.83 2.07 -0.97 59.03 10.93 -2.83 -3.67 2.07 2.47 59.03 10.93 -7.39 -7.51 -8.17 -4.33 59.03 10.93 -7.3916.57 -8.17 -2.41 59.03 10.93 -7.39 -1.55 -8.17 0.31 59.03 10.93 -7.39 -3.83 -8.17 -6.01 59.03 10.93 -7.39 -3.67 -8.17 12.43 59.03 10.93 5.17 -7.51 -2.93 4.11 59.03 10.93 5.1716.57 -2.93 0.23 59.03 10.93 5.17 -1.55 -2.93 -7.25 59.03 10.93 5.17 -3.83 -2.93 1.83 59.03 10.93 5.17 -3.67 -2.93 1.07 59.03 10.93 1.77 -7.51 -9.53 -1.89 59.03 10.93 1.7716.57 -9.53 3.63 59.03 10.93 1.77 -1.55 -9.53 2.35 59.03 10.93 1.77 -3.83 -9.53 7.63 59.03 10.93 1.77 -3.67 -9.53-11.73 59.03 4.97 3.29 -7.51 2.31 1.99 59.03 4.97 3.2916.57 2.31 3.31 59.03 4.97 3.29 -1.55 2.31 1.43 59.03 4.97 3.29 -3.83 2.31 -2.49 59.03 4.97 3.29 -3.67 2.31 -4.25 59.03 4.97 -2.83 -7.51 5.83 0.11 59.03 4.97 -2.8316.57 5.83 -4.77 59.03 4.97 -2.83 -1.55 5.83 3.15 59.03 4.97 -2.83 -3.83 5.83 -0.97

16 P1 M4 C1 Raven Data CentersPrint Ad 17 P1 M4 C2 Raven Data CentersTrade Show 18 P1 M4 C3 Raven Data CentersDirect Mail 19 P1 M4 C4 Raven Data CentersEmail 20 P1 M4 C5 Raven Data CentersSearch Engine 21 P1 M5 C1 Raven Hospitals 22 P1 M5 C2 Raven Hospitals 23 P1 M5 C3 Raven Hospitals 24 P1 M5 C4 Raven Hospitals 25 P1 M5 C5 Raven Hospitals Print Ad Trade Show Direct Mail Email Search Engine

26 P2 M1 C1 Falcon Construction Print Ad 27 P2 M1 C2 Falcon Construction Trade Show 28 P2 M1 C3 Falcon Construction Direct Mail 29 P2 M1 C4 Falcon Construction Email 30 P2 M1 C5 Falcon Construction Search Engine 31 P2 M2 C1 Falcon Shipping 32 P2 M2 C2 Falcon Shipping 33 P2 M2 C3 Falcon Shipping 34 P2 M2 C4 Falcon Shipping Print Ad Trade Show Direct Mail Email

Marketing Optimization

63

J P M C Product

Market

Channel Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Intercept P

PM

Historical Residuals Predicted Value (Actual) (PredictedActual) CM (P+M+C+PM+CM) Value 65.8 44.6 70.6 55.2 46.6 65.2 60.2 80.4 54.8 61.6 61.0 54.0 83.6 64.2 67.2 48.0 50.3 75.7 55.7 49.5 47.9 52.2 71.4 61.2 54.8 58.4 40.0 66.0 50.6 42.0 60.6 69.8 90.0 64.4 71.2 70.6 48.4 78.0 76 35 90 64 30 63 84 73 52 55 54 62 74 57 61 63 40 73 56 55 55 46 78 68 57 49 43 48 47 49 72 67 85 67 74 73 46 64 -10.2 9.56 -19.44 -8.84 16.56 2.16 -23.8 7.4 2.8 6.6 7 -8 9.6 7.2 6.2 -15 10.28 2.68 -0.32 -5.52 -7.12 6.2 -6.6 -6.8 -2.2 9.4 -3.04 17.96 3.56 -7.04 -11.44 2.8 5 -2.6 -2.8 -2.4 2.4 14

35 P2 M2 C5 Falcon Shipping 36 P2 M3 C1 Falcon Airports 37 P2 M3 C2 Falcon Airports 38 P2 M3 C3 Falcon Airports 39 P2 M3 C4 Falcon Airports 40 P2 M3 C5 Falcon Airports

59.03 4.97 -2.83 -3.67 5.83 2.47 59.03 4.97 -7.39 -7.51 -0.21 -4.33 59.03 4.97 -7.3916.57 -0.21 -2.41 59.03 4.97 -7.39 -1.55 -0.21 0.31 59.03 4.97 -7.39 -3.83 -0.21 -6.01 59.03 4.97 -7.39 -3.67 -0.21 12.43 59.03 4.97 5.17 -7.51 -5.57 4.11 59.03 4.97 5.1716.57 -5.57 0.23 59.03 4.97 5.17 -1.55 -5.57 -7.25 59.03 4.97 5.17 -3.83 -5.57 1.83 59.03 4.97 5.17 -3.67 -5.57 1.07 59.03 4.97 1.77 -7.51 -2.37 -1.89 59.03 4.97 1.7716.57 -2.37 3.63 59.03 4.97 1.77 -1.55 -2.37 2.35 59.03 4.97 1.77 -3.83 -2.37 7.63 59.03 4.97 1.77 -3.67 -2.37-11.73 59.03 0.61 3.29 -7.51 -7.13 1.99 59.03 0.61 3.2916.57 -7.13 3.31 59.03 0.61 3.29 -1.55 -7.13 1.43 59.03 0.61 3.29 -3.83 -7.13 -2.49 59.03 0.61 3.29 -3.67 -7.13 -4.25 59.03 0.61 -2.83 -7.51 2.79 0.11 59.03 0.61 -2.8316.57 2.79 -4.77 59.03 0.61 -2.83 -1.55 2.79 3.15 59.03 0.61 -2.83 -3.83 2.79 -0.97 59.03 0.61 -2.83 -3.67 2.79 2.47 59.03 0.61 -7.39 -7.51 -0.45 -4.33 59.03 0.61 -7.3916.57 -0.45 -2.41 59.03 0.61 -7.39 -1.55 -0.45 0.31 59.03 0.61 -7.39 -3.83 -0.45 -6.01 59.03 0.61 -7.39 -3.67 -0.45 12.43 59.03 0.61 5.17 -7.51 8.39 4.11 59.03 0.61 5.1716.57 8.39 0.23 59.03 0.61 5.17 -1.55 8.39 -7.25 59.03 0.61 5.17 -3.83 8.39 1.83 59.03 0.61 5.17 -3.67 8.39 1.07 59.03 0.61 1.77 -7.51 -3.61 -1.89 59.03 0.61 1.7716.57 -3.61 3.63

41 P2 M4 C1 Falcon Data CentersPrint Ad 42 P2 M4 C2 Falcon Data CentersTrade Show 43 P2 M4 C3 Falcon Data CentersDirect Mail 44 P2 M4 C4 Falcon Data CentersEmail 45 P2 M4 C5 Falcon Data CentersSearch Engine 46 P2 M5 C1 Falcon Hospitals 47 P2 M5 C2 Falcon Hospitals 48 P2 M5 C3 Falcon Hospitals 49 P2 M5 C4 Falcon Hospitals 50 P2 M5 C5 Falcon Hospitals 51 P3 M1 C1 Hawk 52 P3 M1 C2 Hawk 53 P3 M1 C3 Hawk 54 P3 M1 C4 Hawk 55 P3 M1 C5 Hawk 56 P3 M2 C1 Hawk 57 P3 M2 C2 Hawk 58 P3 M2 C3 Hawk 59 P3 M2 C4 Hawk 60 P3 M2 C5 Hawk 61 P3 M3 C1 Hawk 62 P3 M3 C2 Hawk 63 P3 M3 C3 Hawk 64 P3 M3 C4 Hawk 65 P3 M3 C5 Hawk 66 P3 M4 C1 Hawk 67 P3 M4 C2 Hawk 68 P3 M4 C3 Hawk 69 P3 M4 C4 Hawk 70 P3 M4 C5 Hawk 71 P3 M5 C1 Hawk 72 P3 M5 C2 Hawk Print Ad Trade Show Direct Mail Email Search Engine

Construction Print Ad Construction Trade Show Construction Direct Mail Construction Email Construction Search Engine Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Data CentersPrint Ad Data CentersTrade Show Data CentersDirect Mail Data CentersEmail Data CentersSearch Engine Hospitals Hospitals Print Ad Trade Show

Marketing Optimization

64

J P M C Product 73 P3 M5 C3 Hawk 74 P3 M5 C4 Hawk 75 P3 M5 C5 Hawk

Market Hospitals Hospitals Hospitals

Channel Direct Mail Email Search Engine

Intercept P

PM

Historical Residuals Predicted Value (Actual) (PredictedActual) CM (P+M+C+PM+CM) Value 58.6 61.6 42.4 45.5 70.9 50.9 44.7 43.1 39.2 58.4 48.2 41.8 45.4 41.2 67.2 51.8 43.2 61.8 57.2 77.4 51.8 58.6 58.0 49.4 79.0 59.6 62.6 43.4 37.9 63.3 43.3 37.1 35.5 31.2 50.4 40.2 33.8 37.4 63 74 42 35 81 62 32 45 40 50 65 37 41 32 81 34 56 62 49 82 52 52 68 50 91 51 60 42 39 61 47 32 38 33 51 33 34 42 -4.4 -12.4 0.4 10.48 -10.12 -11.12 12.68 -1.92 -0.8 8.4 -16.8 4.8 4.4 9.16 -13.84 17.76 -12.84 -0.24 8.2 -4.6 -0.2 6.6 -10 -0.6 -12 8.6 2.6 1.4 -1.12 2.28 -3.72 5.08 -2.52 -1.8 -0.6 7.2 -0.2 -4.6

59.03 0.61 1.77 -1.55 -3.61 2.35 59.03 0.61 1.77 -3.83 -3.61 7.63 59.03 0.61 1.77 -3.67 -3.61-11.73 59.03 -5.03 3.29 -7.51 -6.29 1.99 59.03 -5.03 3.2916.57 -6.29 3.31 59.03 -5.03 3.29 -1.55 -6.29 1.43 59.03 -5.03 3.29 -3.83 -6.29 -2.49 59.03 -5.03 3.29 -3.67 -6.29 -4.25 59.03 -5.03 -2.83 -7.51 -4.57 0.11 59.03 -5.03 -2.8316.57 -4.57 -4.77 59.03 -5.03 -2.83 -1.55 -4.57 3.15 59.03 -5.03 -2.83 -3.83 -4.57 -0.97 59.03 -5.03 -2.83 -3.67 -4.57 2.47 59.03 -5.03 -7.39 -7.51 6.39 -4.33 59.03 -5.03 -7.3916.57 6.39 -2.41 59.03 -5.03 -7.39 -1.55 6.39 0.31 59.03 -5.03 -7.39 -3.83 6.39 -6.01 59.03 -5.03 -7.39 -3.67 6.39 12.43 59.03 -5.03 5.17 -7.51 1.43 4.11 59.03 -5.03 5.1716.57 1.43 0.23 59.03 -5.03 5.17 -1.55 1.43 -7.25 59.03 -5.03 5.17 -3.83 1.43 1.83 59.03 -5.03 5.17 -3.67 1.43 1.07 59.03 -5.03 1.77 -7.51 3.03 -1.89 59.03 -5.03 1.7716.57 3.03 3.63 59.03 -5.03 1.77 -1.55 3.03 2.35 59.03 -5.03 1.77 -3.83 3.03 7.63 59.03 -5.03 1.77 -3.67 3.03-11.73 59.03-11.47 3.29 -7.51 -7.45 1.99 59.03-11.47 3.2916.57 -7.45 3.31 59.03-11.47 3.29 -1.55 -7.45 1.43 59.03-11.47 3.29 -3.83 -7.45 -2.49 59.03-11.47 3.29 -3.67 -7.45 -4.25 59.03-11.47 -2.83 -7.51 -6.13 0.11 59.03-11.47 -2.8316.57 -6.13 -4.77 59.03-11.47 -2.83 -1.55 -6.13 3.15 59.03-11.47 -2.83 -3.83 -6.13 -0.97 59.03-11.47 -2.83 -3.67 -6.13 2.47

76 P4 M1 C1 Osprey Construction Print Ad 77 P4 M1 C2 Osprey Construction Trade Show 78 P4 M1 C3 Osprey Construction Direct Mail 79 P4 M1 C4 Osprey Construction Email 80 P4 M1 C5 Osprey Construction Search Engine 81 P4 M2 C1 Osprey Shipping 82 P4 M2 C2 Osprey Shipping 83 P4 M2 C3 Osprey Shipping 84 P4 M2 C4 Osprey Shipping 85 P4 M2 C5 Osprey Shipping 86 P4 M3 C1 Osprey Airports 87 P4 M3 C2 Osprey Airports 88 P4 M3 C3 Osprey Airports 89 P4 M3 C4 Osprey Airports 90 P4 M3 C5 Osprey Airports Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

91 P4 M4 C1 Osprey Data CentersPrint Ad 92 P4 M4 C2 Osprey Data CentersTrade Show 93 P4 M4 C3 Osprey Data CentersDirect Mail 94 P4 M4 C4 Osprey Data CentersEmail 95 P4 M4 C5 Osprey Data CentersSearch Engine 96 P4 M5 C1 Osprey Hospitals 97 P4 M5 C2 Osprey Hospitals 98 P4 M5 C3 Osprey Hospitals 99 P4 M5 C4 Osprey Hospitals 100 P4 M5 C5 Osprey Hospitals 101 P5 M1 C1 Eagle 102 P5 M1 C2 Eagle 103 P5 M1 C3 Eagle 104 P5 M1 C4 Eagle 105 P5 M1 C5 Eagle 106 P5 M2 C1 Eagle 107 P5 M2 C2 Eagle 108 P5 M2 C3 Eagle 109 P5 M2 C4 Eagle 110 P5 M2 C5 Eagle Print Ad Trade Show Direct Mail Email Search Engine

Construction Print Ad Construction Trade Show Construction Direct Mail Construction Email Construction Search Engine Shipping Shipping Shipping Shipping Shipping Print Ad Trade Show Direct Mail Email Search Engine

Marketing Optimization

65

J P M C Product 111 P5 M3 C1 Eagle 112 P5 M3 C2 Eagle 113 P5 M3 C3 Eagle 114 P5 M3 C4 Eagle 115 P5 M3 C5 Eagle 116 P5 M4 C1 Eagle 117 P5 M4 C2 Eagle 118 P5 M4 C3 Eagle 119 P5 M4 C4 Eagle 120 P5 M4 C5 Eagle 121 P5 M5 C1 Eagle 122 P5 M5 C2 Eagle 123 P5 M5 C3 Eagle 124 P5 M5 C4 Eagle 125 P5 M5 C5 Eagle

Market Airports Airports Airports Airports Airports

Channel Print Ad Trade Show Direct Mail Email Search Engine

Intercept P

PM

Historical Residuals Predicted Value (Actual) (PredictedActual) CM (P+M+C+PM+CM) Value 30.8 56.8 41.4 32.8 51.4 48.0 68.2 42.6 49.4 48.8 52.4 82.0 62.6 65.6 46.4 35 47 54 30 47 39 73 48 48 49 47 99 68 62 33 -4.24 9.76 -12.64 2.76 4.36 9 -4.8 -5.4 1.4 -0.2 5.4 -17 -5.4 3.6 13.4

59.03-11.47 -7.39 -7.51 2.43 -4.33 59.03-11.47 -7.3916.57 2.43 -2.41 59.03-11.47 -7.39 -1.55 2.43 0.31 59.03-11.47 -7.39 -3.83 2.43 -6.01 59.03-11.47 -7.39 -3.67 2.43 12.43 59.03-11.47 5.17 -7.51 -1.33 4.11 59.03-11.47 5.1716.57 -1.33 0.23 59.03-11.47 5.17 -1.55 -1.33 -7.25 59.03-11.47 5.17 -3.83 -1.33 1.83 59.03-11.47 5.17 -3.67 -1.33 1.07 59.03-11.47 1.77 -7.5112.47 -1.89 59.03-11.47 1.7716.5712.47 3.63 59.03-11.47 1.77 -1.5512.47 2.35 59.03-11.47 1.77 -3.8312.47 7.63 59.03-11.47 1.77 -3.6712.47-11.73

Data CentersPrint Ad Data CentersTrade Show Data CentersDirect Mail Data CentersEmail Data CentersSearch Engine Hospitals Hospitals Hospitals Hospitals Hospitals Print Ad Trade Show Direct Mail Email Search Engine

Marketing Optimization Appendix F Multivariate Testing Results for Targeted Lead Generation

66

J P M C Product

Market

Channel

Intercept P

Historical Residuals Predicted Value (Actual) (PredictedPM CM (P+M+C+PM+CM) Value Actual) 12.1 11.5 9.9 9.9 9.5 16.4 16.2 15.8 15.4 12.0 10.8 17.2 13.8 17.6 15.6 9.8 13.2 11.2 12.2 10.4 13.7 10.3 11.1 14.1 8.7 10.5 9.9 8.3 8.3 7.9 13.4 13.2 12.8 12.4 11 13 11 10 8 12 20 16 16 12 10 19 13 17 16 10 12 10 15 10 15 9 9 14 11 11 12 7 8 7 12 13 12 13 1.12 -1.48 -1.08 -0.08 1.52 4.44 -3.76 -0.16 -0.56 0.04 0.8 -1.8 0.8 0.6 -0.4 -0.16 1.24 1.24 -2.76 0.44 -1.28 1.32 2.12 0.12 -2.28 -0.48 -2.08 1.32 0.32 0.92 1.44 0.24 0.84 -0.56

1 P1 M1 C1 Raven Construction Print Ad 2 P1 M1 C2 Raven Construction Trade Show 3 P1 M1 C3 Raven Construction Direct Mail 4 P1 M1 C4 Raven Construction Email 5 P1 M1 C5 Raven Construction Search Engine 6 P1 M2 C1 Raven Shipping 7 P1 M2 C2 Raven Shipping 8 P1 M2 C3 Raven Shipping 9 P1 M2 C4 Raven Shipping 10 P1 M2 C5 Raven Shipping 11 P1 M3 C1 Raven Airports 12 P1 M3 C2 Raven Airports 13 P1 M3 C3 Raven Airports 14 P1 M3 C4 Raven Airports 15 P1 M3 C5 Raven Airports Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

12.30 0.46-0.82-0.18 -1.34 1.70 12.30 0.46-0.82 0.94 -1.34 -0.02 12.30 0.46-0.82-0.38 -1.34 -0.30 12.30 0.46-0.82 1.10 -1.34 -1.78 12.30 0.46-0.82-1.50 -1.34 0.42 12.30 0.46 1.26-0.18 1.18 1.42 12.30 0.46 1.26 0.94 1.18 0.10 12.30 0.46 1.26-0.38 1.18 1.02 12.30 0.46 1.26 1.10 1.18 -0.86 12.30 0.46 1.26-1.50 1.18 -1.66 12.30 0.46 1.90-0.18 0.34 -4.02 12.30 0.46 1.90 0.94 0.34 1.26 12.30 0.46 1.90-0.38 0.34 -0.82 12.30 0.46 1.90 1.10 0.34 1.50 12.30 0.46 1.90-1.50 0.34 2.10 12.30 0.46-0.74-0.18 -0.62 -1.38 12.30 0.46-0.74 0.94 -0.62 0.90 12.30 0.46-0.74-0.38 -0.62 0.22 12.30 0.46-0.74 1.10 -0.62 -0.26 12.30 0.46-0.74-1.50 -0.62 0.54 12.30 0.46-1.62-0.18 0.46 2.30 12.30 0.46-1.62 0.94 0.46 -2.22 12.30 0.46-1.62-0.38 0.46 -0.10 12.30 0.46-1.62 1.10 0.46 1.42 12.30 0.46-1.62-1.50 0.46 -1.38 12.30-1.06-0.82-0.18 -1.42 1.70 12.30-1.06-0.82 0.94 -1.42 -0.02 12.30-1.06-0.82-0.38 -1.42 -0.30 12.30-1.06-0.82 1.10 -1.42 -1.78 12.30-1.06-0.82-1.50 -1.42 0.42 12.30-1.06 1.26-0.18 -0.30 1.42 12.30-1.06 1.26 0.94 -0.30 0.10 12.30-1.06 1.26-0.38 -0.30 1.02 12.30-1.06 1.26 1.10 -0.30 -0.86

16 P1 M4 C1 Raven Data Centers Print Ad 17 P1 M4 C2 Raven Data Centers Trade Show 18 P1 M4 C3 Raven Data Centers Direct Mail 19 P1 M4 C4 Raven Data Centers Email 20 P1 M4 C5 Raven Data Centers Search Engine 21 P1 M5 C1 Raven Hospitals 22 P1 M5 C2 Raven Hospitals 23 P1 M5 C3 Raven Hospitals 24 P1 M5 C4 Raven Hospitals 25 P1 M5 C5 Raven Hospitals Print Ad Trade Show Direct Mail Email Search Engine

26 P2 M1 C1 Falcon Construction Print Ad 27 P2 M1 C2 Falcon Construction Trade Show 28 P2 M1 C3 Falcon Construction Direct Mail 29 P2 M1 C4 Falcon Construction Email 30 P2 M1 C5 Falcon Construction Search Engine 31 P2 M2 C1 Falcon Shipping 32 P2 M2 C2 Falcon Shipping 33 P2 M2 C3 Falcon Shipping 34 P2 M2 C4 Falcon Shipping Print Ad Trade Show Direct Mail Email

Marketing Optimization

67

J P M C Product

Market

Channel Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Intercept P

Historical Residuals Predicted Value (Actual) (PredictedActual) PM CM (P+M+C+PM+CM) Value 9.0 10.8 17.2 13.8 17.6 15.6 9.0 12.4 10.4 11.4 9.6 11.5 8.1 8.9 11.9 6.5 15.7 15.1 13.5 13.5 13.1 14.2 14.0 13.6 13.2 9.8 11.4 17.8 14.4 18.2 16.2 10.0 13.4 11.4 12.4 10.6 12.7 9.3 11 12 16 11 18 18 9 14 10 11 9 9 10 12 9 7 15 14 15 14 13 18 11 14 14 8 11 19 14 17 17 11 13 11 12 11 14 8 -1.96 -1.2 1.2 2.8 -0.4 -2.4 0.04 -1.56 0.44 0.44 0.64 2.52 -1.88 -3.08 2.92 -0.48 0.72 1.12 -1.48 -0.48 0.12 -3.76 3.04 -0.36 -0.76 1.84 0.4 -1.2 0.4 1.2 -0.8 -0.96 0.44 0.44 0.44 -0.36 -1.28 1.32

35 P2 M2 C5 Falcon Shipping 36 P2 M3 C1 Falcon Airports 37 P2 M3 C2 Falcon Airports 38 P2 M3 C3 Falcon Airports 39 P2 M3 C4 Falcon Airports 40 P2 M3 C5 Falcon Airports

12.30-1.06 1.26-1.50 -0.30 -1.66 12.30-1.06 1.90-0.18 1.86 -4.02 12.30-1.06 1.90 0.94 1.86 1.26 12.30-1.06 1.90-0.38 1.86 -0.82 12.30-1.06 1.90 1.10 1.86 1.50 12.30-1.06 1.90-1.50 1.86 2.10 12.30-1.06-0.74-0.18 0.10 -1.38 12.30-1.06-0.74 0.94 0.10 0.90 12.30-1.06-0.74-0.38 0.10 0.22 12.30-1.06-0.74 1.10 0.10 -0.26 12.30-1.06-0.74-1.50 0.10 0.54 12.30-1.06-1.62-0.18 -0.22 2.30 12.30-1.06-1.62 0.94 -0.22 -2.22 12.30-1.06-1.62-0.38 -0.22 -0.10 12.30-1.06-1.62 1.10 -0.22 1.42 12.30-1.06-1.62-1.50 -0.22 -1.38 12.30 0.70-0.82-0.18 2.02 1.70 12.30 0.70-0.82 0.94 2.02 -0.02 12.30 0.70-0.82-0.38 2.02 -0.30 12.30 0.70-0.82 1.10 2.02 -1.78 12.30 0.70-0.82-1.50 2.02 0.42 12.30 0.70 1.26-0.18 -1.26 1.42 12.30 0.70 1.26 0.94 -1.26 0.10 12.30 0.70 1.26-0.38 -1.26 1.02 12.30 0.70 1.26 1.10 -1.26 -0.86 12.30 0.70 1.26-1.50 -1.26 -1.66 12.30 0.70 1.90-0.18 0.70 -4.02 12.30 0.70 1.90 0.94 0.70 1.26 12.30 0.70 1.90-0.38 0.70 -0.82 12.30 0.70 1.90 1.10 0.70 1.50 12.30 0.70 1.90-1.50 0.70 2.10 12.30 0.70-0.74-0.18 -0.66 -1.38 12.30 0.70-0.74 0.94 -0.66 0.90 12.30 0.70-0.74-0.38 -0.66 0.22 12.30 0.70-0.74 1.10 -0.66 -0.26 12.30 0.70-0.74-1.50 -0.66 0.54 12.30 0.70-1.62-0.18 -0.78 2.30 12.30 0.70-1.62 0.94 -0.78 -2.22

41 P2 M4 C1 Falcon Data Centers Print Ad 42 P2 M4 C2 Falcon Data Centers Trade Show 43 P2 M4 C3 Falcon Data Centers Direct Mail 44 P2 M4 C4 Falcon Data Centers Email 45 P2 M4 C5 Falcon Data Centers Search Engine 46 P2 M5 C1 Falcon Hospitals 47 P2 M5 C2 Falcon Hospitals 48 P2 M5 C3 Falcon Hospitals 49 P2 M5 C4 Falcon Hospitals 50 P2 M5 C5 Falcon Hospitals 51 P3 M1 C1 Hawk 52 P3 M1 C2 Hawk 53 P3 M1 C3 Hawk 54 P3 M1 C4 Hawk 55 P3 M1 C5 Hawk 56 P3 M2 C1 Hawk 57 P3 M2 C2 Hawk 58 P3 M2 C3 Hawk 59 P3 M2 C4 Hawk 60 P3 M2 C5 Hawk 61 P3 M3 C1 Hawk 62 P3 M3 C2 Hawk 63 P3 M3 C3 Hawk 64 P3 M3 C4 Hawk 65 P3 M3 C5 Hawk 66 P3 M4 C1 Hawk 67 P3 M4 C2 Hawk 68 P3 M4 C3 Hawk 69 P3 M4 C4 Hawk 70 P3 M4 C5 Hawk 71 P3 M5 C1 Hawk 72 P3 M5 C2 Hawk Print Ad Trade Show Direct Mail Email Search Engine

Construction Print Ad Construction Trade Show Construction Direct Mail Construction Email Construction Search Engine Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Data Centers Search Engine Hospitals Hospitals Print Ad Trade Show

Marketing Optimization

68

J P M C Product 73 P3 M5 C3 Hawk 74 P3 M5 C4 Hawk 75 P3 M5 C5 Hawk

Market Hospitals Hospitals Hospitals

Channel Direct Mail Email Search Engine

Intercept P

Historical Residuals Predicted Value (Actual) (PredictedActual) PM CM (P+M+C+PM+CM) Value 10.1 13.1 7.7 12.7 12.1 10.5 10.5 10.1 14.6 14.4 14.0 13.6 10.2 7.4 13.8 10.4 14.2 12.2 11.2 14.6 12.6 13.6 11.8 11.5 8.1 8.9 11.9 6.5 13.9 13.3 11.7 11.7 11.3 15.2 15.0 14.6 14.2 10.8 8 17 6 13 10 11 11 11 17 12 16 11 11 8 15 11 13 11 10 16 13 12 13 11 9 9 11 7 15 13 10 11 13 15 17 13 15 10 2.12 -3.88 1.72 -0.28 2.12 -0.48 -0.48 -0.88 -2.36 2.44 -1.96 2.64 -0.76 -0.6 -1.2 -0.6 1.2 1.2 1.24 -1.36 -0.36 1.64 -1.16 0.52 -0.88 -0.08 0.92 -0.48 -1.08 0.32 1.72 0.72 -1.68 0.24 -1.96 1.64 -0.76 0.84

12.30 0.70-1.62-0.38 -0.78 -0.10 12.30 0.70-1.62 1.10 -0.78 1.42 12.30 0.70-1.62-1.50 -0.78 -1.38 12.30-0.62-0.82-0.18 0.34 1.70 12.30-0.62-0.82 0.94 0.34 -0.02 12.30-0.62-0.82-0.38 0.34 -0.30 12.30-0.62-0.82 1.10 0.34 -1.78 12.30-0.62-0.82-1.50 0.34 0.42 12.30-0.62 1.26-0.18 0.46 1.42 12.30-0.62 1.26 0.94 0.46 0.10 12.30-0.62 1.26-0.38 0.46 1.02 12.30-0.62 1.26 1.10 0.46 -0.86 12.30-0.62 1.26-1.50 0.46 -1.66 12.30-0.62 1.90-0.18 -1.98 -4.02 12.30-0.62 1.90 0.94 -1.98 1.26 12.30-0.62 1.90-0.38 -1.98 -0.82 12.30-0.62 1.90 1.10 -1.98 1.50 12.30-0.62 1.90-1.50 -1.98 2.10 12.30-0.62-0.74-0.18 1.86 -1.38 12.30-0.62-0.74 0.94 1.86 0.90 12.30-0.62-0.74-0.38 1.86 0.22 12.30-0.62-0.74 1.10 1.86 -0.26 12.30-0.62-0.74-1.50 1.86 0.54 12.30-0.62-1.62-0.18 -0.66 2.30 12.30-0.62-1.62 0.94 -0.66 -2.22 12.30-0.62-1.62-0.38 -0.66 -0.10 12.30-0.62-1.62 1.10 -0.66 1.42 12.30-0.62-1.62-1.50 -0.66 -1.38 12.30 0.50-0.82-0.18 0.42 1.70 12.30 0.50-0.82 0.94 0.42 -0.02 12.30 0.50-0.82-0.38 0.42 -0.30 12.30 0.50-0.82 1.10 0.42 -1.78 12.30 0.50-0.82-1.50 0.42 0.42 12.30 0.50 1.26-0.18 -0.06 1.42 12.30 0.50 1.26 0.94 -0.06 0.10 12.30 0.50 1.26-0.38 -0.06 1.02 12.30 0.50 1.26 1.10 -0.06 -0.86 12.30 0.50 1.26-1.50 -0.06 -1.66

76 P4 M1 C1 Osprey Construction Print Ad 77 P4 M1 C2 Osprey Construction Trade Show 78 P4 M1 C3 Osprey Construction Direct Mail 79 P4 M1 C4 Osprey Construction Email 80 P4 M1 C5 Osprey Construction Search Engine 81 P4 M2 C1 Osprey Shipping 82 P4 M2 C2 Osprey Shipping 83 P4 M2 C3 Osprey Shipping 84 P4 M2 C4 Osprey Shipping 85 P4 M2 C5 Osprey Shipping 86 P4 M3 C1 Osprey Airports 87 P4 M3 C2 Osprey Airports 88 P4 M3 C3 Osprey Airports 89 P4 M3 C4 Osprey Airports 90 P4 M3 C5 Osprey Airports Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

91 P4 M4 C1 Osprey Data Centers Print Ad 92 P4 M4 C2 Osprey Data Centers Trade Show 93 P4 M4 C3 Osprey Data Centers Direct Mail 94 P4 M4 C4 Osprey Data Centers Email 95 P4 M4 C5 Osprey Data Centers Search Engine 96 P4 M5 C1 Osprey Hospitals 97 P4 M5 C2 Osprey Hospitals 98 P4 M5 C3 Osprey Hospitals 99 P4 M5 C4 Osprey Hospitals 100 P4 M5 C5 Osprey Hospitals 101 P5 M1 C1 Eagle 102 P5 M1 C2 Eagle 103 P5 M1 C3 Eagle 104 P5 M1 C4 Eagle 105 P5 M1 C5 Eagle 106 P5 M2 C1 Eagle 107 P5 M2 C2 Eagle 108 P5 M2 C3 Eagle 109 P5 M2 C4 Eagle 110 P5 M2 C5 Eagle Print Ad Trade Show Direct Mail Email Search Engine

Construction Print Ad Construction Trade Show Construction Direct Mail Construction Email Construction Search Engine Shipping Shipping Shipping Shipping Shipping Print Ad Trade Show Direct Mail Email Search Engine

Marketing Optimization

69

J P M C Product 111 P5 M3 C1 Eagle 112 P5 M3 C2 Eagle 113 P5 M3 C3 Eagle 114 P5 M3 C4 Eagle 115 P5 M3 C5 Eagle 116 P5 M4 C1 Eagle 117 P5 M4 C2 Eagle 118 P5 M4 C3 Eagle 119 P5 M4 C4 Eagle 120 P5 M4 C5 Eagle 121 P5 M5 C1 Eagle 122 P5 M5 C2 Eagle 123 P5 M5 C3 Eagle 124 P5 M5 C4 Eagle 125 P5 M5 C5 Eagle

Market Airports Airports Airports Airports Airports

Channel Print Ad Trade Show Direct Mail Email Search Engine

Intercept P

Historical Residuals Predicted Value (Actual) (PredictedActual) PM CM (P+M+C+PM+CM) Value 9.6 16.0 12.6 16.4 14.4 9.8 13.2 11.2 12.2 10.4 14.5 11.1 11.9 14.9 9.5 9 13 16 19 12 10 12 13 12 10 15 11 13 15 8 0.6 3 -3.4 -2.6 2.4 -0.16 1.24 -1.76 0.24 0.44 -0.48 0.12 -1.08 -0.08 1.52

12.30 0.50 1.90-0.18 -0.90 -4.02 12.30 0.50 1.90 0.94 -0.90 1.26 12.30 0.50 1.90-0.38 -0.90 -0.82 12.30 0.50 1.90 1.10 -0.90 1.50 12.30 0.50 1.90-1.50 -0.90 2.10 12.30 0.50-0.74-0.18 -0.66 -1.38 12.30 0.50-0.74 0.94 -0.66 0.90 12.30 0.50-0.74-0.38 -0.66 0.22 12.30 0.50-0.74 1.10 -0.66 -0.26 12.30 0.50-0.74-1.50 -0.66 0.54 12.30 0.50-1.62-0.18 1.22 2.30 12.30 0.50-1.62 0.94 1.22 -2.22 12.30 0.50-1.62-0.38 1.22 -0.10 12.30 0.50-1.62 1.10 1.22 1.42 12.30 0.50-1.62-1.50 1.22 -1.38

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Data Centers Search Engine Hospitals Hospitals Hospitals Hospitals Hospitals Print Ad Trade Show Direct Mail Email Search Engine

Marketing Optimization Appendix G Indexed Values for Total Lead Generation

70

J P M C Product

Market

Channel

Product Expected List Revenue Index Channel Price Predicted Conversion Per Unit Efficiency Value Cost (K) (W) Leads (L) Rate (V) (W*D*E=R) (R/K=F) (F/=XR) $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 86.3 111.7 91.7 85.5 83.9 61.8 81.0 70.8 64.4 68.0 42.6 68.6 53.2 44.6 63.2 68.8 89.0 63.4 70.2 69.6 52.8 82.4 63.0 66.0 46.8 64.1 89.5 69.5 63.3 61.7 59.6 78.8 68.6 62.2 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 $572,934 $741,600 $608,792 $567,621 $556,997 $410,377 $537,872 $470,140 $427,642 $451,547 $282,615 $455,266 $353,004 $295,896 $419,408 $456,860 $590,996 $421,001 $466,156 $462,172 $350,613 $547,169 $418,345 $438,266 $310,771 $564,849 $788,744 $612,449 $557,797 $543,694 $525,359 $694,602 $604,692 $548,277 52.08 39.03 64.08 78.29 69.62 37.31 28.31 49.49 58.99 56.44 25.69 23.96 37.16 40.81 52.43 41.53 31.11 44.32 64.30 57.77 31.87 28.80 44.04 60.45 38.85 51.35 41.51 64.47 76.94 67.96 47.76 36.56 63.65 75.62 1.00 0.75 1.23 1.50 1.33 0.71 0.54 0.95 1.13 1.08 0.49 0.46 0.71 0.78 1.00 0.80 0.60 0.85 1.23 1.11 0.61 0.55 0.84 1.16 0.74 0.98 0.79 1.23 1.47 1.30 0.91 0.70 1.22 1.45

1 P1 M1 C1 Raven Construction Print Ad 2 P1 M1 C2 Raven Construction Trade Show 3 P1 M1 C3 Raven Construction Direct Mail 4 P1 M1 C4 Raven Construction Email 6 P1 M2 C1 Raven Shipping 7 P1 M2 C2 Raven Shipping 8 P1 M2 C3 Raven Shipping 9 P1 M2 C4 Raven Shipping 10 P1 M2 C5 Raven Shipping 11 P1 M3 C1 Raven Airports 12 P1 M3 C2 Raven Airports 13 P1 M3 C3 Raven Airports 14 P1 M3 C4 Raven Airports 15 P1 M3 C5 Raven Airports Print Ad Trade Show Direct Mail Email Print Ad Trade Show Direct Mail Email

5 P1 M1 C5 Raven Construction Search Engine $8,000 $25,540

Search Engine $8,000 $25,540

Search Engine $8,000 $25,540

16 P1 M4 C1 Raven Data Centers Print Ad 17 P1 M4 C2 Raven Data Centers Trade Show 18 P1 M4 C3 Raven Data Centers Direct Mail 19 P1 M4 C4 Raven Data Centers Email 21 P1 M5 C1 Raven Hospitals 22 P1 M5 C2 Raven Hospitals 23 P1 M5 C3 Raven Hospitals 24 P1 M5 C4 Raven Hospitals 25 P1 M5 C5 Raven Hospitals Print Ad Trade Show Direct Mail Email

20 P1 M4 C5 Raven Data Centers Search Engine $8,000 $25,540

Search Engine $8,000 $25,540

26 P2 M1 C1 Falcon Construction Print Ad 27 P2 M1 C2 Falcon Construction Trade Show 28 P2 M1 C3 Falcon Construction Direct Mail 29 P2 M1 C4 Falcon Construction Email 31 P2 M2 C1 Falcon Shipping 32 P2 M2 C2 Falcon Shipping 33 P2 M2 C3 Falcon Shipping 34 P2 M2 C4 Falcon Shipping Print Ad Trade Show Direct Mail Email

30 P2 M1 C5 Falcon Construction Search Engine $8,000 $38,325

Marketing Optimization

71

J P M C Product

Market

Channel Print Ad Trade Show Direct Mail Email

Product Expected List Revenue Index Channel Price Predicted Conversion Per Unit Efficiency Value Cost (K) (W) Leads (L) Rate (V) (W*D*E=R) (R/K=F) (F/=XR) 65.8 44.6 70.6 55.2 46.6 65.2 60.2 80.4 54.8 61.6 61.0 54.0 83.6 64.2 67.2 48.0 50.3 75.7 55.7 49.5 47.9 52.2 71.4 61.2 54.8 58.4 40.0 66.0 50.6 42.0 60.6 69.8 90.0 64.4 71.2 70.6 48.4 78.0 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 $580,011 $392,785 $621,969 $486,222 $410,415 $574,369 $530,648 $708,706 $483,048 $542,989 $537,700 $475,997 $736,913 $565,907 $592,351 $423,108 $477,087 $718,097 $528,325 $469,496 $454,314 $495,305 $677,486 $580,702 $519,975 $554,134 $379,164 $625,868 $479,744 $398,142 $574,630 $662,304 $853,974 $611,066 $675,588 $669,895 $459,248 $740,111 72.50 35.71 32.74 51.18 56.61 71.80 48.24 37.30 50.85 74.89 67.21 43.27 38.78 59.57 81.70 52.89 43.37 37.79 55.61 64.76 56.79 45.03 35.66 61.13 71.72 69.27 34.47 32.94 50.50 54.92 71.83 60.21 44.95 64.32 93.18 83.74 41.75 38.95 1.39 0.68 0.63 0.98 1.08 1.37 0.92 0.71 0.97 1.43 1.29 0.83 0.74 1.14 1.56 1.01 0.83 0.72 1.06 1.24 1.09 0.86 0.68 1.17 1.37 1.33 0.66 0.63 0.97 1.05 1.38 1.15 0.86 1.23 1.78 1.60 0.80 0.75 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 $11,000 $49,940 $19,000 $49,940 $9,500 $49,940 $7,250 $49,940 $11,000 $49,940 $19,000 $49,940 $9,500 $49,940 $7,250 $49,940 $11,000 $49,940 $19,000 $49,940 $9,500 $49,940 $7,250 $49,940 $11,000 $49,940 $19,000 $49,940 $9,500 $49,940 $7,250 $49,940 $11,000 $49,940 $19,000 $49,940

35 P2 M2 C5 Falcon Shipping 36 P2 M3 C1 Falcon Airports 37 P2 M3 C2 Falcon Airports 38 P2 M3 C3 Falcon Airports 39 P2 M3 C4 Falcon Airports 40 P2 M3 C5 Falcon Airports

Search Engine $8,000 $38,325

Search Engine $8,000 $38,325

41 P2 M4 C1 Falcon Data Centers Print Ad 42 P2 M4 C2 Falcon Data Centers Trade Show 43 P2 M4 C3 Falcon Data Centers Direct Mail 44 P2 M4 C4 Falcon Data Centers Email 46 P2 M5 C1 Falcon Hospitals 47 P2 M5 C2 Falcon Hospitals 48 P2 M5 C3 Falcon Hospitals 49 P2 M5 C4 Falcon Hospitals 50 P2 M5 C5 Falcon Hospitals 51 P3 M1 C1 Hawk 52 P3 M1 C2 Hawk 53 P3 M1 C3 Hawk 54 P3 M1 C4 Hawk 55 P3 M1 C5 Hawk 56 P3 M2 C1 Hawk 57 P3 M2 C2 Hawk 58 P3 M2 C3 Hawk 59 P3 M2 C4 Hawk 60 P3 M2 C5 Hawk 61 P3 M3 C1 Hawk 62 P3 M3 C2 Hawk 63 P3 M3 C3 Hawk 64 P3 M3 C4 Hawk 65 P3 M3 C5 Hawk 66 P3 M4 C1 Hawk 67 P3 M4 C2 Hawk 68 P3 M4 C3 Hawk 69 P3 M4 C4 Hawk 70 P3 M4 C5 Hawk 71 P3 M5 C1 Hawk 72 P3 M5 C2 Hawk Print Ad Trade Show Direct Mail Email

45 P2 M4 C5 Falcon Data Centers Search Engine $8,000 $38,325

Search Engine $8,000 $38,325

Construction Print Ad Construction Trade Show Construction Direct Mail Construction Email Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports Print Ad Trade Show Direct Mail Email Print Ad Trade Show Direct Mail Email

Construction Search Engine $8,000 $49,940

Search Engine $8,000 $49,940

Search Engine $8,000 $49,940

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Hospitals Hospitals Print Ad Trade Show

Data Centers Search Engine $8,000 $49,940

Marketing Optimization

72

J P M C Product 73 P3 M5 C3 Hawk 74 P3 M5 C4 Hawk 75 P3 M5 C5 Hawk

Market Hospitals Hospitals Hospitals

Channel Direct Mail Email

Product Expected List Revenue Index Channel Price Predicted Conversion Per Unit Efficiency Value Cost (K) (W) Leads (L) Rate (V) (W*D*E=R) (R/K=F) (F/=XR) $9,500 $49,940 $7,250 $49,940 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 58.6 61.6 42.4 45.5 70.9 50.9 44.7 43.1 39.2 58.4 48.2 41.8 45.4 41.2 67.2 51.8 43.2 61.8 57.2 77.4 51.8 58.6 58.0 49.4 79.0 59.6 62.6 43.4 37.9 63.3 43.3 37.1 35.5 31.2 50.4 40.2 33.8 37.4 0.19 0.19 0.19 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 $556,032 $584,498 $402,317 $476,740 $742,993 $533,345 $468,354 $451,582 $410,910 $612,172 $505,252 $438,164 $475,901 $431,456 $703,998 $542,569 $452,420 $647,393 $599,593 $811,338 $542,988 $614,269 $607,979 $517,831 $828,110 $624,751 $656,198 $454,936 $379,548 $634,050 $433,655 $371,532 $355,501 $312,616 $504,995 $402,794 $338,668 $374,739 58.53 80.62 50.29 43.34 39.10 56.14 64.60 56.45 37.36 32.22 53.18 60.44 59.49 39.22 37.05 57.11 62.40 80.92 54.51 42.70 57.16 84.73 76.00 47.08 43.58 65.76 90.51 56.87 34.50 33.37 45.65 51.25 44.44 28.42 26.58 42.40 46.71 46.84 1.12 1.54 0.96 0.83 0.75 1.07 1.24 1.08 0.72 0.62 1.02 1.16 1.14 0.75 0.71 1.09 1.19 1.55 1.04 0.82 1.09 1.62 1.45 0.90 0.83 1.26 1.73 1.09 0.66 0.64 0.87 0.98 0.85 0.54 0.51 0.81 0.89 0.90

Search Engine $8,000 $49,940

76 P4 M1 C1 Osprey Construction Print Ad 77 P4 M1 C2 Osprey Construction Trade Show 78 P4 M1 C3 Osprey Construction Direct Mail 79 P4 M1 C4 Osprey Construction Email 81 P4 M2 C1 Osprey Shipping 82 P4 M2 C2 Osprey Shipping 83 P4 M2 C3 Osprey Shipping 84 P4 M2 C4 Osprey Shipping 85 P4 M2 C5 Osprey Shipping 86 P4 M3 C1 Osprey Airports 87 P4 M3 C2 Osprey Airports 88 P4 M3 C3 Osprey Airports 89 P4 M3 C4 Osprey Airports 90 P4 M3 C5 Osprey Airports Print Ad Trade Show Direct Mail Email Print Ad Trade Show Direct Mail Email

80 P4 M1 C5 Osprey Construction Search Engine $8,000 $65,515

Search Engine $8,000 $65,515

Search Engine $8,000 $65,515

91 P4 M4 C1 Osprey Data Centers Print Ad 92 P4 M4 C2 Osprey Data Centers Trade Show 93 P4 M4 C3 Osprey Data Centers Direct Mail 94 P4 M4 C4 Osprey Data Centers Email 96 P4 M5 C1 Osprey Hospitals 97 P4 M5 C2 Osprey Hospitals 98 P4 M5 C3 Osprey Hospitals 99 P4 M5 C4 Osprey Hospitals 100 P4 M5 C5 Osprey Hospitals 101 P5 M1 C1 Eagle 102 P5 M1 C2 Eagle 103 P5 M1 C3 Eagle 104 P5 M1 C4 Eagle 105 P5 M1 C5 Eagle 106 P5 M2 C1 Eagle 107 P5 M2 C2 Eagle 108 P5 M2 C3 Eagle 109 P5 M2 C4 Eagle 110 P5 M2 C5 Eagle Print Ad Trade Show Direct Mail Email

95 P4 M4 C5 Osprey Data Centers Search Engine $8,000 $65,515

Search Engine $8,000 $65,515

Construction Print Ad Construction Trade Show Construction Direct Mail Construction Email Shipping Shipping Shipping Shipping Shipping Print Ad Trade Show Direct Mail Email

Construction Search Engine $8,000 $77,075

Search Engine $8,000 $77,075

Marketing Optimization

73

J P M C Product 111 P5 M3 C1 Eagle 112 P5 M3 C2 Eagle 113 P5 M3 C3 Eagle 114 P5 M3 C4 Eagle 115 P5 M3 C5 Eagle 116 P5 M4 C1 Eagle 117 P5 M4 C2 Eagle 118 P5 M4 C3 Eagle 119 P5 M4 C4 Eagle 120 P5 M4 C5 Eagle 121 P5 M5 C1 Eagle 122 P5 M5 C2 Eagle 123 P5 M5 C3 Eagle 124 P5 M5 C4 Eagle 125 P5 M5 C5 Eagle

Market Airports Airports Airports Airports Airports

Channel Print Ad Trade Show Direct Mail Email

Product Expected List Revenue Index Channel Price Predicted Conversion Per Unit Efficiency Value Cost (K) (W) Leads (L) Rate (V) (W*D*E=R) (R/K=F) (F/=XR) $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 30.8 56.8 41.4 32.8 51.4 48.0 68.2 42.6 49.4 48.8 52.4 82.0 62.6 65.6 46.4 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.13 $308,208 $568,721 $414,417 $328,247 $514,614 $480,948 $683,347 $426,841 $494,976 $488,964 $525,035 $821,620 $627,236 $657,296 $464,916 28.02 29.93 43.62 45.28 64.33 43.72 35.97 44.93 68.27 61.12 47.73 43.24 66.02 90.66 58.11 0.54 0.57 0.84 0.87 1.23 0.84 0.69 0.86 1.31 1.17 0.91 0.83 1.26 1.74 1.11

Search Engine $8,000 $77,075

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Hospitals Hospitals Hospitals Hospitals Hospitals Print Ad Trade Show Direct Mail Email

Data Centers Search Engine $8,000 $77,075

Search Engine $8,000 $77,075

Marketing Optimization Appendix H Indexed Values for Targeted Lead Generation

74

J P M C Product

Market

Channel

Product Expected List Revenue Per Index Channel Price Predicted Conversion Efficiency Value Unit Cost (K) (W) Leads (L) Rate (V) (W*D*E=G) (R/K=F) (F/=XG) $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $25,540 $19,000 $25,540 $9,500 $25,540 $7,250 $25,540 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 12.1 11.5 9.9 9.9 9.5 16.4 16.2 15.8 15.4 12.0 10.8 17.2 13.8 17.6 15.6 9.8 13.2 11.2 12.2 10.4 13.7 10.3 11.1 14.1 8.7 10.5 9.9 8.3 8.3 7.9 13.4 13.2 12.8 12.4 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.22 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 $68,100 $64,729 $55,738 $55,738 $53,491 $92,373 $91,249 $89,002 $86,754 $67,650 $60,683 $96,643 $77,539 $98,891 $87,653 $55,289 $74,393 $63,155 $68,774 $58,660 $77,090 $57,986 $62,481 $79,337 $48,996 $72,572 $68,433 $57,396 $57,396 $54,636 $92,716 $91,336 $88,577 $85,817 6.19 3.41 5.87 7.69 6.69 8.40 4.80 9.37 11.97 8.46 5.52 5.09 8.16 13.64 10.96 5.03 3.92 6.65 9.49 7.33 7.01 3.05 6.58 10.94 6.12 6.60 3.60 6.04 7.92 6.83 8.43 4.81 9.32 11.84 0.70 0.38 0.66 0.86 0.75 0.94 0.54 1.05 1.35 0.95 0.62 0.57 0.92 1.53 1.23 0.57 0.44 0.75 1.07 0.82 0.79 0.34 0.74 1.23 0.69 0.74 0.41 0.68 0.89 0.77 0.95 0.54 1.05 1.33

1 P1 M1 C1 Raven Construction Print Ad 2 P1 M1 C2 Raven Construction Trade Show 3 P1 M1 C3 Raven Construction Direct Mail 4 P1 M1 C4 Raven Construction Email 6 P1 M2 C1 Raven Shipping 7 P1 M2 C2 Raven Shipping 8 P1 M2 C3 Raven Shipping 9 P1 M2 C4 Raven Shipping 10 P1 M2 C5 Raven Shipping 11 P1 M3 C1 Raven Airports 12 P1 M3 C2 Raven Airports 13 P1 M3 C3 Raven Airports 14 P1 M3 C4 Raven Airports 15 P1 M3 C5 Raven Airports Print Ad Trade Show Direct Mail Email Print Ad Trade Show Direct Mail Email

5 P1 M1 C5 Raven Construction Search Engine $8,000 $25,540

Search Engine $8,000 $25,540

Search Engine $8,000 $25,540

16 P1 M4 C1 Raven Data Centers Print Ad 17 P1 M4 C2 Raven Data Centers Trade Show 18 P1 M4 C3 Raven Data Centers Direct Mail 19 P1 M4 C4 Raven Data Centers Email 21 P1 M5 C1 Raven Hospitals 22 P1 M5 C2 Raven Hospitals 23 P1 M5 C3 Raven Hospitals 24 P1 M5 C4 Raven Hospitals 25 P1 M5 C5 Raven Hospitals Print Ad Trade Show Direct Mail Email

20 P1 M4 C5 Raven Data Centers Search Engine $8,000 $25,540

Search Engine $8,000 $25,540

26 P2 M1 C1 Falcon Construction Print Ad 27 P2 M1 C2 Falcon Construction Trade Show 28 P2 M1 C3 Falcon Construction Direct Mail 29 P2 M1 C4 Falcon Construction Email 31 P2 M2 C1 Falcon Shipping 32 P2 M2 C2 Falcon Shipping 33 P2 M2 C3 Falcon Shipping 34 P2 M2 C4 Falcon Shipping Print Ad Trade Show Direct Mail Email

30 P2 M1 C5 Falcon Construction Search Engine $8,000 $38,325

Marketing Optimization

75

J P M C Product

Market

Channel Print Ad Trade Show Direct Mail Email

Product Expected List Revenue Per Index Channel Price Predicted Conversion Unit Efficiency Value Cost (K) (W) Leads (L) Rate (V) (W*D*E=G) (R/K=F) (F/=XG) 9.0 10.8 17.2 13.8 17.6 15.6 9.0 12.4 10.4 11.4 9.6 11.5 8.1 8.9 11.9 6.5 15.7 15.1 13.5 13.5 13.1 14.2 14.0 13.6 13.2 9.8 11.4 17.8 14.4 18.2 16.2 10.0 13.4 11.4 12.4 10.6 12.7 9.3 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 $62,362 $74,504 $118,654 $95,199 $121,414 $107,617 $62,362 $85,817 $72,020 $78,919 $66,502 $79,471 $56,016 $61,535 $82,230 $44,978 $117,759 $113,264 $101,278 $101,278 $98,282 $106,672 $105,174 $102,177 $99,181 $73,711 $85,397 $133,340 $107,870 $136,336 $121,354 $75,210 $100,679 $85,697 $93,188 $79,704 $95,286 $69,816 7.80 6.77 6.24 10.02 16.75 13.45 5.67 4.52 7.58 10.89 8.31 7.22 2.95 6.48 11.34 5.62 10.71 5.96 10.66 13.97 12.29 9.70 5.54 10.76 13.68 9.21 7.76 7.02 11.35 18.80 15.17 6.84 5.30 9.02 12.85 9.96 8.66 3.67 0.88 0.76 0.70 1.13 1.88 1.51 0.64 0.51 0.85 1.22 0.94 0.81 0.33 0.73 1.28 0.63 1.20 0.67 1.20 1.57 1.38 1.09 0.62 1.21 1.54 1.04 0.87 0.79 1.28 2.12 1.71 0.77 0.60 1.01 1.45 1.12 0.97 0.41 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 $11,000 $38,325 $19,000 $38,325 $9,500 $38,325 $7,250 $38,325 $11,000 $49,940 $19,000 $49,940 $9,500 $49,940 $7,250 $49,940 $11,000 $49,940 $19,000 $49,940 $9,500 $49,940 $7,250 $49,940 $11,000 $49,940 $19,000 $49,940 $9,500 $49,940 $7,250 $49,940 $11,000 $49,940 $19,000 $49,940 $9,500 $49,940 $7,250 $49,940 $11,000 $49,940 $19,000 $49,940

35 P2 M2 C5 Falcon Shipping 36 P2 M3 C1 Falcon Airports 37 P2 M3 C2 Falcon Airports 38 P2 M3 C3 Falcon Airports 39 P2 M3 C4 Falcon Airports 40 P2 M3 C5 Falcon Airports

Search Engine $8,000 $38,325

Search Engine $8,000 $38,325

41 P2 M4 C1 Falcon Data Centers Print Ad 42 P2 M4 C2 Falcon Data Centers Trade Show 43 P2 M4 C3 Falcon Data Centers Direct Mail 44 P2 M4 C4 Falcon Data Centers Email 46 P2 M5 C1 Falcon Hospitals 47 P2 M5 C2 Falcon Hospitals 48 P2 M5 C3 Falcon Hospitals 49 P2 M5 C4 Falcon Hospitals 50 P2 M5 C5 Falcon Hospitals 51 P3 M1 C1 Hawk 52 P3 M1 C2 Hawk 53 P3 M1 C3 Hawk 54 P3 M1 C4 Hawk 55 P3 M1 C5 Hawk 56 P3 M2 C1 Hawk 57 P3 M2 C2 Hawk 58 P3 M2 C3 Hawk 59 P3 M2 C4 Hawk 60 P3 M2 C5 Hawk 61 P3 M3 C1 Hawk 62 P3 M3 C2 Hawk 63 P3 M3 C3 Hawk 64 P3 M3 C4 Hawk 65 P3 M3 C5 Hawk 66 P3 M4 C1 Hawk 67 P3 M4 C2 Hawk 68 P3 M4 C3 Hawk 69 P3 M4 C4 Hawk 70 P3 M4 C5 Hawk 71 P3 M5 C1 Hawk 72 P3 M5 C2 Hawk Print Ad Trade Show Direct Mail Email

45 P2 M4 C5 Falcon Data Centers Search Engine $8,000 $38,325

Search Engine $8,000 $38,325

Construction Print Ad Construction Trade Show Construction Direct Mail Construction Email Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports Print Ad Trade Show Direct Mail Email Print Ad Trade Show Direct Mail Email

Construction Search Engine $8,000 $49,940

Search Engine $8,000 $49,940

Search Engine $8,000 $49,940

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Hospitals Hospitals Print Ad Trade Show

Data Centers Search Engine $8,000 $49,940

Marketing Optimization

76

J P M C Product 73 P3 M5 C3 Hawk 74 P3 M5 C4 Hawk 75 P3 M5 C5 Hawk

Market Hospitals Hospitals Hospitals

Channel Direct Mail Email

Product Expected List Revenue Per Index Channel Price Predicted Conversion Unit Efficiency Value Cost (K) (W) Leads (L) Rate (V) (W*D*E=G) (R/K=F) (F/=XG) $9,500 $49,940 $7,250 $49,940 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $65,515 $19,000 $65,515 $9,500 $65,515 $7,250 $65,515 $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 10.1 13.1 7.7 12.7 12.1 10.5 10.5 10.1 14.6 14.4 14.0 13.6 10.2 7.4 13.8 10.4 14.2 12.2 11.2 14.6 12.6 13.6 11.8 11.5 8.1 8.9 11.9 6.5 13.9 13.3 11.7 11.7 11.3 15.2 15.0 14.6 14.2 10.8 0.15 0.15 0.15 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 $75,809 $98,282 $57,831 $100,002 $95,285 $82,706 $82,706 $79,561 $115,097 $113,524 $110,380 $107,235 $80,505 $58,177 $108,493 $81,763 $111,638 $95,914 $88,367 $115,097 $99,373 $107,235 $93,084 $90,568 $63,838 $70,127 $93,713 $51,259 $107,288 $102,664 $90,332 $90,332 $87,249 $117,462 $115,921 $112,838 $109,755 $83,549 7.98 13.56 7.23 9.09 5.02 8.71 11.41 9.95 10.46 5.97 11.62 14.79 10.06 5.29 5.71 8.61 15.40 11.99 8.03 6.06 10.46 14.79 11.64 8.23 3.36 7.38 12.93 6.41 9.75 5.40 9.51 12.46 10.91 10.68 6.10 11.88 15.14 10.44 0.90 1.52 0.81 1.02 0.56 0.98 1.28 1.12 1.18 0.67 1.31 1.66 1.13 0.59 0.64 0.97 1.73 1.35 0.90 0.68 1.18 1.66 1.31 0.93 0.38 0.83 1.45 0.72 1.10 0.61 1.07 1.40 1.23 1.20 0.69 1.34 1.70 1.17

Search Engine $8,000 $49,940

76 P4 M1 C1 Osprey Construction Print Ad 77 P4 M1 C2 Osprey Construction Trade Show 78 P4 M1 C3 Osprey Construction Direct Mail 79 P4 M1 C4 Osprey Construction Email 81 P4 M2 C1 Osprey Shipping 82 P4 M2 C2 Osprey Shipping 83 P4 M2 C3 Osprey Shipping 84 P4 M2 C4 Osprey Shipping 85 P4 M2 C5 Osprey Shipping 86 P4 M3 C1 Osprey Airports 87 P4 M3 C2 Osprey Airports 88 P4 M3 C3 Osprey Airports 89 P4 M3 C4 Osprey Airports 90 P4 M3 C5 Osprey Airports Print Ad Trade Show Direct Mail Email Print Ad Trade Show Direct Mail Email

80 P4 M1 C5 Osprey Construction Search Engine $8,000 $65,515

Search Engine $8,000 $65,515

Search Engine $8,000 $65,515

91 P4 M4 C1 Osprey Data Centers Print Ad 92 P4 M4 C2 Osprey Data Centers Trade Show 93 P4 M4 C3 Osprey Data Centers Direct Mail 94 P4 M4 C4 Osprey Data Centers Email 96 P4 M5 C1 Osprey Hospitals 97 P4 M5 C2 Osprey Hospitals 98 P4 M5 C3 Osprey Hospitals 99 P4 M5 C4 Osprey Hospitals 100 P4 M5 C5 Osprey Hospitals 101 P5 M1 C1 Eagle 102 P5 M1 C2 Eagle 103 P5 M1 C3 Eagle 104 P5 M1 C4 Eagle 105 P5 M1 C5 Eagle 106 P5 M2 C1 Eagle 107 P5 M2 C2 Eagle 108 P5 M2 C3 Eagle 109 P5 M2 C4 Eagle 110 P5 M2 C5 Eagle Print Ad Trade Show Direct Mail Email

95 P4 M4 C5 Osprey Data Centers Search Engine $8,000 $65,515

Search Engine $8,000 $65,515

Construction Print Ad Construction Trade Show Construction Direct Mail Construction Email Shipping Shipping Shipping Shipping Shipping Print Ad Trade Show Direct Mail Email

Construction Search Engine $8,000 $77,075

Search Engine $8,000 $77,075

Marketing Optimization

77

J P M C Product 111 P5 M3 C1 Eagle 112 P5 M3 C2 Eagle 113 P5 M3 C3 Eagle 114 P5 M3 C4 Eagle 115 P5 M3 C5 Eagle 116 P5 M4 C1 Eagle 117 P5 M4 C2 Eagle 118 P5 M4 C3 Eagle 119 P5 M4 C4 Eagle 120 P5 M4 C5 Eagle 121 P5 M5 C1 Eagle 122 P5 M5 C2 Eagle 123 P5 M5 C3 Eagle 124 P5 M5 C4 Eagle 125 P5 M5 C5 Eagle

Market Airports Airports Airports Airports Airports

Channel Print Ad Trade Show Direct Mail Email

Product Expected List Revenue Per Index Channel Price Predicted Conversion Unit Efficiency Value Cost (K) (W) Leads (L) Rate (V) (W*D*E=G) (R/K=F) (F/=XG) $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 $11,000 $77,075 $19,000 $77,075 $9,500 $77,075 $7,250 $77,075 9.6 16.0 12.6 16.4 14.4 9.8 13.2 11.2 12.2 10.4 14.5 11.1 11.9 14.9 9.5 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 $73,992 $123,320 $97,115 $126,403 $110,988 $75,842 $102,047 $86,632 $94,340 $80,466 $111,913 $85,707 $91,873 $114,996 $73,375 6.73 6.49 10.22 17.43 13.87 6.89 5.37 9.12 13.01 10.06 10.17 4.51 9.67 15.86 9.17 0.76 0.73 1.15 1.96 1.56 0.78 0.60 1.03 1.46 1.13 1.14 0.51 1.09 1.78 1.03

Search Engine $8,000 $77,075

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Hospitals Hospitals Hospitals Hospitals Hospitals Print Ad Trade Show Direct Mail Email

Data Centers Search Engine $8,000 $77,075

Search Engine $8,000 $77,075

Marketing Optimization Appendix I Blended Final Utility Values

78

J 1 2 3 4 5 6 7 8 9

P M C

Product

Market Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports

Channel Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Total Revenue Targeted Index Value Revenue Index Blended Utility Value ((XR+XG)/2=U) (XR) Value (XG) 1.00 0.75 1.23 1.50 1.33 0.71 0.54 0.95 1.13 1.08 0.49 0.46 0.71 0.78 1.00 0.80 0.60 0.85 1.23 1.11 0.61 0.55 0.84 1.16 0.74 0.98 0.79 1.23 1.47 1.30 0.91 0.70 1.22 1.45 1.39 0.70 0.38 0.66 0.86 0.75 0.94 0.54 1.05 1.35 0.95 0.62 0.57 0.92 1.53 1.23 0.57 0.44 0.75 1.07 0.82 0.79 0.34 0.74 1.23 0.69 0.74 0.41 0.68 0.89 0.77 0.95 0.54 1.05 1.33 0.88 0.85 0.57 0.94 1.18 1.04 0.83 0.54 1.00 1.24 1.02 0.56 0.52 0.81 1.16 1.12 0.68 0.52 0.80 1.15 0.97 0.70 0.45 0.79 1.19 0.72 0.86 0.60 0.96 1.18 1.03 0.93 0.62 1.13 1.39 1.13

P1 M1 C1 Raven P1 M1 C2 Raven P1 M1 C3 Raven P1 M1 C4 Raven P1 M1 C5 Raven P1 M2 C1 Raven P1 M2 C2 Raven P1 M2 C3 Raven P1 M2 C4 Raven

10 P1 M2 C5 Raven 11 P1 M3 C1 Raven 12 P1 M3 C2 Raven 13 P1 M3 C3 Raven 14 P1 M3 C4 Raven 15 P1 M3 C5 Raven 16 P1 M4 C1 Raven 17 P1 M4 C2 Raven 18 P1 M4 C3 Raven 19 P1 M4 C4 Raven 20 P1 M4 C5 Raven 21 P1 M5 C1 Raven 22 P1 M5 C2 Raven 23 P1 M5 C3 Raven 24 P1 M5 C4 Raven 25 P1 M5 C5 Raven 26 P2 M1 C1 Falcon 27 P2 M1 C2 Falcon 28 P2 M1 C3 Falcon 29 P2 M1 C4 Falcon 30 P2 M1 C5 Falcon 31 P2 M2 C1 Falcon 32 P2 M2 C2 Falcon 33 P2 M2 C3 Falcon 34 P2 M2 C4 Falcon 35 P2 M2 C5 Falcon

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Data Centers Search Engine Hospitals Hospitals Hospitals Hospitals Hospitals Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Marketing Optimization

79

P M C

Product

Market Airports Airports Airports Airports Airports

Channel Print Ad Trade Show Direct Mail Email Search Engine

Total Revenue Targeted Index Value Revenue Index Blended Utility (XR) Value (XG) Value ((XR+XG)/2=U) 0.68 0.63 0.98 1.08 1.37 0.92 0.71 0.97 1.43 1.29 0.83 0.74 1.14 1.56 1.01 0.83 0.72 1.06 1.24 1.09 0.86 0.68 1.17 1.37 1.33 0.66 0.63 0.97 1.05 1.38 1.15 0.86 1.23 1.78 1.60 0.80 0.75 1.12 0.76 0.70 1.13 1.88 1.51 0.64 0.51 0.85 1.22 0.94 0.81 0.33 0.73 1.28 0.63 1.20 0.67 1.20 1.57 1.38 1.09 0.62 1.21 1.54 1.04 0.87 0.79 1.28 2.12 1.71 0.77 0.60 1.01 1.45 1.12 0.97 0.41 0.90 0.72 0.66 1.05 1.48 1.44 0.78 0.61 0.91 1.33 1.11 0.82 0.54 0.93 1.42 0.82 1.02 0.70 1.13 1.41 1.23 0.98 0.65 1.19 1.46 1.18 0.77 0.71 1.12 1.58 1.54 0.96 0.73 1.12 1.61 1.36 0.89 0.58 1.01

36 P2 M3 C1 Falcon 37 P2 M3 C2 Falcon 38 P2 M3 C3 Falcon 39 P2 M3 C4 Falcon 40 P2 M3 C5 Falcon 41 P2 M4 C1 Falcon 42 P2 M4 C2 Falcon 43 P2 M4 C3 Falcon 44 P2 M4 C4 Falcon 45 P2 M4 C5 Falcon 46 P2 M5 C1 Falcon 47 P2 M5 C2 Falcon 48 P2 M5 C3 Falcon 49 P2 M5 C4 Falcon 50 P2 M5 C5 Falcon 51 P3 M1 C1 Hawk 52 P3 M1 C2 Hawk 53 P3 M1 C3 Hawk 54 P3 M1 C4 Hawk 55 P3 M1 C5 Hawk 56 P3 M2 C1 Hawk 57 P3 M2 C2 Hawk 58 P3 M2 C3 Hawk 59 P3 M2 C4 Hawk 60 P3 M2 C5 Hawk 61 P3 M3 C1 Hawk 62 P3 M3 C2 Hawk 63 P3 M3 C3 Hawk 64 P3 M3 C4 Hawk 65 P3 M3 C5 Hawk 66 P3 M4 C1 Hawk 67 P3 M4 C2 Hawk 68 P3 M4 C3 Hawk 69 P3 M4 C4 Hawk 70 P3 M4 C5 Hawk 71 P3 M5 C1 Hawk 72 P3 M5 C2 Hawk 73 P3 M5 C3 Hawk

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Data Centers Search Engine Hospitals Hospitals Hospitals Hospitals Hospitals Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Data Centers Search Engine Hospitals Hospitals Hospitals Print Ad Trade Show Direct Mail

Marketing Optimization

80

P M C

Product

Market Hospitals Hospitals Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Airports Airports Airports Airports Airports

Channel Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Total Revenue Targeted Index Value Revenue Index Blended Utility (XR) Value (XG) Value ((XR+XG)/2=U) 1.54 0.96 0.83 0.75 1.07 1.24 1.08 0.72 0.62 1.02 1.16 1.14 0.75 0.71 1.09 1.19 1.55 1.04 0.82 1.09 1.62 1.45 0.90 0.83 1.26 1.73 1.09 0.66 0.64 0.87 0.98 0.85 0.54 0.51 0.81 0.89 0.90 0.54 1.52 0.81 1.02 0.56 0.98 1.28 1.12 1.18 0.67 1.31 1.66 1.13 0.59 0.64 0.97 1.73 1.35 0.90 0.68 1.18 1.66 1.31 0.93 0.38 0.83 1.45 0.72 1.10 0.61 1.07 1.40 1.23 1.20 0.69 1.34 1.70 1.17 0.76 1.53 0.89 0.93 0.66 1.03 1.26 1.10 0.95 0.64 1.16 1.41 1.14 0.67 0.68 1.03 1.46 1.45 0.97 0.75 1.14 1.64 1.38 0.91 0.61 1.04 1.59 0.90 0.88 0.62 0.97 1.19 1.04 0.87 0.60 1.07 1.30 1.04 0.65

74 P3 M5 C4 Hawk 75 P3 M5 C5 Hawk 76 P4 M1 C1 Osprey 77 P4 M1 C2 Osprey 78 P4 M1 C3 Osprey 79 P4 M1 C4 Osprey 80 P4 M1 C5 Osprey 81 P4 M2 C1 Osprey 82 P4 M2 C2 Osprey 83 P4 M2 C3 Osprey 84 P4 M2 C4 Osprey 85 P4 M2 C5 Osprey 86 P4 M3 C1 Osprey 87 P4 M3 C2 Osprey 88 P4 M3 C3 Osprey 89 P4 M3 C4 Osprey 90 P4 M3 C5 Osprey 91 P4 M4 C1 Osprey 92 P4 M4 C2 Osprey 93 P4 M4 C3 Osprey 94 P4 M4 C4 Osprey 95 P4 M4 C5 Osprey 96 P4 M5 C1 Osprey 97 P4 M5 C2 Osprey 98 P4 M5 C3 Osprey 99 P4 M5 C4 Osprey 100 P4 M5 C5 Osprey 101 P5 M1 C1 Eagle 102 P5 M1 C2 Eagle 103 P5 M1 C3 Eagle 104 P5 M1 C4 Eagle 105 P5 M1 C5 Eagle 106 P5 M2 C1 Eagle 107 P5 M2 C2 Eagle 108 P5 M2 C3 Eagle 109 P5 M2 C4 Eagle 110 P5 M2 C5 Eagle 111 P5 M3 C1 Eagle

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Data Centers Search Engine Hospitals Hospitals Hospitals Hospitals Hospitals Construction Construction Construction Construction Construction Shipping Shipping Shipping Shipping Shipping Airports Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad

Marketing Optimization

81

P M C

Product

Market Airports Airports Airports Airports

Channel Trade Show Direct Mail Email Search Engine

Total Revenue Targeted Index Value Revenue Index Blended Utility (XR) Value (XG) Value ((XR+XG)/2=U) 0.57 0.84 0.87 1.23 0.84 0.69 0.86 1.31 1.17 0.91 0.83 1.26 1.74 1.11 0.73 1.15 1.96 1.56 0.78 0.60 1.03 1.46 1.13 1.14 0.51 1.09 1.78 1.03 0.65 0.99 1.41 1.40 0.81 0.65 0.94 1.39 1.15 1.03 0.67 1.18 1.76 1.07

112 P5 M3 C2 Eagle 113 P5 M3 C3 Eagle 114 P5 M3 C4 Eagle 115 P5 M3 C5 Eagle 116 P5 M4 C1 Eagle 117 P5 M4 C2 Eagle 118 P5 M4 C3 Eagle 119 P5 M4 C4 Eagle 120 P5 M4 C5 Eagle 121 P5 M5 C1 Eagle 122 P5 M5 C2 Eagle 123 P5 M5 C3 Eagle 124 P5 M5 C4 Eagle 125 P5 M5 C5 Eagle

Data Centers Print Ad Data Centers Trade Show Data Centers Direct Mail Data Centers Email Data Centers Search Engine Hospitals Hospitals Hospitals Hospitals Hospitals Print Ad Trade Show Direct Mail Email Search Engine

Marketing Optimization Appendix J Saturation Point Analysis


Total Market Saturation Analysis
600 Average Total Leads Generated Average Total Leads Generated 500 400 300 200 100 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

82

Channel Saturation Analysis - P1


600 Average Total Leads Generated
1 2 3 4 5 6 7 8 9 10 11 12

Channel Saturation Analysis - P2


140 120 100 80 60 40 20 0
1 2 3 4 5 6

500 400 300 200 100 0

Total Promotions within a Single Market

Print Ads within a Single Market

Trade Shows within a Single Market

Total Promotions within Average Total a Single Market Leads Generated 1 59.0 2 135.3 3 174.8 4 223.2 5 292.7 6 385.0 7 368.4 8 468.0 9 531.8 10 515.9 11 500.4 12 450.3 13 358.3 14 360.6 15 345.9

Print Ads within a Single Average Total Market Leads Generated 1 65.7 2 135.9 3 164.7 4 231.1 5 303.9 6 391.1 7 402.8 8 490.9 9 473.9 10 460.9 11 480.0 12 374.4

Trade Shows within a Single Market 1 2 3 4 5 6

Average Total Leads Generated 60.9 118.8 122.8 97.5 75.5 56.5

Channel Saturation Analysis - P3


250 Average Total Leads Generated Average Total Leads Generated 200 150 100 50 0
1 2 3 4 5 6 7

Channel Saturation Analysis - P4


250 200 150 100 50 0
1 2 3 4 5 6 7

Channel Saturation Analysis - P5


200 180 160 140 120 100 80 60 40 20 0
1 2 3 4 5

Direct Mailings within a Single Market

Emails Sent within a Single Market

Average Total Leads Generated

Search Eng. Campaigns within a Single Market

Direct Mailings within a Average Total Single Market Leads Generated 1 67.7 2 135.4 3 164.3 4 228.1 5 234.9 6 220.8 7 185.9

Emails Sent within a Single Market 1 2 3 4 5 6 7

Average Total Leads Generated 47.8 129.8 177.0 235.5 226.1 195.1 178.6

Seach Engine Campaigns within a Single Market 1 2 3 4 5

Average Total Leads Generated 59.3 133.9 173.3 146.3 116.3

Marketing Optimization Appendix K Linear Programming Constraint Values


Category General Item Total Budget Nonnegativity Integers Raven Falcon Product Mix Requirements Hawk Osprey Eagle Construction Shipping Market Mix Requirements Airports Data Centers Hospitals Print Ad Trade Show Channel Mix Requirements Direct Mail Email Search Engine Construction Shipping Market Saturation Airports Data Centers Hospitals Print Ad Trade Show Channel Saturation - Construction Direct Mail Email Search Engine Print Ad Trade Show Channel Saturation - Shipping Direct Mail Email Search Engine Channel Saturation - Airports Print Ad Trade Show Relationship Limit $ 250,000

83

0 Integers 4 4 4 4 4 2 2 2 2 2 4 3 4 4 4 9 9 9 9 9 8 2 4 4 3 8 2 4 4 3 8 2

Marketing Optimization
Category Email Search Engine Print Ad Trade Show Channel Saturation - Data Centers Direct Mail Email Search Engine Print Ad Trade Show Channel Saturation - Hospitals Direct Mail Email Search Engine Item Direct Mail Relationship Limit

84

4 4 3 8 2 4 4 3 8 2 4 4 3

Marketing Optimization Appendix L Marketing Plan Summary Tables for All Four Models
TOTAL QUANTITY COST (Q) (K*Q) 4 4 1 2 3 2 4 1 $44,000 $38,000 $19,000 $14,500 $24,000 $38,000 $29,000 $8,000 EXPECTED TOTAL REVENUE (R*Q) $2,291,735 $2,418,767 $625,868 $796,283 $1,723,889 $1,622,676 $2,457,075 $607,979

85

P M C Product

Market Shipping Airports Airports Airports

Channel Direct Mail Trade Show Email Search Engine

EXPECTED TARGETED REVENUE (G*Q) $272,399 $354,307 $133,340 $272,672 $364,063 $230,194 $428,940 $93,084 $459,984 $2,608,982 68099.856 193286.72 155078.88 242827.2 144040.68 272672.4 115661.04

1 P1 M1 C1 Raven BLENDED MODEL 33 P2 M2 C3 Falcon 62 P3 M3 C2 Hawk 64 P3 M3 C4 Hawk 65 P3 M3 C5 Hawk 92 P4 M4 C2 Osprey 94 P4 M4 C4 Osprey 95 P4 M4 C5 Osprey 124 P5 M5 C4 Eagle

Construction Print Ad

Data Centers Trade Show Data Centers Email Data Centers Search Engine Hospitals Email

4 $29,000 $2,629,182 TOTALS $243,500 $15,173,454 1 2 2 2 2 2 2 2 2 2 1 1 1 $11,000 572933.712 $38,000 910531.648 $19,000 706007.328 $14,500 $19,000 $16,000 $22,000 $22,000 $19,000 $11,000 $9,500 820829.52 966096.6 804633.28

1 P1 M1 C1 Raven 12 P1 M3 C2 Raven PREVIOUS (HEURISTIC) MODEL 13 P1 M3 C3 Raven 39 P2 M3 C4 Falcon 43 P2 M4 C3 Falcon 64 P3 M3 C4 Hawk 75 P3 M5 C5 Hawk 80 P4 M1 C5 Osprey 81 P4 M2 C1 Osprey 106 P5 M2 C1 Eagle 107 P5 M2 C2 Eagle 111 P5 M3 C1 Eagle 118 P5 M4 C3 Eagle 121 P5 M5 C1 Eagle

Construction Print Ad Airports Airports Airports Airports Hospitals Shipping Shipping Shipping Airports Hospitals Trade Show Direct Mail Email Email Search Engine Print Ad Print Ad Trade Show Print Ad Print Ad

Data Centers Direct Mail

$14,500 796283.312

Construction Search Engine

$16,000 903163.584 159122.832 821820.16 230193.504 625232.4 504995.4 308207.51 426841.35 525034.9 $9,692,611 234924.6 115920.8 73992 86632.3 111912.9 $2,204,366

Data Centers Direct Mail

1 $11,000 TOTALS $242,500

Marketing Optimization

86

P M C Product

Market

Channel

TOTAL QUANTITY COST (Q) (K*Q) 1 3 1 1 2 4 3 1 3 2 $9,500 $21,750 $19,000 $9,500 $19,000 $44,000 $21,750 $8,000 $24,000 $38,000

EXPECTED TOTAL REVENUE (R*Q) $608,792 $1,702,864 $788,744 $612,449 $1,209,384 $2,649,217 $2,026,765 $669,895 $1,942,179 $1,656,219

EXPECTED TARGETED REVENUE (G*Q) $55,738 $167,215 $68,433 $57,396 $177,153 $300,839 $279,564 $79,704 $287,742 $127,676

3 P1 M1 C3 Raven TOTAL REVENUE MODEL 4 P1 M1 C4 Raven 27 P2 M1 C2 Falcon 28 P2 M1 C3 Falcon 33 P2 M2 C3 Falcon 66 P3 M4 C1 Hawk 69 P3 M4 C4 Hawk 70 P3 M4 C5 Hawk 90 P4 M3 C5 Osprey 97 P4 M5 C2 Osprey 124 P5 M5 C4 Eagle

Construction Direct Mail Construction Email Construction Trade Show Construction Direct Mail Shipping Direct Mail Data Centers Print Ad Data Centers Email Data Centers Search Engine Airports Hospitals Hospitals Search Engine Trade Show Email

4 $29,000 $2,629,182 $459,984 TOTALS $243,500 $16,495,690 $2,061,444 2 2 1 3 4 1 1 3 2 1 1 $19,000 $14,500 $19,000 $24,000 $44,000 $8,000 $19,000 $21,750 $19,000 $19,000 $7,250 $940,281 $855,284 $621,969 $1,723,107 $1,908,347 $454,314 $625,868 $1,194,425 $1,010,503 $811,338 $614,269 $178,004 $173,509 $118,654 $322,850 $471,034 $98,282 $133,340 $409,009 $220,759 $115,097 $107,235

8 P1 M2 C3 Raven TARGETED REVENUE MODEL 9 P1 M2 C4 Raven 37 P2 M3 C2 Falcon 40 P2 M3 C5 Falcon 51 P3 M1 C1 Hawk 55 P3 M1 C5 Hawk 62 P3 M3 C2 Hawk 64 P3 M3 C4 Hawk 83 P4 M2 C3 Osprey 92 P4 M4 C2 Osprey 94 P4 M4 C4 Osprey 124 P5 M5 C4 Eagle

Shipping Shipping Airports Airports

Direct Mail Email Trade Show Search Engine

Construction Print Ad Construction Search Engine Airports Airports Shipping Trade Show Email Direct Mail

Data Centers Trade Show Data Centers Email Hospitals Email

4 $29,000 $2,629,182 $459,984 TOTALS $243,500 $13,388,887 $2,807,755

Marketing Optimization Appendix M Reference Table of Constraint Satisfaction Values for All Four Models
Heuristic Model Actual Value Targeted Model Actual Value Revenue Model Actual Value Blended Model Actual Value

87

Category General

Item Total Budget Nonnegativity Integers Raven Falcon Hawk Osprey Eagle Construction Shipping Airports Data Centers Hospitals Print Ad Trade Show Direct Mail Email Search Engine Construction Shipping

Relationship 5 4 4 4 6 3 5 9 3 3 7 3 5 4 4 3 5 9 3 3 1 0 0 0 2 4 1 0 0

Limit

$243,500 (All) (All) 4 4 6 7 4 4 4 6 7 4 4 3 4 10 4 4 4 6 7 4 4 0 0 0 0 0 0 4 0

$243,500 (All) (All) 4 4 8 5 4 6 2 3 8 6 4 3 4 10 4 6 2 3 8 6 0 1 2 3 0 0 0 2 0

$243,500 (All) (All) 4 4 9 4 4 5 6 8 2 4 4 3 4 10 4 5 6 8 2 4 4 0 0 0 1 0 0 4 2

$242,500 (All) (All)

Product Mix Requirements

Market Mix Requirements

Channel Mix Requirements

Market Saturation

Airports Data Centers Hospitals Print Ad Trade Show Direct Mail Email Search Engine

Channel Saturation Construction

Channel Saturation Shipping

Print Ad Trade Show Direct Mail Email

$250,000 0 = Integers 4 4 4 4 4 2 2 2 2 2 4 3 4 4 4 9 9 9 9 9 8 2 4 4 3 8 2 4 4

Marketing Optimization

88

Heuristic Model Actual Value

Targeted Model Actual Value

Revenue Model Actual Value

Blended Model Actual Value

Category

Item Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine Print Ad Trade Show Direct Mail Email Search Engine

Relationship

Limit 3 8 2 4 4 3 8 2 4 4 3 8 2 4 4 3

0 0 1 0 2 3 0 2 0 4 1 0 0 0 4 0

0 0 0 0 0 3 4 0 0 3 1 0 2 0 4 0

0 0 2 0 3 3 0 1 0 1 0 0 0 0 4 0

0 1 2 2 4 0 0 0 3 0 0 1 0 0 0 2

Channel Saturation Airports

Channel Saturation Data Centers

Channel Saturation Hospitals

Marketing Optimization Appendix N Table of Costs, Revenues and Efficiencies for All Four Models
Expected Total Revenue $ 2,291,734.85 $ 2,418,767.40 $ 3,146,040.22 $ 4,687,729.28 $ 2,629,182.40 $ 2,291,734.85 $ 2,418,767.40 $ 3,146,040.22 $ 4,687,729.28 $ 2,629,182.40 $ 2,291,734.85 $ 2,248,543.58 $ 2,418,767.40 $ 5,882,540.27 $ 2,331,868.05 $15,173,454.14 $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ 2,189,472.69 1,786,926.12 1,600,916.59 1,724,983.74 2,390,311.56 1,476,097.30 1,952,047.96 3,541,859.32 1,392,937.95 1,329,668.18 2,853,228.68 1,415,527.05 2,098,945.28 1,617,112.83 1,707,796.86 9,692,610.70

89

Raven Falcon Hawk Osprey Eagle Construction Shipping Airports Data Centers Hospitals Print Ad Trade Show Direct Mail Email Search Engine GRAND TOTALS Raven Falcon Hawk Osprey Eagle Construction Shipping Airports Data Centers Hospitals Print Ad Trade Show Direct Mail Email Search Engine GRAND TOTALS By Channel By Market By Product By Channel By Market By Product

Total Cost $ 44,000 $ 38,000 $ 57,500 $ 75,000 $ 29,000 $ 44,000 $ 38,000 $ 57,500 $ 75,000 $ 29,000 $ 44,000 $ 57,000 $ 38,000 $ 72,500 $ 32,000 $243,500 $ 68,000 $ 33,500 $ 30,500 $ 38,000 $ 72,500 $ 27,000 $ 63,000 $ 97,000 $ 28,500 $ 27,000 $ 77,000 $ 57,000 $ 47,500 $ 29,000 $ 32,000 $242,500

Expected Targeted Total Efficiency Targeted Efficiency Revenue (Revenue/Cost) (Revenue/Cost) $ 272,399.42 52.08 6.19 $ 354,306.96 63.65 9.32 $ 770,074.80 54.71 13.39 $ 752,217.02 62.50 10.03 $ 459,983.60 90.66 15.86 $ 272,399.42 52.08 6.19 $ 354,306.96 63.65 9.32 $ 770,074.80 54.71 13.39 $ 752,217.02 62.50 10.03 $ 459,983.60 90.66 15.86 $ 272,399.42 52.08 6.19 $ 363,533.30 39.45 6.38 $ 354,306.96 63.65 9.32 $ 1,161,595.81 81.14 16.02 $ 457,146.31 72.87 14.29 $ 2,608,981.81 62.31 10.71 $ 416,465.46 $ 386,867.88 $ 388,333.44 $ 389,316.34 $ 623,382.60 $ 227,222.69 $ 581,038.90 $ 937,857.20 $ 230,672.98 $ 227,573.94 $ 719,122.86 $ 309,207.52 $ 385,751.86 $ 515,499.60 $ 274,783.87 $ 2,204,365.71 32.20 53.34 52.49 45.39 32.97 54.67 30.98 36.51 48.88 49.25 37.05 24.83 44.19 55.76 53.37 39.97 6.12 11.55 12.73 10.25 8.60 8.42 9.22 9.67 8.09 8.43 9.34 5.42 8.12 17.78 8.59 9.09

HEURISTIC MODEL

BLENDED MODEL

Marketing Optimization

90

Raven Falcon Hawk Osprey Eagle Construction Shipping Airports Data Centers Hospitals Print Ad Trade Show Direct Mail Email Search Engine GRAND TOTALS Raven Falcon Hawk Osprey Eagle Construction Shipping Airports Data Centers Hospitals Print Ad Trade Show Direct Mail Email Search Engine GRAND TOTALS By Channel By Market By Product By Channel By Market By Product

Total Cost $ 31,250 $ 47,500 $ 73,750 $ 62,000 $ 29,000 $ 59,750 $ 19,000 $ 24,000 $ 73,750 $ 67,000 $ 44,000 $ 57,000 $ 38,000 $ 72,500 $ 32,000 $243,500 $ 33,500 $ 43,000 $ 92,750 $ 45,250 $ 29,000 $ 52,000 $ 52,500 $ 83,750 $ 26,250 $ 29,000 $ 44,000 $ 57,000 $ 38,000 $ 72,500 $ 32,000 $243,500

Expected Total Revenue $ 2,311,656.05 $ 2,610,576.36 $ 5,345,877.24 $ 3,598,398.27 $ 2,629,182.40 $ 3,712,848.71 $ 1,209,383.70 $ 1,942,179.07 $ 5,345,877.24 $ 4,285,401.60 $ 2,649,217.12 $ 2,444,963.03 $ 2,430,624.40 $ 6,358,811.54 $ 2,612,074.23 $16,495,690.32 $ 1,795,564.16 $ 2,345,076.09 $ 4,182,954.42 $ 2,436,109.76 $ 2,629,182.40 $ 2,362,661.40 $ 2,806,067.52 $ 4,165,369.11 $ 1,425,606.40 $ 2,629,182.40 $ 1,908,347.23 $ 2,059,174.58 $ 1,950,784.00 $ 5,293,159.53 $ 2,177,421.50 $13,388,886.83

Expected Targeted Total Efficiency Targeted Efficiency Revenue (Revenue/Cost) (Revenue/Cost) $ 222,953.98 73.97 7.13 $ 302,982.12 54.96 6.38 $ 660,106.92 72.49 8.95 $ 415,417.51 58.04 6.70 $ 459,983.60 90.66 15.86 $ 348,782.62 62.14 5.84 $ 177,153.48 63.65 9.32 $ 287,741.88 80.92 11.99 $ 660,106.92 72.49 8.95 $ 587,659.23 63.96 8.77 $ 300,838.56 60.21 6.84 $ 196,108.75 42.89 3.44 $ 290,287.50 63.96 7.64 $ 906,763.21 87.71 12.51 $ 367,446.12 81.63 11.48 $ 2,061,444.14 67.74 8.47 $ 351,512.13 $ 441,504.00 $ 1,111,664.40 $ 443,091.05 $ 459,983.60 $ 569,316.00 $ 572,271.47 $ 983,852.40 $ 222,331.70 $ 459,983.60 $ 471,034.08 $ 367,090.75 $ 398,762.93 $ 1,149,735.70 $ 421,131.72 $ 2,807,755.18 53.60 54.54 45.10 53.84 90.66 45.44 53.45 49.74 54.31 90.66 43.37 36.13 51.34 73.01 68.04 54.99 10.49 10.27 11.99 9.79 15.86 10.95 10.90 11.75 8.47 15.86 10.71 6.44 10.49 15.86 13.16 11.53

TARGETED REVENUE MODEL

TOTAL REVENUE MODEL