Price and Quality

14 downloads 0 Views 207KB Size Report
Email: stephen.jones@atu.edu. Tami L. Knotts. Department of Management. Missouri State University. 901 S. National Ave. Springfield, MO 65897. U. S. A..
Characterizing Viability of Small Manufacturing Enterprises in the Market

Kee S. Kim Department of Finance Missouri State University 901 S. National Ave Springfield, MO 65897 U. S. A. Fax: 417-836-6224 Email: [email protected] Corresponding Author

Stephen C. Jones Department of Business & Economics Arkansas Tech University Russellville, AR 72801 U. S. A. Fax: 479-968-0677 Email: [email protected]

Tami L. Knotts Department of Management Missouri State University 901 S. National Ave. Springfield, MO 65897 U. S. A. Fax: 417-836-3004 Email: [email protected]

Characterizing Viability of Small Manufacturing Enterprises (SME) in the Market

Introduction Small business failure rates may be lower than previously thought. Headd (2003) found that over half of small firms with employees remained in business after four years and that surviving firms were likely to have paid employees, adequate start-up capital ($50,000 or more), and experienced owners. He stressed, however, that many firms lacking these and other “success” qualities did not necessarily fail. Often their owners still described the ventures as a success even after closure because it allowed the owners to be independent, to earn extra income at a crucial time, or to successfully sell the business as part of an exit strategy. The definition of failure may be the problem, as failure in this case doesn’t necessarily mean that the firm went under. Headd’s work also calls into question the meaning of success. If a firm survives, is the firm successful? We suggest that it is. While Headd (2003) examined firm survival for four years, our study looks at the survival status of small firms over an eight to ten year period. This paper examines the survival characteristics of small manufacturing enterprises (SMEs) competing to be mass merchandiser suppliers. Using an artificial intelligence approach, we attempt to determine the factors that lead to survival for these SMEs. The next section presents a literature review on successful small manufacturers and discusses the artificial intelligence technique, Adaptive Learning Network (ALN). The data description and model development are also presented, followed by the empirical results and discussion. Literature Review Successful manufacturers possess many characteristics which separate them from the competition and allow them to survive long-term in the marketplace. Chaneski (2004) suggested that successful manufacturing companies excel in quality, continuous improvement, customer

1

service, employee input, and employee training. Regarding quality, Roth and Miller (1992) found that “superstar” manufacturing firms emphasized formal quality initiatives, including continuous improvement and zero defects, more than low-performing firms. Abdul-Aziz, Chan and Metcalfe (2000) found that use of an in-process inspection system (along with pre- and postprocess inspection) was critical for manufacturing success, and Yusof and Aspinwall (2000) found that successful small manufacturers emphasized process improvement, process control, and employee involvement. Sun and Cheng (2002) noted that small manufacturers may favor more informal means of quality assessment over formal programs; however, Bentley (2003) suggested that small firms must embrace lean manufacturing techniques such as Six Sigma or ISO 9000 for successful performance and quality improvement. Successful manufacturers value their customers by providing solutions that make their life easier and by keeping them informed of money-saving ideas and industry changes (Chaneski, 2004). Bentley (2003) stressed that successful manufacturing companies develop an intimate relationship with their customers and are more capable than larger firms in adjusting to customer preferences. Gadenne (1998) identified a similar competitive advantage factor for successful SMEs that allows manufacturing firms to create a niche by changing their product mix to meet customer demands, and Pellham (2000) found that market-oriented manufacturing firms were better performers because they respond quickly to negative customer feedback, competitor activities, and customer changes. Excellent customer service not only involves external customers but internal ones as well—the employees (Chaneski, 2004). Keeping employees happy means making sure that the materials and information they receive and use is up to quality standards. Previous work by Roth and Miller (1992) and Corbett and Harrison (1992) noted that employee relationships were as important as the physical aspects of the business in terms of

2

superior performance. Dossenbach (2005) supported the need for good labor relations and suggested that manufacturing success depends on company-wide solidarity, which is achieved through open communication, integrity, and employee empowerment. As mentioned previously, success for some firms means simply staying in business, and this can be a difficult goal to achieve. This may be particularly true for small manufacturers in the mass merchandising marketplace as Udell, Atehortua, and Parker (1995) found that mass retail buyers considered only 1 in 300 small firms to be viable suppliers, firms likely to be longterm quality partners. How do they decide which firms are the best candidates for long-term partnerships? Quality and price seem to be the top two criteria used by buyers in the supplier selection process. For example, Pearson and Ellram (1995) found that small and large electronics firms selected suppliers based on quality and price, while Piercy and Cravens (1997) found that the main requirement for British importers was a quality product. Donovan (1996) agreed that pricing and product quality are important, but he suggested small manufacturers must improve their order processing in order to be considered as potential mass retail suppliers. Knotts, Jones, and Udell (2003) identified other important factors (management experience, employee input, and cash flow analysis) which helped manufacturers get their product reviewed by a mass retail buyer. Jones, Knotts, and Udell (2004) and Kim, Jones, and Knotts (2005) found that small manufacturers vying to be mass merchandising suppliers had a better chance at obtaining buyer review when they met high standards in the areas of technology transfer (product imitability), commercialization stage (market-readiness), merchandising potential (local vs. national appeal), demand stability, and perceived appearance. In summary, small firm success is often a result of the firm’s ability to match a quality product with a well-run firm. Customers wanting long-term partnerships with small

3

manufacturers are interested in getting quality products at reasonable prices, but for these small firms to accomplish this goal they must generally be run by experienced managers with welltrained and motivated employees. The firms must be healthy and properly capitalized in order to be able to take advantage of any reasonable opportunities presented in the marketplace. To this end, our study reviews the results of a long-term screening and educational project used by small manufacturers to assess their viability as suppliers to the mass retail market.

The Study The sample for this study consisted of small manufacturers who participated in a supplier evaluation program developed at a regional Midwest university. All of the participating firms were independently-owned manufacturers who were interested in supplying their product(s) to a major mass merchandiser. Of 2113 potential suppliers, 1690 (80.0 percent) completed both parts of the evaluation process, which included a firm self-assessment and an independent product evaluation. Nineteen percent (321 firms) were female-owned and managed. The respondents were from all states, and racial, ethnic, and other minority information were not kept as part of the main database. All firms supplied products exclusively for consumer purchase, and none of the firms was dominant in its industry. Products varied in suggested retail price from inexpensive and/or point-of-purchase to major purchase levels. The supplier evaluation program required a firm to complete two assessments: a selfappraisal of its management practices and an external review of its submitted product. Specific evaluation items for the firm assessment and production evaluation are found in the Appendix. Each product was either rejected from the program or sent on to the mass merchandiser for buyer

4

review based upon the results of these evaluations. The final decision as to whether the retailer accepted the forwarded product for sale on its shelf was left entirely to the retailer. Firm Assessment The firm assessment measure, which evaluated the management practices of potential suppliers, was a self-administered instrument used by program participants. The 34 items were based on prior research conclusions and discussion with potential buyers from the mass merchandiser industry. The items generally fell into the areas of marketing management, strategic management, production operations, and financial management. The firm self-assessment items were structured with evaluation statements and multiple levels of measurement scored from one to five points. The three-point (or middle) response was the minimum performance level acceptable to retail buyers. Firm owners rated their management practices by answering questions like the one below:

Marketing Plan. Does your firm have a marketing plan for this project? (1) We do not need a marketing plan for this project. (2) We have an informal, unwritten marketing plan. (3) We have an informal, written plan. (4) A formal, written marketing plan is in progress. (5) We have a formal, written marketing plan.

Product Evaluation The product evaluation instrument consisted of 41 items based on the Product Innovation Evaluation System (PIES) developed at the University of Oregon (Udell, O’Neill, & Baker, 1977). Product areas included societal impact, business risk, demand analysis, market acceptance, competitive capabilities, and experience and strategy. An independent, trained evaluator completed this portion of the assessment process. The independent evaluator was

5

typically a current or former retail buyer or an experienced small firm owner with a retail background whose role was to assess the mass market potential of the product. The product evaluation instrument was similar in structure to the firm self-assessment. Products were judged objectively on a five-point ordinal scale using specific achievement levels rather than a sliding subjective scale. The three-point (or middle) response was the minimum performance level acceptable to retail buyers. The independent evaluators rated each product using items like the one below:

Functional Feasibility. In terms of its intended functions, will it do what it is intended to do? This product: (1) is not sound; cannot be made to work. (2) won’t work now, but might be modified. (3) will work, but major changes might be needed. (4) will work, but minor changes might be needed. (5) will work; no changes necessary.

Adaptive Learning Network (ALN) The Adaptive Learning Networks (ALN) approach (R. L. Barron, Mucciardi, Cook, Craig, and A.R. Barron, 1984)is refined from the Group Method of Data Handling (GMDH) algorithm of the Ivakhnenko and Ivakhnenko (1974) method and developed from using advanced statistics, expert systems, and artificial intelligence research including artificial neural networks. The ALN model automatically generates the trained network from the database and performs a traditional task of fitting model coefficients to bases of observational data. The trained network is considered to be powerful because it allows decomposing complex problems into much smaller and simpler ones, and to solve them. The adaptive model obtained from the AIM synthesis process is a layered network of feed-forward functional elements (nodes), in which the coefficients, number and types of

6

network elements, and the connectivity are learned inductively and automatically. Each processing unit (node) has a unique equation of multi-variable configurations: singles, doubles, triples, normalizers, white elements, unitizers, and wire elements (Montgomery, 1989). Normalizers transform the original input variables into standardized normal variables with a mean of zero and a variance of one. The white element is a linear combination of all inputs to the current layer. Unitizers convert the normalized data back into the original data to assess the output values. The algebraic form of singles, doubles, and triples is shown in the following equations:

Single = W0 + W1*X1 + W2*X12 + W3*X13 Double = W0 + W1*X1 + W2*X2 + W3*X12 + W4*X22 + W5*X1*X2 + W6*X13 + W7*X23 Triple =W0 + W1*X1 + W2*X2 + W3*X3 + W4*X12 + W5*X22 + W6*X32 + W7*X1*X2 + W8*X1*X3 + W9*X2*X3 + W10*X1*X2*X3 + W11*X13 + W12*X23 + W13*X33

Where Xi and Wi denote specific input variables and coefficients respectively. These elements are homogeneous multinomials of degree 3 in one, two, three variables and allow interaction among input variables. It is well known that a suitably high degree multinomial---a polynomial of n variables in which all cross products appear and combinations of the variables to a different degree are included---can approximate arbitrary functions of many variables very accurately (Barron et al., 1984). All these terms in the equation may not always appear in a node since ModelQuestTM will throw out the terms that do not contribute significantly to output. The output

7

of elements in one layer will then feed into subsequent layers, together with the original input variables. Networks are synthesized from layer to layer until the adaptive network model ceases to improve based on Predicted Squared Error (PSE) criterion. The objective of the ALN algorithm is to train and identify the model that minimizes the predicted squared error (PSE), the errors on as yet unforeseen data, without over fitting the data (A.R. Barron, 1984). PSE consists of the training squared error (TSE) and overfit penalty as shown in the following:

PSE = TSE + 2  p

2

K N

Where TSE is the average squared error of the model on the training sample observations, K is the number of coefficients that are estimated to minimize TSE, σp2 is the prior estimate of true error variance, and N is the size of the training sample observation. The minimum PSE is always attainable because as each coefficient is added to the model, TSE decreases at a decreasing rate while the overfit penalty increases linearly. If the adaptive model is obtained by minimizing TSE alone, the model will perform well on the training data set, but it can perform poorly on evaluation samples. When the model has an overly complex structure and many coefficients, it will give a poor estimate of error on the test data set. By adding a term for overfit penalty, the minimum expected squared difference between the estimated model and the true model on the future data set can be obtained (A.R. Barron, 1984).

(Insert Figure 1 here)

An application of ModelQuestTM entails the following four steps as shown in Figure 1.

8

The first step involves identifying and characterizing data for a better solution. The second step involves representing and transforming the data through mapping, sampling, and feature extraction routines to provide additional inputs and to compensate for outliers and sparse regions. The third step is to split data into training and evaluation subsets, and train a model from a training data set. The ModelQuestTM automatically synthesizes the optimal ALN network including model size, connectivity, and parameter values. The last step involves applying the model to an evaluation data set and predicting output values. The accuracy and consistency of prediction determines how well the model works. Empirical Results The final ALN model is synthesized and presented in Figure 2. It is a layered network of feed-forward functional elements, which contain the best network structure, node types, coefficients, and connectivity to minimize the predicted squared error (PSE). The equations in Appendix B show the final ALN model in polynomial equation forms, and each equation number represents the node number of the ALN network as shown in Figure 1.

(Insert Figure 2 here)

The ALN model as shown in Figure 2 uses ten input variables to characterize viability of SMEs. These input variables are first transformed into standardized normal variables with a mean of zero and a variance of 1 using normalisers in Appendix B. These standardized variables were next fed into the first layer to generate a series of intermediate output values. For example, the node T(Triplet)100 was synthesized using normalized values of Per40 (Recommended Distribution Channels), Far07 (Company Orientation), and Per17 (Product Life Cycle), and then

9

fed into the node D(Doublet)117 together with T111 that was also synthesized using Far03 (Price Determination), Far29 (Cash Flow), and Per05 (Functional Feasibility). The value of the node D117 together with two other variables, Far27 (In-Process Inspection), and Per43 (EICBT and CCL is fed into the subsequent node T127 in the third layer, and the value of T127 is again fed into T133 in the fourth layer with two other variables: Far09 (Job Description) and Per39 (Management and Production experience) to synthesize. T133 is finally converted back to the original variables, Alive. Once this final ALN model is validated using a hold-out sample, it becomes a knowledge base from which the survivability of SMEs can be predicted.

Insert Table 1 here

The result of testing the ALN model using the hold-out sample is shown in Table 1. The contingency table shows that the ALN model predicted 86.17 percent of the firms that are currently in operation correctly using only 10 out of 77 input variables. It is not surprised to see that 74.43 percent of the non-operating firms were also categorized as currently operating firms primarily because these firms could have been survived unless they are acquired or merged with another firms; many successful SMEs are frequently acquired and merged as indicated earlier in the paper. The accuracy of the ALN model clearly indicates that the final ALN model is quite successful in evaluating the holdout sample. The ALN model first selected three input variables including recommended distribution channel (Per 40), company orientation (Far07), and product life cycle (Per17), which indicates that viable SMEs are producing quality products oriented toward more customer satisfaction with a longer product life cycle and more effective channels of distribution. Small manufacturers first

10

must focus on matching its product selection with current customer needs. In addition, these manufacturers must use more effective channels of distribution in order to make the products available at the retail stores as customers want. These findings support previous works by Gadenne (1998) and Pellham (2000), which focused on customer orientation, and by Kim et al. (2005), which emphasized appropriate distribution level and demand stability. Consider the manufacturer that attempts to sell local or regional products nationwide. Without a proper placement of the product according to its demand radius, the manufacturer could attempt to sell Poughkeepsie football jerseys in Los Angeles, or it might undersell jerseys from a team with national appeal (say, the Raiders or the Cowboys) by placing them in just their local markets. Additionally, focusing on the customers’ priorities through the proper company orientation is fundamental. Long gone are the days that manufacturers can simply create products for the marketplace and expect the public to buy whatever is on the shelf. Firms unable to adapt to the vagaries of the marketplace will find themselves without a loyal customer base. Finally, the SME must make a product that has a long enough life cycle to meet the needs of both the firm and the consumer. A firm unable to match its strategies with either long- or short-term product life cycles will find itself either under- or over-capitalizing, producing, marketing, etc. The viable manufacturers must have an ability to supply retailers with its product on a consistent basis. Quality manufacturing techniques such as quality control systems (Per43) and proper product and process inspections (Far 27) are integral to a company aiming at maintaining a consistent level of production (Abdul-Aziz et al., 2000; Bentley, 2003; Yusof & Aspinwall, 2000). Using these techniques, experienced manufacturers should be able to make products that function as they are supposed to (Par05: functional feasibility) at prices that consumers are willing to pay (Far03: price determination). Keeping track of company finances on a timely

11

basis (Far29: cash flow) (Kim et al., 2006) and helping employees know exactly what their roles within the firm are (Far09: job descriptions) round out the list of practical and critical variables selected by the model. These variables support prior works by Piercy and Cravens (1997), Donovan (1996), and Knotts et al. (2003).

(Insert Table 2)

The Sensitivity and Importance values of ten input variables are shown in Table 2. The sensitivity value indicates that the degree of influence one input variable has on the output variable with all other input values remaining at their median values. For example, over 26 percent of output (dependent variable) variability is attributable to the company orientation (Far07) variable. According to the sensitivity values, about 86 percent of the output variability is attributable to the first five variables including Far07, Far03 Per39, Per17, & Per43 as shown in Table 2. These five variables are the most influential predictor variables and therefore must be used to develop strategic planning for a long-term survival. This result indicates that SMEs must have an explicit goal of the organization and require having ample product and management experiences. The company must also produce products with a longer life cycle, and must have in-process inspection system to have quality control of products that are critical to consumer satisfaction. The importance values in Table 2 indicate that the degree of influence each selected input value has on the output variable over all ranges of inputs. The importance values show that Far07 (Company Orientation) is the most influential variable in which over 30 percent of the variability of the output variable is attributable to Far07. The study also finds that, if the SMEs use their

12

product and management experiences to produce their products appealing to the customer, they could survive over a longer period of time. In order to survive over a long time, the firms must produce quality product with a longer product life cycle at a competitive price. These findings reaffirm the strategic implications found through in sensitivity analysis.

Discussion and Conclusions This selection of independent or predictor variables for this model suggests two key goals for small manufacturers in their quest to make viable consumer products available to the marketplace. First, the SME must focus on the matching of its product selection with current customer needs. The first triple (recommended distribution channel, company orientation, and product life cycle) strongly hints at making long-term products available at the right consumer marketing level and making them as the customer wants them. This supports previous work by Gadenne (1998) and Pellham (2000), which focused on customer orientation, and by Kim et al. (2005), which emphasized appropriate distribution level and demand stability. Consider the manufacturer that attempts to sell local or regional products nationwide. Without a proper placement of the product according to its demand radius, the manufacturer could attempt to sell Poughkeepsie football jerseys in Los Angeles, or it might undersell jerseys from a team with national appeal (say, the Raiders or the Cowboys) by placing them in just their local markets. Additionally, focusing on the customers’ priorities through the proper company orientation is fundamental. Long gone are the days that manufacturers can simply create products for the marketplace and expect the public to buy whatever is on the shelf. Firms unable to adapt to the vagaries of the marketplace will find themselves without a loyal customer base. Finally, the SME must make a product that has a long enough life cycle to meet the needs of both

13

the firm and the consumer. A firm unable to match its strategies with either long- or short-term product life cycles will find itself either under- or over-capitalizing, producing, marketing, etc. The second focus is that of creating a viable firm with a reliable ability to supply retailers with its product on a consistent basis. Quality manufacturing techniques such as quality control systems and proper product and process inspections are integral to a company aiming at maintaining a consistent level of production (Abdul-Aziz et al., 2000; Bentley, 2003; Yusof & Aspinwall, 2000). Using these techniques, experienced manufacturers should be able to make products that function as they are supposed to (functional feasibility) at prices that consumers are willing to pay (price determination). Keeping track of company finances on a timely basis (cash flow) and helping employees know exactly what their roles within the firm are (job descriptions) round out the list of practical and critical variables selected by the model. These variables support prior work by Piercy and Cravens (1997), Donovan (1996), and Knotts et al. (2003). In the end, this study finds that firms that use viable organizations to create and market products appealing to the customer are those that are also more likely to survive over a longer period of time. While long-term survival is not the only measure of the success of a small manufacturing firm, it is one that encompasses other success measures such as sales performance and profitability. Generalization of the results to all firms is inappropriate since the participants were specifically small manufacturers wishing to enter the mass retail arena, but this segment of the market is fairly large compared to more focused niches. The conclusions are broadly fundamental, too, since the results show that the core underlying principles of quality business practices really are those that firms should use on a day-to-day basis. The crux of the matter is this: can a firm present itself as a healthy and responsive maker of consumer products to prospective retailers in the hopes of getting its products on the right shelves in order to be

14

successful? A firm that is unable to do this will not survive. A firm that places a reasonable focus on these two goals will tend to succeed. While nothing is certain in a dynamic marketplace, the firm that is best prepared to concentrate on getting these two strategies right will have the best chance at surviving over the long term.

15

References Abdul-Aziz, Z., Chan, J., & A. Metcalfe (2000). Quality practices in the manufacturing industry in the UK and Malaysia. Total Quality Management, 11 (December): 1053-1065. Barron, A. R. (1984) "Predicted Squared Error: A Criterion for Automatic Model Selection,” in Self-Organizing Methods in Modeling: GMDH Type Algorithms. Ed. S.J. Farlow. MarcelDekker, Inc.: New York. Chaneski, W. S. (2004). What Differentiates Today’s Successful Manufacturers? Modern Machine Shop, 76(8). Dossenbach, T. (2005). The circles of solidarity for manufacturing success. Wood & Wood Products: Management Matters, (March): 25-26. Gadenne, D. (1998). Critical success factors for small business: An inter-industry comparison. International Small Business Journal, 17: 36-51 Headd, B. (2003). Redefining business success: Distinguishing between closure and failure. Small Business Economics, 21: 51-61. Ivakhnenko, A., and N. Ivakhnenko (1974). “Long-Term Prediction of FMDH Algorithms Using the Unbiased Criterion and the Balance-of-Variables Criterion,” Soviet Automation Control 7 (4), 40-45. Jones, S. C., Knotts, T. L., & G. G. Udell (2004). The effect of product-related factors on small business failure. Business Journal, 19(1-2): 68-72. Kim, K. Jones, S. & T. Knotts (2005). Selecting and developing suppliers for mass merchandisers. International Journal of Manufacturing Technology and Management (IJMTM), 7(5/6): 566-580. Kim, K. Keller, C., & R. Bottin (2006). Examining the financial profile of small manufacturing enterprises (SME). Forthcoming in the International Journal of Manufacturing Technology and Management (IJMTM) Knotts, T. L., Jones, S. C., & G. G. Udell (2003). Small business failure: The role of management practices and product characteristics. Journal of Business and Entrepreneurship, 15(2): 48-63. Pearson, J. N. and L. M. Ellram (1995). Supplier selection and evaluation in small versus large electronics firms. Journal of Small Business Management, (October): 53-65. Pelham, A. M. (2000). Market orientation and other potential influences on performance in small and medium-sized manufacturing firms. Journal of Small Business Management, 38: 4867.

16

Piercy, N. F. and D. W. Cravens (1997). Examining the role of buyer-seller relationships in export performance. Journal of World Business, 32 (Spring): 73-86. Roth, A. V. and J. G. Miller (1992). Success factors in manufacturing. Business Horizons, 35 (July-August): 73-81. Sun, H., and T. Cheng. (2002). Comparing reasons, practices and effects of ISO 9000 certification and TQM implementation in Norwegian SMEs and large firms. International Small Business Journal, 20 (November): 421-442. Udell, G. G., O'Neill, M. F., & K. G. Baker (1977). Guide to Invention and Iinnovation Evaluation. Washington, D.C.: National Science Foundation. Udell, G. G.., Atehortua, C., & R. Parker (1995). The Support American Made Manual of Venture Assessment. United States. Yusof, S. M. and E. M. Aspinwall (2000) Critical success factors in small and medium enterprises: Survey results. Total Quality Management, 11 (July): 448-459.

17

Figure 1. ModelQuestTM Modeling Procedure

Feedback

Raw Data

Identify

Transfe r

Mode l

Analyz e

Solution

Figure 2. The Final Network Model Input Variable

First Layer

Per40

1

Far 07

2

Per17

3

Far03

4

Far29

5

Per05

6

Far27

7

Per43

8

Far09

9

Per39

10

Second Layer

Third Layer

Fourth Layer

T127

T133

Output

T100 D117

ALIVE

T111

18

Table 1. Contingency Table for Viability

Prediction Results Survived Not Forwarded Total

Survived 218 35 253

Sensitivity Specificity Type I Error Type II Error Overall Accuracy

86.17 25.57 13.83 74.43 61.31%

Actual Results Not Survived 131 45 176

Total 349 80 429

Table 2. Variable Sensitivity and Importance Input Description

Sensitivity

Importance

Far07 (Company Orientation) Far03 (Price Determination) Per39 (Production Experience) Per17 (Product Life Cycle) Per43 (Quality Control) Per40 (Distribution Channels) Per05 (Functional Feasibility) Far29 (Cash Flow) Far09 (Job Description) Far27 (In-Process Inspection)

.2627 .1773 .1705 .1361 .1122 .0504 .0424 .0124 .0197 .0073

.3040 .1874 .0990 .1563 .0905 .0751 .0275 .0256 .0267 .0079

19

Appendix A: Firm Assessment and Product Evaluation Rating Item Firm Evaluation (FAR) Marketing Management: 01 Marketing Plan 02 Marketing Organization 03 Price Determination 04 Market Demand 05 Competitive Product Analysis 06 Promotional Plan 07 Company Orientation Strategic Management: 08 Mission Statement 09 Job Descriptions 10 Employee Input 11 Management Experience 12 Quality 13 Firm’s Primary Objective 14 Use of Consultants 15 Business Plan 16 Board of Directors 17 Board Involvement Production Operations: 18 Product Testing 19 Research and Development 20 Manufacturing Technology 21 Management Planning and Control Systems 22 Delivery Schedule Reliability 23 Quality Control Measures 24 Maintenance Program 25 Cost Containment 26 First Piece Approval 27 In-Process Inspection 28 Continuous Improvement Program Financial Management: 29 Cash Flow 30 Budgetary Planning Cycle 31 Budget Update Cycle 32 Cost Accounting 33 Accounting 34 Financial Planning Product Evaluation (PER) Societal Impact: 01 Legality 02 Safety 03 Environmental Impact 04 Societal Impact

09 Payback Period 10 Profitability 11 Marketing Research 12 Research and Development Demand Analysis: 13 Potential Market 14 Potential Sales 15 Trend of Demand 16 Stability of Demand 17 Product Life Cycle 18 Product Line Potential Market Acceptance: 19 Use Pattern Compatibility 20 Learning 21 Need 22 Dependence 23 Visibility 24 Promotion 25 Distribution 26 Service Competitive Capabilities: 27 Appearance 28 Function 29 Durability 30 Price 31 Existing Competition 32 New Competition 33 Protection Experience and Strategy: 34 Technology Transfer 35 New Venture 36 Marketing Experience 37 Technical Experience 38 Financial Experience and Resources 39 Management and Production Experience 40 Promotional Requirements 41 Merchandising Potential 42 Product Capability 43 Product Quality Control 44 Marketing Capability 45 Engineering & Technical Capability 46 Financial Capability Overall evaluation (PER) 47 48

Overall Venture Readiness Overall Product Readiness

Business Risk: 05 Functional Feasibility 06 Production Feasibility 07 Commercialization Stage 08 Investment Costs

20

Appendix B. ALN Results and Equations Normalizers: Perrecom Far07 Per17 Far03 Far29 = Per05 = Far27 = Per43 = Far09 = Per39 =

= –1.6511 + .5583X1 = –2.6727 + .6521X1 = – 1.7511 + .6659X1 = –1.9971 + .761X1 –2.5375 + .617X1 –3.1212 + .706X1 –2.3371 + .6364X1 –2.3531 + .7506X1 –1.7063 + .5816X1 –2.7151 + .9363X1

Triplets: Triplet(100) = –0.0589 – 0.3219*X1 + 0.2124*X12 + 0.1005*X13 + 0.0624*X2 – 0.0262*X1*X2 – 0.079*X12*X2 + 0.1077*X1*X22 – 0.0045*X23 + 0.1367*X3 + 0.0832*X1*X3 + 0.0179*X12*X3 – 0.0847*X2*X3 + 0.0527*X1*X2*X3 – 0.1162*X22*X3 – 0.0212*X32 + 0.0403*X1*X32 + 0.0583*X2*X32 – 0.0109*X33 Triplet(111) = –0.0638 – 0.0511*X1 – 0.1145*X12 + 0.0495*X13 – 0.0226*X2 + 0.0236*X1*X2 + 0.0188*X12*X2 + 0.2597*X22 – 0.0832*X1*X22 + 0.0987*X23 – 0.2849*X3 – 0.1293*X1*X3 – 0.0448*X12*X3 + 0.3496*X2*X3 + 0.0303*X1*X2*X3 + 0.1465*X22*X3 + 0.6251*X32 – 0.038*X1*X32 + 0.0128*X2*X32 + 0.1736*X33 Doublet(117) =–0.0354 + 1.1304*X1 – 0.2649*X12 – 0.1756*X13 + 1.0643*X2 – 0.5943*X1*X2 + 0.1476*X12*X2 + 0.0732*X22 – 1.3497*X1*X22 – 0.1063*X23 Triplet(127) = 0.1225 + 0.8458*X1 – 0.1391*X12 + 0.2585*X1*X2 – 0.0499*X12*X2 – 0.17*X22 + 0.1363*X1*X22 – 0.0733*X23 + 0.1552*X3 + 0.2162*X1*X3 – 0.0482*X12*X3 + 0.0169*X1*X2*X3 – 0.0299*X22*X3 – 0.0786*X32 + 0.0752*X1*X32 + 0.0331*X2*X32 – 0.0397*X33 Triplet(133) = 0.0644 + 0.8676*X1 – 0.188*X12 – 0.1847*X13 + 0.0671*X2 + 0.2273*X1*X2 – 0.1177*X12*X2 – 0.0803*X22 + 0.1832*X1*X22 – 0.0358*X23 – 0.0969*X3 – 0.0719*X12*X3 + 0.0193*X1*X2*X3 + 0.1246*X32 + 0.0082*X1*X32 + 0.0425*X33 Unitizers: Alive =

.5963 + .4908X1