Ece UYGUN - Simplified Scoring Models

Simplified Scoring Models

 In the simplified scoring model, each criterion is ranked according to its relative importance. Our choice of projects will thus reflect our desire to maximize the impact of certain criteria on our decision. In order to score our simplified checklist, we assign a specific weight to each of our four criteria:

Criterion	Importance Weight
Importance Weight	3
Profit Potential	2
Development Risks	2
Cost	1

Example: Scoring Models

 Using the criterion weighting values we developed above, SAP Corporation is attempting to determine the optimal project to fund. As you can see in Table, although adding a scoring component to our simple checklist complicates our decision, it also gives us a more precise screening model—one that more closely reflects our desire to emphasize certain criteria over others.

Table 2 (Simple Scoring Model)

Project	Criteria	Importance Weight (A)	Score (B)	Weighted Score (A)x(B)
Project Alpha
	Cost	1	3	3
	Profit Potential	2	1	2
	Development Risk	2	1	2
	Time to Market	3	2	6
	Total Score			13

Project	Criteria	Importance Weight (A)	Score (B)	Weighted Score (A)x(B)
Project Beta
	Cost	1	2	2
	Profit Potential	2	2	4
	Development Risk	2	2	4
	Time to Market	3	3	9
	Total Score			19

Table 2 (Continued)

Project	Criteria	Importance Weight (A)	Score (B)	Weighted Score (A)x(B)
Project Gamma
	Cost	1	3	3
	Profit Potential	2	3	6
	Development Risk	2	3	6
	Time to Market	3	1	3
	Total Score			18

Project	Criteria	Importance Weight (A)	Score (B)	Weighted Score (A)x(B)
Project Delta
	Cost	1	1	1
	Profit Potential	2	1	2
	Development Risk	2	2	4
	Time to Market	3	3	9
	Total Score			16

SOLUTION

Table 1 (Simplified Checklist Model for Project Selection)

Project                  Criteria                             High                       Medium                              Low     

Project Alpha         
                               Cost                                   X                                                                     
                               Profit Potential                                                                                                X
                               Time to Market                                                     X                                
                               Development Risk                                                                                           X
Project Beta       
                               Cost                                                                     X                                                                     
                               Profit Potential                                                      X
                               Time to Market                   X                                
                               Development Risk                                                 X
Project Gamma         
                               Cost                                   X                                                                     
                               Profit Potential                    X
                               Time to Market                                                                                               X
                               Development Risk               X
Project Delta       
                               Cost                                                                                                               X
                               Profit Potential                                                                                                X
                               Time to Market                   X                                
                               Development Risk                                                  X

 In Table 2, the numbers in the column labeled Importance Weight specify the numerical values that we have assigned to each criterion: Time to Market always receives a value of 3, profit potential a value of 2, development risk a value of 2, and cost a value of 1. We then assign relative values to each of our four dimensions. The numbers in the column labeled Score replace the X’s of Table 1 with their assigned score values:

(High = 3, Medium = 2, Low = 1)

In Project Alpha, for example, the High rating given Cost becomes a 3 in Table 2 because High is here valued at 3. Likewise, the Medium rating given Time to Market in Table 1 becomes a 2. But notice what happens when we calculate the numbers in the column labeled Weighted Score. When we multiply the numerical value of Cost (1) by its rating of High (3), we get a Weighted Score of 3. But when we multiply the numerical value of Time to Market (3) by its rating of Medium (2), we get a Weighted Score of 6. We add up the total Weighted Scores for each project, and according to Table, Project Beta (with a total of 19) is the best alternative, compared to the other options: Project Alpha (with a total of 13), Project Gamma (with a total of 18), and Project Delta (with a total of 16).

Thus the simple scoring model consists of the following steps:

• Assign importance weights to each criterion: Develop logic for differentiating among various levels of importance and devise a system for assigning appropriate weights to each criterion. Relying on collective group judgment may help to validate the reasons for determining importance levels. The team may also designate some criteria as “must” items. Safety concerns, for example, may be stipulated as nonnegotiable. In other words, all projects must achieve an acceptable safety level or they will not be considered further.

• Assign score values to each criterion in terms of its rating (High = 3, Medium = 2, Low = 1): The logic of assigning score values is often an issue of scoring sensitivity—of making differences in scores distinct. Some teams, for example, prefer to widen the range of possible values—say, by using a 1-to-7 scale instead of a 1-to-3 scale in order to ensure a clearer distinction among scores and, therefore, among project choices. Such decisions will vary according to the number of criteria being applied and, perhaps, by team members’ experience with the accuracy of outcomes produced by a given approach to screening and selection.

• Multiply importance weights by scores to arrive at a weighted score for each criterion: The weighted score reflects both the value that the team gives each criterion and the ratings that the team gives each criterion as an output of the project.

• Add the weighted scores to arrive at an overall project score: The final score for each project becomes the sum of all its weighted criteria.

The pharmaceuticals company Hoechst Marion Roussel uses a scoring model for selecting projects that identifies not only five main criteria—reward, business strategy fit, strategic leverage, probability of commercial success, and probability of technical success—but also a number of more specific subcriteria. Each of these 19 subcriteria is scored on a scale of 1 to 10. The score for each criterion is then calculated by averaging the scores for each criterion. The final project score is determined by adding the average score of each of the five subcategories. Hoechst has had great success with this scoring model, both in setting project priorities and in making go/no-go decisions. The simple scoring model has some useful advantages as a project selection device. First, it is easy to use it to tie critical strategic goals for the company to various project alternatives. In the case of the pharmaceutical company Hoechst, the company has assigned several categories to strategic goals for its project options, including Business strategy fit and Strategic leverage. These strategic goals become a critical hurdle for all new project alternatives. Second, the simple scoring model is easy to comprehend and use. With a checklist of key criteria, evaluation options (high, medium, and low), and attendant scores, top managers can quickly grasp how to employ this technique.

Limitations of Scoring Models

   The simple scoring model illustrated here is an abbreviated and unsophisticated version of the weightedscoring approach. In general, scoring models try to impose some structure on the decision-making process while, at the same time, combining multiple criteria. Most scoring models, however, share some important limitations. A scale from 1 to 3 may be intuitively appealing and easy to apply and understand, but it is not very accurate. From the perspective of mathematical scaling, it is simply wrong to treat evaluations on such a scale as real numbers that can be multiplied and summed. If 3 means High and 2 means Medium, we know that 3 is better than 2, but we do not know by how much. Furthermore, we cannot assume that the difference between 3 and 2 is the same as the difference between 2 and 1. Thus in Table, if the score for Project Alpha is 13 and 19 is the score for Project Beta, may we assume that Beta is 46 percent better than Alpha? Unfortunately, no. Critics of scoring models argue that their ease of use may blind novice users to the sometimes-false assumptions that underlie them. From a managerial perspective, another drawback of scoring models is the fact that they depend on the relevance of the selected criteria and the accuracy of the weight given them.

    In other words, they do not ensure that there is a reasonable link between the selected and weighted criteria and the business objectives that prompted the project in the first place. Here’s an example. As a means of selecting projects, the Information Systems steering committee of a large bank adopted three criteria: contribution to quality, financial performance, and service. The bank’s strategy was focused on customer retention, but the criteria selected by the committee did not reflect this fact. As a result, a project aimed at improving service to potential new markets might score high on service even though it would not serve existing customers (the people whose business the bank wants to retain). Note, too, that the criteria of quality and service could overlap, leading managers to double-count and overestimate the value of some factors.Thus, the bank employed a project selection approach that neither achieved its desired ends nor matched overall strategic goals.