What Is a Decision Matrix?
Decision matrix is often used in the strategic operation and management of enterprises. It is a matrix expression that expresses the relationship between the decision plan and related factors. Often used for quantitative decision analysis.
Decision matrix
- Chinese name
- Decision matrix
- Foreign name
- selection matrix or grid
- Content
- One of the commonly used analysis methods of risky decision
- Applications
- When making decisions based on several criteria
- Decision matrix is often used in the strategic operation and management of enterprises. It is a matrix expression that expresses the relationship between the decision plan and related factors. Often used for quantitative decision analysis.
- 1. State variables: refer to various objective external conditions or natural states that may affect the consequences of decision-making. Is an uncontrollable factor.
- 2. Decision variables: Refer to the various action plans taken by decision makers, which are controllable factors.
- 3. Probability: refers to the probability of occurrence of various natural states.
- 4. Profit and loss value: The profit and loss value of the results obtained by selecting a certain scheme in a natural state.
- The decision matrix consists of alternatives, natural states (and their probability of occurrence), and profit and loss values. Generally, the matrix decision table is listed with conditions given by actual problems. In business management, the description of decision-making problems is concentrated on the decision-making matrix. Decision analysis is based on the decision-making matrix and uses different analysis standards and methods to select the best solution from several feasible solutions.
- 1 Use brainstorming to find applicable evaluation criteria. This process is best done with customer involvement.
- 2 Discuss and modify the evaluation criteria, and distinguish between "must" and "must not". The most important of these criteria may be the use of list reduction methods and multi-round voting methods.
- 3 Assign a weight to each standard according to its importance, with a total score of 10 points. The distribution of weights can be done through discussion and voting. Or each group member assigns a weight to each standard, adds the weights obtained by each standard, and sorts by the total weight sum.
- 4 Draw the L-shaped matrix. The evaluation criteria are placed at the top, and the options are arranged on the left. It is customary to use fewer entries as column entries.
- 5Evaluate each option according to the standard. There are three options.
- Option 1: Establish a level for each standard, such as:
- 1,2,3: 1-slightly, 2-part, 3-largely
- Or 1,2,3: 1-low, 2-medium, 3-high
- Or 1,2,3,4,5: 1-a little, ..., 5-a lot
- Or l, 4, 9: 1-low, 4-medium, 9-high
- Ensure that the levels established are consistent. Write down the criteria and make the maximum (5 or 3) the most desirable choice: the most impactful, most important, easiest, and most likely to succeed for the customer.
- Option 2: Sort the options according to how well each option meets the criteria. l indicates the option that does not conform to the standard.
- Option 3: Decision matrix: Establish a benchmark, which can be a choice or the current product or service: For each criterion, compare each choice with the benchmark and score, poor (-1), same (0), good (+ 1). You can also use more detailed standards, such as "2,1,0,1,2" five levels or "3,2,1,0, -1, -2, -3" seven levels. Similarly, a positive number indicates the level you want to achieve.
- 6 Multiply the ordering of each option by the weights and then add up. The highest scoring option is not necessarily a required option, but the relative size of the scoring is very meaningful for the discussion of the problem and can help us finally reach an agreement.
- Figure 5.50 is a decision matrix used by the customer service team of the "Parisian style" hotel to decide which of the "long waiting times" should be solved first. Problems arise in the areas of "customers waiting for reception", "customers waiting for waiters", "customers waiting for food" and "customers waiting for payment".
- The evaluation criteria are "inconvenience to customers" (to what extent does it have an adverse effect on customers?), "Easiness to resolve", "impact on other systems", and "speed of resolution". Originally "easiness of solving" was often written as "degree of difficulty of solving", but that reversed the ordering. The maximum value of each standard now represents the most desired choice: inconvenience to customers, easy resolution, great impact on other systems, and quick resolution.
- A five-point weight for "inconvenience to customers" indicates that the group believes this is the most important evaluation criterion. "Ease of Resolution" and "Speed of Resolution" each have a weight of 2 points. "Impact on other systems" has a weight of 1 point.
- Use 3, 2, and 1 to indicate that each question is divided into three levels: high, medium, and low. For example, in the question "Customers are waiting for food", the inconvenience caused to customers is rated intermediate (2) because the surrounding atmosphere is very good. Because it involves waiters and kitchen staff, this problem is not easy to solve (1-not easy). The impact on other systems is moderate, because the waiter must go to the kitchen more often. Because the kitchen is crowded and inflexible, this problem cannot be solved quickly (1-low speed). (Note: The team assumes that the solution involves a redesign of the kitchen, which may or may not be a good assumption.)
- Multiply each score with a weight. For example, the item "customer waiting for reception" is evaluated as "high (3)" on the question "inconvenience to customer" (weight 5), and the score is 15. Add the scores of each row to get the total score for each question. "Customer waiting for reception" received a maximum score of 28. Since the second highest score is 18, it becomes obvious which one to choose.
- You can also read the case of Medrad to see how the decision matrix can be used to determine which problem to prioritize.