Decision Making & Decision Support Systems: Models, Types & Tools

TAKEAWAYS:

• Decision-making ranges from rational (optimal) to bounded rationality (satisficing) to intuitive, with each model suited to different business contexts.
• Decision Support Systems (DSS) combine data, models, and user interaction to improve decision quality, while Group DSS (GDSS) enhance collaboration in team settings.
• Types of DSS (model-driven, data-driven, communication-driven, knowledge-driven) serve distinct business needs, from financial forecasting to strategic planning.
• Real-world examples like Nepal Rastra Bank’s monetary policy decisions (model-driven DSS) or Daraz’s dynamic pricing (data-driven DSS) show how these systems drive efficiency.
• OLAP cubes and what-if analysis are key tools in DSS for exploring business scenarios interactively.
• Exam focus: Be ready to compare DSS types, explain Simon’s decision-making models, and apply DSS concepts to case studies (e.g., NTC’s network optimization).

1. Decision Making: Models and Process

Decisions are the backbone of business strategy. Herbert Simon’s decision-making models classify how humans and organizations make choices under uncertainty.

1.1 Simon’s Decision-Making Models

Simon proposed three models to explain how decisions are made in real-world constraints:

mindmap
  root((Decision-Making Models))
    Rational Model
      "Optimal choice under certainty"
      "All alternatives & outcomes known"
      "Maximizing utility"
    Bounded Rationality
      "Limited information & cognitive ability"
      "Satisficing (good enough) rather than optimizing"
      "Heuristics used for speed"
    Intuitive Model
      "Experience-based, subconscious judgment"
      "Fast, but may lack logical rigor"
      "Used in crises or expert domains"

Key Takeaway:

• Rational Model: Assumes perfect information (theoretical, rarely achievable).
• Bounded Rationality: Most practical for businesses (e.g., a manager choosing a supplier based on partial data).
• Intuitive Model: Used by experts (e.g., a chef adjusting recipes based on taste, not data).

1.2 The Decision-Making Process

A structured approach to decision-making involves:

1. Problem Identification: Recognize the issue (e.g., declining sales in a Daraz category).
2. Data Collection: Gather relevant data (e.g., customer feedback, market trends).
3. Alternative Generation: Brainstorm solutions (e.g., discount offers, new product lines).
4. Evaluation: Assess alternatives using criteria (e.g., cost, customer response).
5. Selection: Choose the best option (e.g., launch a limited-time discount).
6. Implementation: Execute the decision.
7. Feedback: Monitor results and adjust (e.g., track sales impact).

A visual of the 7-step process with arrows connecting each stage. (Image: DavidLevinson, CC BY-SA 3.0, via Wikimedia Commons)

2. Decision Support Systems (DSS)

DSS are interactive computer-based systems that help managers make better decisions by combining data, models, and user input.

2.1 Components of a DSS

A typical DSS consists of:

• Database: Stores raw data (e.g., sales records, customer demographics).
• Model Base: Contains analytical models (e.g., regression, simulation).
• User Interface: Allows interaction (e.g., dashboards, query tools).
• Dialogue System: Manages user-DSS communication (e.g., prompts for inputs).

flowchart TD
  A["User Input"] --> B["Dialogue System"]
  B --> C["Database"]
  B --> D["Model Base"]
  C --> E["Data Analysis"]
  D --> E
  E --> F["Output/Recommendations"]
  F --> A

2.2 Types of DSS

DSS can be classified based on their primary function:

3. Group Decision Support Systems (GDSS)

GDSS are designed to improve collaboration in decision-making by:

• Reducing groupthink (pressure to conform).
• Enhancing anonymous input (prevents bias from dominant personalities).
• Providing structured voting and consensus-building tools.

Real-World Example: Kathmandu Traffic Management The Kathmandu Metropolitan City uses a GDSS to optimize traffic signals. Sensors collect real-time data, and a centralized system adjusts signal timings dynamically. This reduces congestion by 15% in peak hours.

flowchart LR
  A["Traffic Sensors"] --> B["Central GDSS Server"]
  B --> C["Data Analysis Engine"]
  C --> D["Signal Timing Model"]
  D --> E["Adjusted Traffic Lights"]
  E --> F["Reduced Congestion"]

4. Tools and Techniques in DSS

4.1 OLAP (Online Analytical Processing)

OLAP enables multidimensional analysis of data, allowing users to:

• Slice and dice data (e.g., view sales by region, product, and time).
• Perform drill-down (e.g., from total sales to daily sales per store).
• Use pivot tables for dynamic reporting.

Example: Nabil Bank’s Loan Portfolio Analysis Nabil Bank uses OLAP to analyze loan defaults across branches. A manager can:

1. Slice: View defaults by branch.
2. Dice: Compare defaults by loan type (e.g., personal vs. business).
3. Drill-down: See individual borrower details.

4.2 What-If Analysis

A tool to simulate scenarios before making decisions. For example:

• Daraz’s Warehouse Location: What if we open a new warehouse in Pokhara? The DSS can simulate delivery times and costs.
• NTC’s Network Expansion: What if we add 100 new towers? The model predicts coverage and ROI.

A 3D cube showing dimensions like Time, Product, and Region. (Image: Margaret Rouse, olap.com, Guru99, CC BY-SA 4.0, via Wikimedia Commons)

5. Advantages and Limitations of DSS

6. Case Study: Nepal Stock Exchange (NEPSE) and DSS

Problem: NEPSE needed to predict market trends and identify high-risk stocks to prevent crashes. Solution: Implemented a hybrid DSS combining:

• Data-Driven: Historical stock prices, trading volumes.
• Model-Driven: Machine learning for trend prediction.
• Knowledge-Driven: Expert rules from economists.

Outcome:

• Reduced false trading signals by 30%.
• Enabled early warnings for volatile stocks (e.g., during the 2020 COVID-19 dip).

In the Real World

1. eSewa’s Fraud Detection

   • Idea Used: Knowledge-Driven DSS with rule-based systems.
   • How: eSewa’s DSS flags suspicious transactions (e.g., sudden large payments) by comparing them against user behavior patterns. If a transaction deviates from the norm (e.g., a farmer in Pokhara suddenly paying for a luxury watch), the system triggers a manual review.

2. Pathao’s Dynamic Pricing

   • Idea Used: Data-Driven DSS with real-time demand-supply analysis.
   • How: During peak hours (e.g., 7–9 PM in Kathmandu), Pathao’s algorithm increases fares by 20–50% if driver supply is low. This balances demand and ensures drivers earn more during busy times. The system uses OLAP to analyze historical surge pricing data.

3. NTC’s Network Optimization

   • Idea Used: Model-Driven DSS for linear programming.
   • How: NTC uses DSS to optimize the placement of 4G towers in remote areas like Dolpa. The model calculates the minimum number of towers needed to cover 90% of the population while minimizing costs. This reduced capital expenditure by 18% in the last fiscal year.

Exam Tip

1. Compare DSS Types: Always relate examples to the four DSS types (model-driven, data-driven, etc.). For instance:

   • "Nepal Rastra Bank uses a model-driven DSS for inflation forecasting, while Daraz uses a data-driven DSS for pricing."

2. Simon’s Models: Expect short-answer questions on the differences between rational, bounded rationality, and intuitive models. Use the mindmap above to recall key points.

3. Case Study Application: In long-answer questions, apply DSS concepts to real scenarios:

   • "How would you design a DSS for Khalti’s loan approval process?" Answer: Use a hybrid DSS—data-driven for credit scoring, knowledge-driven for fraud rules, and communication-driven for team reviews.

4. OLAP and What-If: Be ready to explain OLAP operations (slice, dice, drill-down) and what-if scenarios with numerical examples. For instance:

   • "If Nabil Bank increases its loan interest rate by 2%, how would it affect default rates?" (Use a hypothetical table with before/after scenarios.)

5. Group DSS: Highlight collaboration benefits in exams. For example:

   • "GDSS helps Nepal Investment Bank’s loan committees avoid groupthink by allowing anonymous voting on risky loans."

Final Note: Business Intelligence exams test both theory and application. Always tie concepts to real-world examples (e.g., NTC, Daraz, banks) to score full marks. Practice drawing DSS architectures and OLAP cubes in your notes—they often appear in diagrams!