CHAPTER TWO: DECISION SUPPORT SYSTEMS AND BUSINESS PROCESSES
Managerial Decision-Making Challenges and Organizational Levels
Managerial Decision-Making Challenges: Managers face three primary challenges when making decisions in a modern business environment:
Analyzing Large Amounts of Information: Innovations in communication and globalization have increased the variables and dimensions required for decision-making. Example: Analyzing data from hotels to determine room discounts based on occupancy patterns.
Making Decisions Quickly: Time is critical, and manual data sifting is no longer feasible. Example: Resolving a missing reservation for an important customer when the hotel is fully booked.
Applying Sophisticated Analysis Techniques: Competitive environments require strategies like Porter's or forecasting for strategic decisions. Example: Implementing a loyalty program across hotels.
The Six-Step Decision-Making Process:
Problem Identification: Clearly define the problem. Example: A customer wants to return a product after the -day policy expired.
Data Collection: Gather facts (who, what, where, when, why, how) without rumors or opinions. Example: Gather details on the product issue and the customer's history.
Solution Generation: Detail every possible solution, including farfetched ones. Example: Allow return, deny return, or offer an exchange.
Solution Test: Evaluate based on feasibility (can it be done?), suitability (is it permanent or temporary?), and acceptability (consensus).
Solution Selection: Choose the best fit for business needs. Example: Allow the return to preserve customer loyalty.
Solution Implementation: Monitor results. If the problem is solved, the decision was correct; if not, restart the process.
Three Primary Organizational Levels:
Operational Level (Structured Decisions): Employees maintain core activities required for day-to-day operations. Decisions are structured, meaning established processes offer potential solutions.
Managerial Level (Semistructured Decisions): Employees evaluate operations to identify and adapt to change. Decisions are semistructured, meaning a few established processes help but are not enough for a definite recommendation.
Strategic Level (Unstructured Decisions): Managers develop overall business strategies and long-term goals. Decisions are unstructured, where no procedures or rules guide the decision-maker toward the correct choice.
Measuring Organizational Business Decisions: CSFs, KPIs, and MIS Metrics
Core Measurement Definitions:
Project: A temporary activity undertaken to create a unique product, service, or result.
Metrics: Measurements used to evaluate if a project is meeting its goals.
Critical Success Factors (CSFs): Crucial steps organizations perform to achieve goals and implement strategies. Examples include:
Creating high-quality products.
Retaining competitive advantages.
Reducing product costs.
Increasing customer satisfaction.
Hiring and retaining top professionals.
Key Performance Indicators (KPIs): Quantifiable metrics used to evaluate progress toward CSFs. Examples include:
Employee turnover rates.
Number of product returns.
Number of new customers.
Average customer spending.
Percentage of help desk calls answered within the first minute.
Efficiency MIS Metrics: Focus on the performance of the MIS system itself:
Throughput: Information volume that can travel through a system at any given time.
Transaction Speed: Time taken to perform a transaction.
System Availability: Number of hours the system is available for users.
Information Accuracy: Extent to which the system generates correct results across repeated executions.
Response Time: Time taken to respond to user interactions (e.g., a mouse click).
Effectiveness MIS Metrics: Focus on the impact MIS has on business processes:
Usability: Ease with which people find information or perform transactions.
Customer Satisfaction: Measured via surveys, retention percentages, and revenue increases per customer.
Conversion Rates: The number of first-time "touches" that result in a purchase; vital for evaluating internet ads.
Financial Metrics: Includes Return on Investment (ROI), Cost-Benefit Analysis, and Break-even Analysis (where revenues equal ongoing costs).
Benchmarking: The process of continuously measuring system results against baseline values (Benchmarks) to identify improvements.
Management Information Systems (MIS) for Decision Support
Models: Simplified representations or abstractions of reality used to calculate risks, understand uncertainty, change variables, and manipulate time.
Operational Support Systems:
Transactional Information: Information contained in a single business process to support daily operational decisions.
Transaction Processing System (TPS): Basic system assisting in structured decisions. Process steps include CRUD (Create, Read, Update, Delete), calculate, and summarize. Input includes source documents (original transaction records).
Online Transaction Processing (OLTP): Use of technology to capture and update transaction/event information.
Managerial Support Systems:
Analytical Information: Organizational information used for semistructured decisions.
Decision Support System (DSS): Models information to support managers. Capabilities include what-if analysis, sensitivity analysis, goal-seeking, and optimization.
Online Analytical Processing (OLAP): Manipulation of information to create business intelligence.
Strategic Support Systems:
Executive Information System (EIS): Specialized DSS for senior executives handling unstructured, nonroutine decisions requiring insight. They rely on external data and internal TPS/DSS feeds.
Granularity: The level of detail in a model. Operational levels have fine granularity; strategic levels have coarse granularity.
Visualization: Produces graphical displays of patterns in large data sets. Infographics are graphical formats designed for instant understanding.
Digital Dashboards: Tools that track KPIs and CSFs by tailing information from multiple sources.
Artificial Intelligence (AI) and Machine Learning in Business
Artificial Intelligence (AI): Simulates human intelligence, including reasoning and learning.
Expert Systems: Computerized advisory programs imitating expert reasoning.
Genetic Algorithms: Systems mimicking evolutionary "survival-of-the-fittest" processes to generate better solutions.
Robotic Process Automation (RPA): Software with AI/ML capabilities handling high-volume, repeatable tasks.
Machine Learning (ML): Enables computers to understand concepts and learn from the environment. Types include Supervised, Unsupervised, and Transfer modern learning.
Data Augmentation: Adding training examples by transforming existing ones.
Overfitting: The model matches training data too closely and fails on new data.
Underfitting: The model fails to learn the underlying complexity of the training data.
AI for Business Decision Support:
Automation: Building robots operating like humans.
Complex Analytics: Analyzing massive data to find patterns.
Fraud Detection: Identifying unfamiliar spending patterns.
Resource Scheduling: Maximizing efficiency in hospitals or airports.
Bias in Machine Learning:
Sample Bias: Using incorrect/incomplete training data.
Prejudice Bias: Results influenced by cultural stereotypes.
Measurement Bias: Skewed data collection.
Variance Bias: A mathematical property of the algorithm itself.
Human Cognitive Biases: Affinity bias (similar backgrounds), Conformity bias (following others), Confirmation bias (seeking backing for preconceived ideas), and Name bias (preference for certain names).
Neural Networks and Advanced Concepts:
Neural Network: Emulates the human brain.
Fuzzy Logic: Math method for handling imprecise/subjective info.
Deep Learning: Specialized algorithms for complex data modeling.
Reinforcement Learning: Training models for a sequence of decisions.
Autonomous Robotics: Robots capable of independent decision-making.
Computerized Vision:
Machine Vision: Ability of a computer to "see" by digitizing images and acting on data.
Sensitivity: Ability to see in dim light/invisible wavelengths.
Resolution: Extent of differentiating between objects; better resolution often limits the field of vision.
Virtual Reality (VR): Computer-simulated environments.
Augmented Reality (AR): Physical world viewed with computer-generated layers added.
Managing and Modeling Business Processes
Business Processes Across Departments:
Accounting/Finance: Statements, accounts payable, accounts receivable.
Marketing/Sales: Discounts, campaigns, attracting customers, sales processing.
Operations Management: Ordering inventory, production schedules, manufacturing.
Human Resources: Hiring, health care enrollment, tracking vacation/sick time.
Process Classifications:
Customer-facing Process: Results in a product/service received by an external customer (e.g., billing, shipping).
Business-facing Process: Invisible to the external customer but essential (e.g., strategic planning, training, purchasing raw materials).
Core Processes: Primary activities in the value chain.
Static Processes: Systematic approach for continuous efficiency improvement.
Dynamic Processes: Continuously changing for ever-evolving operations.
Business Process Patent: Protects a specific set of procedures for a business activity.
Business Process Modeling and Notation (BPMN): Graphical notation for process steps.
Event (Circle): Anything happening during a process (e.g., customer request).
Activity (Rounded Rectangle): A task or work being performed.
Gateway (Diamond): Controls flow; handles forking, merging, or joining paths.
Flow (Arrow): Displays the path of the process.
As-Is and To-Be Models:
As-Is process model: Represents the current state of operations without changes.
To-Be process model: Represents the result of applied improvements.
Example: A burger order process simplified from "taking separate orders for fries and drinks" (As-Is) to a single "combo meal" choice (To-Be).
MIS-Driven Business Process Improvement
Workflow: Tasks and responsibilities required for each step in a process.
Digitization: Automating manual, paper-based processes into digital formats.
Levels of Process Change:
Automation (Operational): Computerizing manual tasks to lower costs. Static and routine.
Streamlining (Managerial): Eliminating unnecessary steps to improve efficiency. Semidynamic.
Bottleneck: Resource reaching full capacity.
Redundancy: Tasks unnecessarily repeated.
Business Process Reengineering (BPR) (Strategic): Radical analysis and redesign of workflow. Dynamic and nonroutine.
Metaphor: Automation is moving from foot to car on the same road; BPR is using an airplane to ignore the road entirely.
Case Study: Progressive Insurance: Reengineered the claims process. Traditional resolution took to weeks (moving from tow to customer); Progressive mobile claims resolution takes minutes to hours (on-site offer, towing, and ride home).
Questions & Discussion
What are the informational requirements across the organizational pyramid? At the bottom (Operational), data is fine-grained, transactional, and structured via OLTP. At the top (Strategic), data is coarse-grained, analytical, and unstructured via OLAP.
How do TPS, DSS, and EIS interact? The Transaction Processing System (TPS) feeds sales, manufacturing, and transportation data into the Decision Support System (DSS) and the Executive Information System (EIS). The EIS also captures external industry and market outlooks to generate executive reports.