Business Analytics for Managerial Decision-Making

Managerial Decision-Making and Business Analytics

Types of Managerial Decisions

To effectively plan, coordinate, and lead, managers make several types of decisions:

• Strategic Decisions: Address high-level issues and the overall direction of the organization. They define future goals and are long-term and complex.
• Tactical Decisions: Focus on how to achieve the goals and objectives set by the strategy. These are typically made by mid-level management for the medium term.
• Operational Decisions: Pertain to day-to-day operations. They are made by operations managers and are often simple and routine.

The Decision-Making Process (DMP)

A structured approach to decision-making involves several key steps:

1. Identify and define the problem.
2. Determine the criteria for evaluation.
3. Determine the set of alternative solutions.
4. Evaluate the alternatives.
5. Choose an alternative.

The Role of Business Analytics

Business analytics transforms data into actionable insights, aiding decision-making by:

• Providing accurate forecasting.
• Helping to quantify risk.
• Yielding better alternatives through analysis and optimization.

Types of Analytics

Descriptive Analytics

Describes what has happened in the past. Examples include reports, data dashboards, data queries, and data mining techniques like cluster analysis, sentiment analysis, and credit ratings.

Predictive Analytics

Uses data to predict the future or ascertain the impact of one variable on another. Examples include surveys and techniques such as linear regression, time series analysis, data mining, and simulation.

Prescriptive Analytics

Indicates the best course of action to take. A predictive model combines predictions with rules (a rule-based model). For example, portfolio models use investment return data to recommend optimal investments.

• Simulation Optimization: Combines probability and statistics to model uncertainty with optimization techniques, helping to find good decisions in highly complex and uncertain settings.
• Decision Analysis: Used to develop an optimal strategy when facing several decision alternatives and an uncertain set of future events. It often employs utility theory.

The Four V's of Big Data

• Volume: The scale of data.
• Velocity: The speed at which data is generated.
• Variety: The different forms of data.
• Veracity: The uncertainty or quality of data.

Data Visualization and Reporting

Data Dashboards

A data dashboard is a visualization tool that displays multiple metrics and automatically updates as new data becomes available.

Principles of Effective Data Dashboards

Effective dashboards often feature Key Performance Indicators (KPIs), which are critical metrics for success.

• Examples of KPIs: An automobile dashboard shows speed and fuel level; a business dashboard might show financial position, inventory levels, or customer service metrics.
• Timeliness: Should provide timely summary information on relevant KPIs.
• Clarity: Should present all KPIs on a single screen for a quick scan of the business's current state.
• Relevance: KPIs must be meaningful and related to the user's decisions.
• Attention-Grabbing: Should highlight unusual measures that may require attention.
• Use of Color: Color should be used sparingly to call attention to specific values or differentiate categories.

Data Visualization Techniques

Table Design Principles

• Avoid using vertical lines unless necessary for clarity.
• Use horizontal lines mainly to separate column titles from data or to indicate a calculation.

Common Chart and Table Types

• Crosstabulation: A table used to describe and summarize the relationship between two variables.
• Bar and Column Charts: Use horizontal (bar) or vertical (column) bars to display the magnitude of a quantitative variable, making them ideal for comparing categorical variables.
• Pie Chart: A circular chart used to compare categorical data as proportions of a whole.
• Bubble Chart: A two-dimensional graph that visualizes three variables, often as an alternative to 3-D graphs.
• Heat Map: A two-dimensional graphical representation of data where different shades of color indicate magnitude.
• Scatter Chart: A graphical presentation of the relationship between two quantitative variables.

Key Concepts in Analytics and Modeling

Modeling and Optimization Concepts

• Influence Diagram: A visual representation showing which entities influence others within a model.
• Optimal Point: In geometric terms, the point where binding constraints intersect.
• Unbounded Solution: In linear programming, an unbounded solution typically indicates that the problem formulation is improper or incomplete.
• Proportionality Assumption: A core assumption in linear programming stating that the contribution to the objective function and the resources used are proportional to the value of each decision variable.
• Binding vs. Non-Binding Constraints: A constraint is binding if it is met exactly at the optimal solution (LHS = RHS), meaning the resource is fully used. A non-binding constraint has leftover capacity.
  • Slack: The amount of unused resource in a non-binding less-than-or-equal-to (<=) constraint.
  • Surplus: The amount of over-satisfaction in a non-binding greater-than-or-equal-to (>=) constraint.
• Feasible Solution: A set of decision variables that satisfies all problem constraints.
• Optimal Solution: A feasible solution that achieves the best possible objective function value.
• Feasible Region: The area defined by the problem's constraints. For 'greater than' constraints, it is typically to the right of the line; for 'less than' constraints, it is to the left.
• Impact of Changing Constraints (RHS): Increasing the right-hand side (RHS) of a '<=' constraint expands the feasible region, while decreasing it shrinks the region. The opposite is true for '>=' constraints. An expanded feasible region cannot result in a worse optimal solution.
• Utility Theory: The study of the total worth or desirability of an outcome, reflecting a decision-maker's attitude toward factors like profit, loss, and risk.

Excel Functions and Features

Cell Referencing

• Relative Referencing (e.g., A1): The default; cell references change when the formula is copied.
• Absolute Referencing (e.g., $A$1): The column and row are both frozen and do not change when the formula is copied.
• Mixed Referencing (e.g., $A1 or A$1): Either the column or the row is frozen. For example, in A$9, only the row (9) is frozen.

Common Functions

• =SUM(range): Adds all the numbers in a range of cells.
• =SUMPRODUCT(array1, array2): Multiplies corresponding components in the given arrays and returns the sum of those products.
• =IF(logical_test, value_if_true, value_if_false): Returns one value if a condition is true and another value if it's false. Example: =IF(C2>B2, "Over Budget", "Within Budget").
• =AND(logical1, logical2, ...): Returns TRUE if all arguments are true. Often used within an IF function. Example: =IF(AND(C4="Excellent", E4<20), "Yes", "No").
• =OR(logical1, logical2, ...): Returns TRUE if any argument is true. Can be nested with AND and IF.
• =COUNTIF(range, criteria): Counts the number of cells within a range that meet a given criterion. Example: =COUNTIF(E15:E25, ">30000").
• =MAX(range) / =MIN(range): Returns the largest or smallest value in a set of values, respectively.
• =AGGREGATE(...): Performs a calculation on a range of cells while optionally ignoring hidden rows or error values.

Data Handling

• Filtering: Displaying only the rows in a dataset that meet certain conditions.
• Sorting: Arranging data in a specific order (e.g., ascending or descending).
• Pivot Tables: A powerful tool to summarize, group, and aggregate large datasets.

Statistical Definitions

• Mean: The average of a dataset, calculated by summing the values and dividing by the count of observations.
• Median: The middle value in a dataset when arranged in ascending order.
• Mode: The value that appears most frequently in a dataset.
• Variance: A measure of how spread out the data values are from the mean.
• Range: The difference between the largest and smallest values in a dataset.
• Time Series Data: Data collected over a period of time (e.g., minutes, days, years).
• Histogram: A graphical display of data using bars of different heights to show the frequency of data points in successive intervals or bins.
• Categorical Data: Data identified by labels or names (e.g., city, product type). Arithmetic operations cannot be performed on this type of data.