1.2 Types of Analytics: Descriptive, Predictive, and Prescriptive

Business analytics is all about using data to make smarter decisions. There are three main types: descriptive, predictive, and prescriptive. Each type builds on the last, giving you more insight and power to improve your business.

These analytics types form a progression from understanding the past to shaping the future. Descriptive looks back, predictive looks forward, and prescriptive tells you what to do next. Knowing when to use each type is key to getting the most value from your data.

Descriptive vs Predictive vs Prescriptive Analytics

Analyzing Historical Data

Top images from around the web for Analyzing Historical Data

• 

  Gegevensanalyse - Data analysis - other.wiki View original

  Is this image relevant?

• 

  Performance Patterns - HPC Wiki View original

  Is this image relevant?

• 

  Historical Modeling makes offline data analysis easy | Q.E.D. Code View original

  Is this image relevant?

• 

  Gegevensanalyse - Data analysis - other.wiki View original

  Is this image relevant?

• 

  Performance Patterns - HPC Wiki View original

  Is this image relevant?

1 of 3

Top images from around the web for Analyzing Historical Data

• 

  Gegevensanalyse - Data analysis - other.wiki View original

  Is this image relevant?

• 

  Performance Patterns - HPC Wiki View original

  Is this image relevant?

• 

  Historical Modeling makes offline data analysis easy | Q.E.D. Code View original

  Is this image relevant?

• 

  Gegevensanalyse - Data analysis - other.wiki View original

  Is this image relevant?

• 

  Performance Patterns - HPC Wiki View original

  Is this image relevant?

1 of 3

• Descriptive analytics focuses on analyzing historical data to gain insights into what has happened in the past
  • Uses data aggregation and mining techniques to identify patterns and trends
  • Helps understand past performance and generate reports for stakeholders
  • Does not provide insights into future outcomes or recommend actions

Forecasting the Future

• Predictive analytics utilizes historical data, statistical modeling, and machine learning techniques to make predictions about future outcomes or events
  • Employs techniques such as regression analysis, time series analysis, and machine learning algorithms (decision trees, neural networks, support vector machines)
  • Helps forecast future trends, identify risks and opportunities, and support decision-making
  • Accuracy of predictions depends on the quality and relevance of input data and robustness of models used
  • Used in areas such as demand forecasting, customer churn prediction, and fraud detection

Recommending Actions

• Prescriptive analytics goes beyond prediction by recommending specific actions or decisions based on insights gained from descriptive and predictive analytics
  • Uses optimization and simulation techniques such as linear programming, integer programming, and goal programming
  • Requires clear understanding of business objectives, constraints, and trade-offs
  • Ability to implement and monitor recommended actions is crucial
  • Helps optimize decision-making, resource allocation, and business processes (production planning, supply chain optimization, dynamic pricing)

Applications and Limitations of Analytics

Understanding Past Performance

• Descriptive analytics is useful for understanding past performance, identifying trends, and generating reports
  • Aggregates data from various sources (databases, spreadsheets, transactional systems) to provide a comprehensive view
  • Identifies patterns and trends that may not be immediately apparent
  • Generates reports and visualizations to communicate insights to stakeholders
  • Limited in its ability to predict future outcomes or recommend actions

Supporting Decision-Making

• Predictive analytics can help forecast future trends, identify risks and opportunities, and support decision-making
  • Utilizes historical data and relevant external data to make predictions
  • Supports decision-making in areas such as inventory management, targeted marketing campaigns, and risk assessment
  • Accuracy of predictions relies on the quality and relevance of input data and the robustness of models used
  • Requires ongoing monitoring and updating of models to maintain accuracy

Optimizing Business Processes

• Prescriptive analytics can optimize decision-making and resource allocation by providing specific recommendations
  • Incorporates data from descriptive and predictive analytics, as well as business objectives and constraints
  • Provides actionable recommendations to optimize processes (manufacturing schedules, delivery routes, pricing strategies)
  • Requires a clear understanding of business objectives, constraints, and trade-offs
  • Success depends on the ability to implement and monitor recommended actions

Data Requirements for Analytics

Descriptive Analytics Data Needs

• Descriptive analytics typically requires structured, historical data from various sources
  • Data sources include databases, spreadsheets, and transactional systems
  • Data should be accurate, complete, and consistent
  • Techniques used include data aggregation, data mining, and statistical analysis
  • Example: Analyzing sales data from the past year to identify top-selling products and peak sales periods

Predictive Analytics Data Needs

• Predictive analytics requires historical data as well as relevant external data
  • Historical data provides the basis for identifying patterns and trends
  • External data (economic indicators, weather data, social media sentiment) can improve prediction accuracy
  • Techniques used include regression analysis, time series analysis, and machine learning algorithms
  • Example: Using customer data and market trends to predict which customers are most likely to churn

Prescriptive Analytics Data Needs

• Prescriptive analytics requires data from descriptive and predictive analytics, as well as information on business objectives, constraints, and decision variables
  • Data on business objectives defines the goals to be optimized (maximizing profits, minimizing costs)
  • Constraints data outlines the limitations within which decisions must be made (budget, resource availability)
  • Decision variables data represents the factors that can be controlled or influenced (product prices, production quantities)
  • Techniques used include optimization algorithms and simulation models
  • Example: Using sales data, inventory levels, and supplier information to optimize production planning and minimize costs

Choosing the Right Analytics Type

Aligning with Business Objectives

• The choice of analytics type should align with the specific business questions to be answered
  • Descriptive analytics for understanding past performance and identifying trends
  • Predictive analytics for forecasting future outcomes and supporting decision-making
  • Prescriptive analytics for optimizing decision-making and business processes
  • Consider the level of insight and actionability required to meet business objectives

Considering Data Availability and Quality

• The available data and its quality should be considered when choosing an analytics type
  • Descriptive analytics requires accurate and complete historical data
  • Predictive analytics benefits from relevant external data in addition to historical data
  • Prescriptive analytics requires data from descriptive and predictive analytics, as well as business objectives and constraints
  • Assess the availability, accuracy, and completeness of data before selecting an analytics type

Evaluating Organizational Readiness

• The organization's analytical maturity and capabilities should be considered when choosing an analytics type
  • Descriptive analytics is a foundation for more advanced analytics and requires basic data management and reporting capabilities
  • Predictive analytics requires skilled data scientists, robust data infrastructure, and an understanding of statistical modeling
  • Prescriptive analytics demands a high level of analytical maturity, including the ability to implement and monitor recommended actions
  • Assess the organization's current capabilities and invest in necessary resources and skills to support the chosen analytics type