Prescriptive Analytics & Decision Making

Prescriptive Analytics Overview

• Prescriptive analytics utilizes data to determine the best course of action
• This analysis provides recommendations for the next steps, making it valuable for data-driven decision-making
• Machine learning algorithms process large datasets faster and more efficiently than humans
• Algorithms use "if" and "else" statements to analyze data and make recommendations based on requirements

Importance of Human Judgement

• Algorithms offer data-informed recommendations, but cannot replace human judgment
• Prescriptive analytics informs strategies and requires human discernment to provide context and algorithmic outputs

Venture Capital: Investment Decisions

• Investment decisions are strengthened by algorithms that weigh risks and recommend investments
• A Harvard Business Review experiment tested algorithms against angel investors in startup investment picks
• Algorithms outperformed inexperienced angel investors unskilled in controlling cognitive biases

Angel Investors in Venture Capital Decisions

• Experienced angel investors who controlled cognitive biases outperformed algorithms
• Prescriptive analytics plays a complementary role in decision-making by aiding decision-making when experience isn't present
• Algorithms are only as unbiased as their training data, requiring human judgment

Sales: Lead Scoring

• Prescriptive analytics uses lead scoring to rank leads based on their likelihood to convert
• Lead scoring assigns point values to actions taken along the sales funnel

Lead Scoring Actions

• Assign values to:
  • Page views
  • Email interactions
  • Site searches
  • Content engagement including webinars, e-books, and videos
• Assign high point values to actions implying purchase intent, like visiting a product page
• Assign negative points to actions showing non-purchase intent, like viewing job postings

Content Curation: Algorithmic Recommendations

• Social media platforms and dating apps use prescriptive analytics for algorithmic content recommendations
• Algorithms gather data from user engagement history on platforms and potentially other sources
• Algorithm triggers can release specific recommendations based on behavior combinations

Tik Tok "For You" Example

• TikTok’s "For You" feed exemplifies prescriptive analytics
• Website states user interactions weight a user’s level of interest
• TikTok ranks and delivers videos to each user based on the analysis of potential interest
• Prescriptive analytics can increase customer engagement, customer satisfaction, and ad retargeting with ads based on user behavioral history

Banking: Fraud Detection

• Prescriptive analytics is used in banking to detect and flag fraudulent activity algorthmically
• Algorithms analyze and scan new transactional data for anomalies, using customers' historical transaction data

Example anomaly

• Spending is usually $3,000 per month, but suddenly there is a $30,000 charge

Bank Alerts and actions

• Algorithms analyze transactional data, alerts banks, and recommends a course of action
• Recommended action may be to cancel the credit card for potential stealing

Product Management: Development and Improvement

• Prescriptive analytics informs product development and improvements
• Product managers use surveys, beta tests, market research, and behavioral data to gather user data
• Data is analyzed to identify trends, reasons for trends, and predict trend recurrence
• Prescriptive analytics can determine which product features to include/exclude and what changes ensure better user experience

Marketing: Email Automation

• Email automation uses prescriptive analytics
• Marketing sorts leads into categories based on motivations, mindsets, and intentions
• Email content is delivered based on these categories, and lead interactions may shift leads to different categories triggering new messages
• Email automation personalizes messaging at scale, improving lead conversion using relevant content

Leveraging Prescriptive Analytics

• Adopt prescriptive analytics to improve decision-making
• Start with a question or process, gathering related data to analyze with various types of analytics
• The analysis types are:
  • Descriptive
  • Diagnostic
  • Prescriptive
• Use proprietary algorithms, third-party tools, or manual analysis to assess next steps and their impact with company data
• Prescriptive analytics optimizes strategies and helps reach organizational goals.

Data Visualization Importance

• Data visualization translates information into visual contexts
• Examples of visual contexts are maps and graphs
• Data visualization assists the human brain to understand and extract insights
• The goal is to identify patterns, trends, and outliers in datasets

Data Visualization Details

• Data visualization is also known as information graphics, information visualization and statistical graphics
• Data is visualized for conclusions after having been collected, processed, and modeled
• It identifies, locates, manipulates, formats, and delivers data for efficiency
• Executives share information with stakeholders using data visualization