How to Use Cassandra for Forecasting

Forecasting in Cassandra enables you to predict future performance based on historical trends, simulate budget changes, and interpret confidence intervals to make data-driven marketing decisions.

This guide walks you through setting up a forecasting model, running budget simulations, and analyzing projection limits.

1. Setting Up a Forecasting Model Using Past Trends

Why Forecasting Matters

• Predicts future sales, conversions, or revenue based on historical performance.
• Helps plan budget allocations and optimize marketing investments.
• Identifies potential seasonal trends and market fluctuations.

Steps to Build a Forecasting Model

1. Ensure High-Quality Historical Data
   • Minimum 12 months of historical data is recommended.
   • Include marketing spend, conversions, sales, seasonality markers, and external factors.
2. Select Key Performance Indicators (KPIs)
   • Revenue, Return on Ad Spend (ROAS), Customer Acquisition Cost (CAC), or any primary KPI.
3. Define the Forecasting Window
   • Cassandra allows projections from weeks to months ahead based on data history.
   • Select an appropriate time horizon for short-term (4-8 weeks) or long-term (3-12 months) predictions.
4. Run Forecasting Model
   • Navigate to Cassandra’s Forecasting Dashboard.
   • Select the dataset and time window.
   • Review initial forecast output and refine as needed.

2. Simulating Budget Changes and Predicting Future Performance

What is Budget Simulation?

• Simulates various marketing spend scenarios to predict future performance.
• Helps identify optimal budget allocations to maximize ROI.
• Evaluates the impact of increasing or decreasing spend per channel.

Steps to Run Budget Simulations

1. Access Cassandra’s Budget Simulation Tool
   • Choose Optimize Budget or Scenario Planning mode.
2. Set the Budget Parameters
   • Adjust total spend and allocation per channel.
   • Define constraints (e.g., minimum/maximum spend per channel).
3. Run and Analyze Simulations
   • Cassandra generates expected revenue, conversions, and ROI based on inputs.
   • Compare multiple budget scenarios to identify the most efficient spend strategy.
4. Apply Learnings to Real-World Budgeting
   • Use insights to guide quarterly planning, ad spend shifts, and investment strategies.
   • Iterate based on ongoing market performance and business goals.

3. Interpreting Confidence Intervals and Projection Limits

What Are Confidence Intervals?

• Measure the uncertainty in forecasted results.
• Provide a range within which the actual outcome is likely to fall.
• Wider confidence intervals indicate greater uncertainty, while narrower ones signal higher confidence in projections.

How to Interpret Forecast Ranges

• High Certainty Scenario: Narrow confidence intervals suggest stable predictions.
• Uncertain Scenario: Wide intervals mean external factors or data limitations affect accuracy.
• Risk Assessment: If intervals are too wide, consider gathering more historical data or refining model assumptions.

Handling Projection Limits

• Short-Term Forecasting (1-3 months): Generally reliable with smaller confidence intervals.
• Long-Term Forecasting (6+ months): Increased uncertainty; results should be used for directional insights rather than precise planning.
• External Factors Impact: Economic shifts, competitive changes, or unexpected events can influence forecasts beyond historical trends.

4. Summary & Next Steps

• Ensure clean historical data before running forecasts.
• Use budget simulations to predict performance under different investment strategies.
• Interpret confidence intervals to assess forecast reliability.
• Refine model inputs based on new insights and external factors.
• Apply learnings to real-world decision-making for optimal budget allocation.