Getting started with Cassandra
      Back to home
      1. Knowledge base
      2. Getting started with Cassandra

      Glossary of Key Terms

      This glossary provides definitions for key terms used within Cassandra’s Marketing Mix Modeling (MMM) platform. Use this guide to better understand model outputs, budget optimization, and data integration.

      A

      • Adstock Effect – The lasting impact of advertising campaigns, where exposure continues to influence consumer behavior over time.
      • Algorithmic Attribution – A data-driven approach to assigning credit to marketing channels based on statistical modeling.
      • Attribution Model – A framework for distributing credit across touchpoints in the customer journey.
      • Average Error on Prediction – A measure of how far off the model’s predictions are from actual values, indicating forecast accuracy.

      B

      • Baseline – Sales that would occur even without marketing spend, influenced by external factors such as seasonality and economic conditions.
      • Bayesian Modeling – A statistical method that incorporates prior knowledge and updates probabilities based on new data.
      • Budget Allocator – A feature in Cassandra that optimizes media spend based on historical performance and diminishing returns.
      • Business Error – When a model’s predictions do not align with real-world expectations, such as underestimating a highly impactful channel.

      C

      • Calibration – Adjusting model parameters to improve alignment with observed data and real-world performance.
      • Channel Contribution – The percentage of total revenue attributed to a specific marketing channel.
      • Confidence Interval – A statistical range estimating the likely impact of a marketing channel, indicating the certainty of attributions.
      • Conversion Lift Test – An experiment measuring the true impact of a marketing campaign by comparing exposed vs. non-exposed audiences.
      • CPA (Cost per Acquisition) – The cost incurred to acquire a new customer through marketing spend.
      • CPM (Cost per Mille) – The cost per 1,000 impressions in advertising.
      • CTR (Click-Through Rate) – The percentage of users who click on an ad after viewing it.

      D

      • Data Integration – The process of importing and structuring marketing data into Cassandra.
      • Dataset – The collection of structured data, including marketing spend, conversions, and external factors, used to analyze performance.
      • Diminishing Returns – The principle that increasing investment in a marketing channel eventually leads to reduced incremental gains.

      E

      • Elasticity – A measure of how changes in marketing spend impact conversions or revenue.
      • Experimentation – Testing marketing hypotheses through GeoLift or Conversion Lift studies.

      G

      • GeoLift Experiment – A test that measures the effectiveness of marketing spend by comparing regions where advertising was paused versus active.
      • Granularity – The level of detail in a dataset, such as daily vs. weekly spend data.

      I

      • Incrementality – The additional revenue generated by a marketing channel beyond what would have happened organically.
      • Impressions – The number of times an ad is displayed to users.

      M

      • MMM (Marketing Mix Modeling) – A statistical approach to measuring and optimizing the impact of marketing investments.
      • Model Refresh – Updating a Cassandra model with new data to reflect the latest business trends.
      • Multicollinearity – When multiple marketing variables are highly correlated, making it difficult to isolate their individual impact.
      • Media Spending – The total investment allocated to advertising and marketing efforts across different channels.

      O

      • Optimization – Finding the best way to allocate marketing budget for maximum efficiency and ROI.
      • Organic Growth – Revenue or customer acquisition that occurs without paid marketing.
      • Organic Variables – Factors influencing sales that are not directly tied to paid marketing, such as word of mouth or brand reputation.
      • Output Variable – The key performance metric (e.g., revenue, conversions) that a model predicts based on marketing inputs.

      P

      • Paid Attribution – Revenue attributed to paid marketing activities.
      • Platform Discrepancies – Differences between Cassandra’s model outputs and platform-reported metrics due to different attribution methodologies.
      • Predictive Analytics – Using historical data to forecast future performance.

      R

      • ROAS (Return on Ad Spend) – Revenue generated per dollar spent on advertising.
      • ROI (Return on Investment) – The profitability of marketing activities, calculated as revenue divided by marketing spend.

      S

      • Saturation – The point at which additional marketing spend no longer drives incremental sales growth.
      • Saturation Curve – A graphical representation of how marketing spend affects returns at different levels.
      • Seasonality – Recurring fluctuations in demand based on time of year (e.g., holiday sales trends).
      • Sales Decomposition – Breaking down total sales into different components, such as baseline sales, paid media impact, and seasonality effects.

      T

      • Touchpoints – The various interactions a customer has with marketing before conversion.
      • Time Lag Effect – The delay between marketing exposure and resulting conversions.

      U

      • Uncertainty Measurement – A calculation in Cassandra that indicates how confident the model is in its attribution estimates.
      • User Journey – The path a user takes from awareness to conversion across different marketing channels.

      Summary & Next Steps

      • Use this glossary to understand Cassandra’s key terms and improve your data analysis.
      • If you need further clarification, visit Cassandra’s Help Center

      This glossary will be continuously updated as new features and methodologies are introduced in Cassandra.