Baseline: The expected outcome under the counterfactual scenario where all treatment variables are set to their baseline values. For paid and organic media, the baseline values are zero. For non-media treatment variables, the baseline value can be set to the observed minimum value of the variable (default), the maximum, or a user-provided float.
Contribution: Each treatment variable's incremental outcome as a percent of expected outcome.
Control variables: Variables in the model that aren't treatment variables. Control variables are used to estimate baseline outcome, and it is not possible to estimate causal effects or contribution percentages for control variables. See Control Variables for important practical advice on selecting controls. Also, see the related concept Mediator variables.
Confounding variables: Variables that have a causal effect on both the treatment and the KPI. Including these as control variables debiases the causal estimates of the treatment on the KPI.
Predictor variables: Variables that have a causal effect on the KPI, but nothing else. Including these as control variables does nothing to debias the causal estimates of the treatment on the KPI. However, strong predictors can reduce the variance of causal estimates.
Cost Per Incremental KPI (CPIK): Total spend divided by total incremental KPI. When the KPI is not revenue and revenue per KPI data is not available, then CPIK equals one over ROI.
Effectiveness: Incremental outcome divided by total media units.
Expected outcome: The expected outcome when all treatment variables are set to actual historical values. This is the sum total of baseline outcome plus the incremental outcome of all treatment variables.
Flighting pattern: The relative distribution of media units across geographic regions and time periods for a given media variable. This is used to allocate media units across geographic regions and time periods when the total budget of a channel is scaled up or down, which applies to budget optimization and response curves.
Incremental outcome: The change in expected outcome driven by each treatment variable. For paid and organic media, this is the change in expected outcome when one variable is set to zero. For non-media treatment variables, this is the change in expected outcome when a variable is set to its baseline value (observed minimum value of the variable (default), the maximum, or a user-provided float) for every geographic region and time period. See Incremental Outcome for details.
KPI: The response (target, dependent) variable of the model. It can be revenue, sales units, conversions, or anything else that the treatment variables may have a causal effect upon.
Lagged effect: A causal effect of treatment variables from previous time periods affecting the outcome in a later time period. Meridian models lagged effects using an adstock function.
Marginal ROI (mROI): The derivative of the response curve and is approximately the ROI on the next monetary unit (such as dollar) spent beyond current spend level.
Media execution: A general term referring to the media unit values of a given channel across geographic regions and time periods.
Mediator variables: Variables that are causally affected by the treatment and have a causal effect on the KPI. Including these as control variables causes a bias in causal estimates of the treatment on the KPI. They should be excluded from the model.
Outcome: The primary metric of interest that Meridian measures the causal effect of treatment variables upon. This is typically revenue, but when the KPI is not revenue and revenue per KPI data is not available, then Meridian defines the outcome to be the KPI itself. It is not necessarily the response variable of the model (see the KPI definition).
Response curve: A plot of incremental outcome versus spend level for a given media variable. As the spend varies, media units are allocated across geographic regions and time periods according to the flighting pattern.
Return on Investment (ROI): Meridian defines ROI as incremental outcome divided by spend. When the KPI is revenue or revenue per KPI data is available, the incremental outcome is incremental revenue. Otherwise, the incremental outcome is incremental KPI.
Revenue: For non-revenue KPIs, this is the revenue per KPI multiplied by the KPI. For revenue KPIs, this is the same as the KPI. When the KPI is not revenue and revenue per KPI data is not available, revenue is undefined.
Revenue per KPI: The assumed revenue generated per KPI unit. This can vary by time, geographic region, or both. Meridian multiplies incremental KPI units by the revenue per KPI to estimate incremental revenue of the treatment variables.
Saturation: Meridian assumes that paid and organic media have diminishing marginal returns, and that there is an asymptotic limit on the media effect over a given time period. As spending increases along the response curve, the mROI diminishes. As the spending becomes large and mROI becomes small, a channel is considered to have become saturated. Saturation is a general term, and no specific threshold is defined.
Treatment variables: Includes all variables for which the MMM estimates a causal effect, namely paid media, organic media, and non-media treatments. The term treatment comes from the field of causal inference and is elsewhere often used synonymously with intervention or exposure.
Paid media variables: Includes all media channels for which spend data is available. This includes both channels modeled with a single variable (for example, spend, impressions, or clicks) and channels modeled with reach and frequency data.
Organic media variables: Includes all media channels that don't have an associated cost, or where the cost is unknown. These channels are modeled with adstock and diminishing returns, just like paid media. The primary difference is that ROI and mROI cannot be measured for organic channels, and so ROI and mROI priors cannot be used for organic channels. This includes both organic media channels modeled with a single variable (for example, spend, impressions, or clicks) and channels modeled with reach and frequency data.
Non-media treatment variables: Includes any non-media tactics such as pricing and promotions. These are variables for which Meridian estimates a causal effect, but the effect size is assumed to be linear rather than having adstock and diminishing returns.