Google Analytics is evolving to provide a deeper understanding of your true marketing impact with the introduction of incremental performance metrics and enhanced cross-channel budgeting tools powered by marketing mix modeling (MMM). These tools isolate the “incremental” impact of your marketing efforts. Unlike attribution, which assigns credit to marketing touchpoints that are associated with a conversion, incrementality isolates the conversions, or sales, that were causally driven by those marketing efforts and would not have happened otherwise. These tools can help you make data-driven marketing decisions, rooted in evidence of incremental lift.
There are 2 key methodologies to help you understand the “incremental” impact of ads:
- MMM is a comprehensive, cross-channel tool that captures the impact of your marketing over long time horizons. It accounts for historical factors outside of ads, like seasonality or organic demand, that impact your business outcomes like conversions or sales. As a result, MMM helps identify the true, incremental impact of your marketing investment, making it a useful long-term planning tool.
- Incremental attribution uses the MMM model for the holistic view of incremental performance, while also adding the granular, event-level view from data-driven attribution (DDA). This makes incremental attribution reporting uniquely helpful for frequent, tactical optimization to ensure your campaign strategy is set up to drive growth for your business. You can also use incremental attribution to evaluate new campaigns and determine if they are truly bringing in new customers or revenue.
For a technical deep dive on the underlying methodology, see the Meridian-Based Budgeting and Incremental Attribution in Google Analytics white paper.
On this page
Before you begin
To get started on your modern measurement journey, consider the following requirements:
- Your Google Ads account must be linked to your Google Analytics property.
- Your property must have conversions created. Learn more about creating conversions.
- You must import at least 2 years of campaign cost data. Learn more about importing campaign data.
- No further steps are needed to enable or implement this reporting.
Your property must have at least 2 years of accumulated conversion and revenue data. Conversions are backfilled with event data when available.
About Meridian in Google Analytics 360
Marketing mix modeling (MMM) is a durable, cross-channel measurement solution that helps measure marketing-driven incremental sales (vs expected “baseline” revenue) over a longer time horizon. The model accounts for historical factors outside of ads, like seasonality or organic demand, that impact your business outcomes. MMM doesn’t rely on event-level attribution, and instead looks at aggregate trends, enabling you to see a holistic picture of your marketing efforts and compare impact across channels in an apples-to-apples way. This makes MMM an actionable long-term planning tool, empowering you to make data-driven budgeting decisions based on the true, or incremental impact of your marketing.
The MMM solution in Google Analytics 360 leverages Meridian, an open-source Bayesian MMM framework built by Google. Meridian is grounded in causal inference theory and uses Bayesian regression modeling to measure the impact of historical marketing activity. It combines prior knowledge of your aggregate conversions and sales data with patterns learned from campaign data to estimate media effects. Meridian in Google Analytics 360 uses informed incrementality estimates, grounding the model’s assumptions in proven conversion lift experimentation and industry benchmarks. Meridian’s optimization engine explores different “what if” scenarios to understand the impact to your conversions or revenue at different budgets. This allows Meridian-based budgeting tools in Google Analytics to pinpoint the best recommended budget allocation to maximize your incremental ROI.
Model inputs
- Historical data: Leverages 2+ years of conversion, revenue, cost, and aggregate impression data across Google and third-party platforms.
- Incrementality estimates (Priors): Utilizes Google’s Smart Defaults, which are informed calibration signals calculated blending lift studies and third-party benchmarks data.
- Control variables: Incorporates control variables like Google Query Volume (GQV), which is organic search volume derived from your brand name. This allows the model to effectively account for natural brand intent and better isolate the effects of paid marketing efforts.
Historical data and lookback window
Historical data refers to the total timeframe of aggregated data used to build and calibrate the model. A 2-year minimum is set so the model has an extensive amount of data to establish a baseline of "business as usual" and accurately isolate variables, as well as account for seasonality and economic shifts. It also helps the model calculate an accurate view of baseline sales: the revenue you would generate even if you turned off all advertising today.
Lookback windows in the context of MMM are different from the historical data input. In this context, an MMM lookback window, or adstock window, measures the "memory" of your marketing. This is the maximum number of days or weeks a single ad exposure continues to drive incremental sales after it runs. Media rarely has a purely immediate impact. While a consumer might click a search ad and buy instantly, they might see a video ad and wait 3 weeks before converting. The lookback window defines the timeframe the model uses to capture that delayed response, known as adstock or carryover effect. With Meridian in Google Analytics 360, this window is set to 16 weeks to best account for extended customer journeys and high-consideration purchases.
Incrementality estimates
To calibrate our MMM within Google Analytics Modern Measurement, we rely on incrementality estimates. These signals are grounded in trusted measurement data, drawing from Google Conversion Lift experiments, aggregated by channel, conversion type, and advertiser vertical, and established third-party benchmarks.
Incrementality estimates (effectively, priors) serve as educated initial baselines for the model. Rather than evaluating a media channel's effectiveness from an entirely blank slate, priors supply the model with an informed starting point regarding that channel's expected return on investment (ROI).
These baselines are calculated using established performance metrics, specifically, incrementality data, which isolates the conversions driven directly by advertising from those that would have occurred organically. All available incrementality data for a specific advertiser is consolidated into a single, calibrated ROI estimate to guide the model toward more accurate outcomes.
Model quality checks
We rigorously test the underlying Meridian models to ensure you receive a stable, deterministic solution that learned from your historical data. Before appearing in your property, a model must pass a series of health tests, which are broadly aligned with the open-source Meridian health checks. The model’s predictions are validated to make sure that these are grounded in reality, confirming that baseline sales estimates make logical business sense, so the model accurately isolates your true media impact without overestimating or underestimating the results. These health checks help ensure you’re only making decisions based on accurate, high-quality models.
In some cases, even if your property meets the initial data eligibility requirements, a model might not be available for your property. This typically occurs because the underlying data lacks the necessary signals to confidently measure incrementality, or because the model's outputs failed our strict quality assurance tests.
When a model doesn’t pass, it’s generally due to one of 2 core reasons:
- Data limitations: The model needs a rich, diverse dataset to learn accurately. It will fail if there is insufficient historical history, missing cost and impression data from third-party platforms, or a lack of cross-channel diversity in your media mix.
- Model convergence: Model convergence refers to the model being able to detect consistent and repeatable patterns in the data, which increases the confidence level that those patterns are accurate. The model tries to find a stable set of parameter estimates that explain the relationship between estimated channel contributions (marketing spend) and ROI (business outcomes). Models aren’t able to detect consistent patterns when they have incomplete data, missing data, or sales data that doesn’t match the conversion patterns.
- Misalignment with reality: A model will fail if it generates impossible scenarios, like negative baseline sales. This is a sign that there is a statistical error or misspecification in your model that needs to be addressed before it should be used for decision making.
- Model fit: The model’s predicted outputs significantly differed from historical, observed data for the property.
Powered surfaces
Meridian-powered insights are available throughout the Advertising section in Google Analytics 360. Specifically, in these surfaces:
- Advertising snapshot: Review channel performance based via MMM incremental metrics (iROAS or iCPA) from the past 4 weeks.
- Scenario planner: Optimize your future cross-channel budgets using Meridian-powered insights.
- Projections report: Monitor in-flight performance against your goals and optimize your tactical strategy based on projected outcomes.
- Analytics Advisor: Run Meridian-based budget optimizations directly within Analytics Advisor.
About incremental attribution
The incremental attribution model measures the performance of your marketing efforts across all channels by isolating the impact of your ads. It identifies campaigns that genuinely drive growth rather than just capturing existing demand. While standard attribution models, such as last-click or data-driven attribution (DDA), assign credit to touchpoints along a conversion path, incremental attribution focuses on causality. It identifies the "lift", the additional conversions your ads generated that would not have occurred without the ad.
To achieve this, we leverage the Meridian model. Calibrated with real-world ROI anchors, such as Conversion Lift data and third-party benchmarks, your imported cost and performance data will be used to generate powerful incrementality outcomes that represent the true ROI across your channels. By matching these MMM incrementality outputs directly to your DDA performance, we can calculate highly granular, incremental attribution results for every channel.
What is the difference between "incremental" and "attributed" conversions?
Attributed conversions measure correlation, while incremental conversions measure causation. Attribution assigns credit to marketing touchpoints that are associated with a conversion, whereas incrementality isolates the conversions that were causally driven by those marketing efforts and would not have happened otherwise.
How incremental attribution works
- Meridian MMM output: Provides periodic estimates of incremental conversions at a highly defined granularity.
- Standard DDA model output: Attributes raw credit to individual touchpoints, such as clicks and impressions, along user conversion paths.
- Comparison engine: Aggregates the DDA-assigned credits up to the exact channel and time granularity of the MMM output, then compares the totals.
- Lookback window: Aligns the attribution window precisely with the MMM’s 16-week lag window so that causal lift and touchpoint credits are measured across the same timeframe.
Powered surfaces
Incremental attribution metrics are available throughout the Advertising section of Google Analytics. Specifically, in these surfaces:
- Advertising snapshot: Review a summary of performance between MMM, DDA, and Incremental Conversions
- Incremental conversions report: Analyze granular, channel-level performance to identify the true causal lift of your campaigns and make tactical budget optimizations.
- Analytics Advisor: Dive deeper into incremental conversions reporting with Analytics Advisor.
How to use incremental performance metrics
Measuring incremental performance suits scenarios needing true causal impact. It focuses on specific marketing activity, not just correlation with conversions.
- Optimize spending: Identify the effective spending levels for existing campaigns by understanding their diminishing returns or increasing efficiency.
- Compare marketing initiatives: Measure the net value each program contributes to allocate resources more effectively.
- Data-driven strategy: Shift marketing budgets and strategies based on empirical evidence of incremental lift.
Learn more About the incremental conversions report.