MMM v2.0 Now Available

We’re excited to announce the release of Media Mix Modeling (MMM) 2.0, our biggest update to the tool since its initial launch.

While version 1.1 focused primarily on improving the platform’s infrastructure and reporting capabilities, version 2.0 introduces some significant enhancements to the underlying modeling process, expands support beyond digital media, and introduces goal-based budget optimization.

With the addition of traditional media, more accurate projections, and the ability to automatically calculate recommended budgets based on specific conversion goals, MMM is becoming an even more comprehensive tool for media planning and optimization.

Here’s what’s new.

Introducing Traditional Media Support

One of the biggest additions in 2.0 is the ability to include traditional media channels alongside digital media in the model.

Previously, MMM was primarily focused on digital channels such as Paid Search, Paid Social, and Display. With this update, we can now incorporate channels such as Broadcast, Print, and Out-of-Home (OOH) into the same model, allowing users to get a more comprehensive view of their overall media investment.

We’ve also introduced support for blended media metrics, meaning that channel activity can now be measured using a weighted combination of metrics, such as impressions and clicks, rather than relying on a single measurement.

These changes allow us to accommodate a wider range of media strategies and build models that better reflect how our clients are actually investing their marketing dollars.

More Accurate and Reliable Projections

We’ve made some substantial improvements to the underlying modeling process to produce more accurate and consistent projections.

One important change is how we handle historical and projected conversions.

Previously, conversions associated with non-paid sources such as Organic, Direct, or Email could be excluded from certain reporting views depending on how the underlying data was structured.

With 2.0, Historical conversions now include all recorded conversions, regardless of whether they were directly associated with a paid media channel.

Projected conversions also account for these non-paid conversions, holding their contribution constant while modeling the incremental impact of changes to paid media investment.

This means that when a user first opens a project with the default budget allocation, Historical and Projected conversions will match. As they begin adjusting budgets and channel allocations, the projected results will change based on the model’s estimated media impact.

We’ve also improved how projections are calculated by aligning them with the same time period as the selected budget and calibrating the model against actual historical performance.

Additionally, we’ve introduced an automatic fallback process that allows models to move from daily to weekly or monthly data when a stable result can’t be achieved at the more granular level.

Combined, these updates make the model more reliable and the resulting projections easier to interpret.

Set Conversion Targets and Automatically Optimize Your Budget

Another major addition in 2.0 is the ability to work backward from a specific conversion goal.

Users can now enter a Target Conversions value directly in the sidebar, and MMM will automatically calculate the recommended total budget and channel allocation needed to achieve that goal.

Rather than only asking what might happen if a budget increases or decreases by a certain amount, users can now approach planning from the opposite direction: If we need to generate a certain number of conversions, how much should we spend, and where should we allocate that budget?

This introduces a new layer of functionality to the tool, allowing users to move beyond scenario exploration and generate model-driven budget recommendations based on specific performance objectives.

We’ve also added a new “Show Media Mix As” toggle that allows users to switch between viewing channel allocations as percentages of the overall budget or as actual dollar amounts.

The overall budget slider has been refined as well, making it easier to move through the modeled budget range in whole-percentage increments.

Together, these enhancements provide more flexibility when building scenarios and give agency teams a more direct way to translate client performance goals into media investment recommendations.

Improved Location and Segment Reporting

For projects that include multiple locations or segments, we’ve updated the reporting logic to ensure that each individual page displays the appropriate data.

Previously, some location-level views could display historical or projected information associated with the overall account rather than the selected location.

With 2.0, each location or segment page now displays its own historical performance and corresponding projections.

This should make it easier to evaluate performance and potential investment opportunities at a more granular level without confusing individual location results with those of the overall account.

Better Client Access and Project Management

We’ve also made some important changes to how users access and manage their MMM projects.

Each client now has a dedicated administration page where authorized email addresses can be added and managed. Users granted access through these pages will only be able to view the projects associated with that specific client.

On the administrative side, we’ve introduced more granular controls over the model execution process.

Administrators can now separately initiate data ingestion, model optimization, or both processes together. Each step includes live status updates, additional detail panels, and the ability to retry individual failed processes without restarting everything.

These updates should make project management and troubleshooting significantly more straightforward as we continue onboarding additional clients.

Additional Improvements and Bug Fixes

Beyond the major changes outlined above, we’ve made several smaller improvements throughout the application:

• Faster optimization: Improvements to the processing workflow have reduced optimization run times.
• Cleaner reporting UI: Historical and Projected spend allocation charts now appear side by side, with redundant conversion pie charts removed.
• More consistent reporting totals: Monthly and 12-month tables now reconcile with the Historical and Projected totals, including when a reporting period begins in the middle of a month.
• More reliable file uploads: Resolved an issue that caused CSV uploads larger than 1 MB to fail.
• Clearer project statuses: Projects now correctly display “Ingesting” while data is being loaded, rather than incorrectly showing “Optimizing.”
• Enhanced access controls: Additional improvements to authentication and client-level permissions.

What’s Next?

MMM 2.0 represents an important step toward making the platform a more comprehensive media planning and optimization tool.

With the addition of traditional media, improvements to the underlying modeling process, and new ways to automatically optimize budgets against specific conversion targets, we’re continuing to expand the types of questions agencies can answer using MMM.

The goal remains the same: give our teams a way to move beyond historical reporting and use data to make more informed decisions about future media investments.

As always, we’ll continue refining the platform based on feedback from our agency teams and clients.

If you have questions, feedback, or a client you think could benefit from MMM, please reach out to the RADaR Analytics team.