strataQED is a professional Marketing Mix Modeling (MMM) platform developed by ScanmarQED. It is designed to help organizations measure, understand, and improve the return on marketing and promotional investments.

The software enables you to build statistical models that explain what is driving your sales or other key business metrics, quantify the contribution and ROI of each marketing activity, and then use that knowledge to forecast and optimize future plans.

strataQED is built for marketing analysts, data scientists, and commercial planners at consumer goods companies, retailers, and any organization that needs to understand and justify its marketing spend. It is particularly well suited to teams that work with sales data across multiple geographies, regions, or market segments and need to move beyond high-level national averages to granular, actionable insights.

strataQED is built around answering the questions that matter most to marketing and commercial teams:

• What is the overall ROI of our advertising and promotions activities?
• Which channels or regions deliver the strongest return?
• Where are we seeing diminishing returns, and where should we spend more?
• What sales can we expect if we run a specific marketing plan next year?
• How should we redistribute our budget to maximize return for the same total spend?
• What is the right total budget to achieve our sales target?
• How much of our TV or digital impact is flowing through indirect channels like search?

strataQED is structured as a modular platform. The core platform covers data management, model building, and reporting. Optional licensed modules extend the platform further:

• Core platform — data import & repository, automated and manual model search, response curves, contributions, ROI and period comparison reporting.
• Nested Modeling module — multi-KPI modeling to capture direct and indirect marketing effects across linked models (e.g. sales, search).
• Forecasting & Optimization module — plan creation, future performance simulation, plan comparison, fixed-budget optimization, and budget sizing.

strataQED works with two main data structures:

• Time Series — a single stream of data over time, such as national weekly sales.
• Time Series Cross-Sectional — data that spans multiple cross sections simultaneously, such as sales by region, market, or store across time.

Typical inputs include weekly or monthly sales volumes, advertising spend by channel, promotional calendars, price and distribution data, and competitive activity. The software can handle national, regional, and granular market-level datasets within the same project.

Yes. strataQED is designed to work across multiple levels of geographic granularity within a single project. You can import national, regional, and market-level data files into the same repository and build models at each level. The software supports user-defined hierarchies, making it easy to aggregate or disaggregate data as needed and to compare results — for example, a national model versus a market-level model — to understand where performance differs.

strataQED offers two complementary approaches to model building:

• Automated Model Search — a genetic algorithm searches through thousands of model combinations per minute, testing different variable selections, response curve transformations, carryover (Adstock) rates, and lag lengths. The engine ranks each candidate model against a configurable Model Rating Template.
• Manual Model Building — an interactive model builder with a built-in Manual Model Assistant that gives live traffic-light feedback on statistical significance and variable direction as you add or remove each variable.

Both approaches can be used in combination — for example, starting with an automated search and then refining the best candidate manually.

strataQED supports both Ordinary Least Squares (OLS) and Ridge regression. Ridge regression is particularly useful when datasets contain correlated predictor variables, as the penalty parameter reduces variance at the cost of a small degree of bias.

Yes. strataQED supports both fixed effects and varying effects cross-sectional modeling, allowing a single model to be built across dozens of markets or regions simultaneously. This pooled approach uses more data than individual market models, improving statistical reliability, while still capturing how marketing response rates differ between markets. You can also use a hold-out test sample to validate model robustness before finalizing.

Once a model is built, strataQED provides seven core reporting screens:

• Response Curves — visualize the diminishing-returns relationship between spend level and incremental sales, including long-term and profit curves.
• Brand Painting — a stacked period-by-period chart showing the week-by-week contribution of each driver to the dependent variable.
• Contributions — a summary bar or waterfall chart showing each driver's total contribution over a selected date range.
• Return On Investment (ROI) — displays incremental revenue, cost, and ROI for each marketing activity, with the option to include long-term carryover effects.
• Period Comparison — compare driver contributions between two time periods to understand how marketing effectiveness has changed year-on-year.
• Model Definition — the full statistical specification and key diagnostics.
• Advanced Model Statistics — detailed coefficient-level statistics and the model equation.

All reporting screens can be filtered by cross section (e.g. by region or market) and grouped by series, category, or category type.

Standard Marketing Mix Models measure the direct effect of each marketing input on a single KPI such as sales. However, marketing often works through intermediate KPIs — for example, TV advertising drives search and search then drives sales. Standard models may miss or understate this chain of effects.

strataQED's Nested Modeling module links multiple separate models together and reattributes the indirect effects back to their original driver. The result is a complete picture of the true total ROI of each activity, including both its direct and indirect contributions.

Yes. The Nested Modeling module includes a Combine Models feature that lets you aggregate two or more models — for example, Brand A sales and Brand B sales — into a single combined total sales model. This is useful for portfolio-level analysis and for reporting total marketing effectiveness across a brand portfolio.

The Forecasting module allows you to take a completed model and use it to simulate the future. You combine a model (the historical statistical relationships) with a plan (your proposed marketing inputs for a future period) to generate a forecast of the dependent variable — typically sales or revenue. The module helps you to:

• Project the sales impact of a proposed marketing plan before committing budget.
• Test multiple alternative plans and compare their expected outcomes side by side.
• Set a sales target and track the percentage of that target your plan is forecast to achieve.
• Review forecast contributions, ROI, and response curves for the plan period.

The Optimization module uses your model's response curve knowledge to algorithmically redistribute your marketing budget across channels and time periods to maximize the forecast outcome. Two optimization modes are available:

• Fixed Budget Optimization — given a fixed total spend, find the channel mix and timing that maximizes your KPI (typically sales or gross profit).
• Budget Sizing — identify the optimal total budget level by modeling the marginal return on additional investment, helping you build the business case for a budget increase or reduction.

Yes — strataQED is designed to produce optimizations that are commercially realistic, not just mathematically optimal.

Yes. strataQED is a Windows desktop application that runs locally on your machine. Tutorial example files, including sample datasets and pre-built model files, are installed alongside the software and accessible via the Start Menu, giving you a ready-to-use environment to explore the software's capabilities before working on your own data.

strataQED is built to guide analysts through the modeling process with structured wizards, step-by-step workflows, and a built-in Help file. The automated model search significantly lowers the barrier to building an initial model. The Manual Model Assistant provides live statistical feedback as variables are added, making it suitable for both experienced econometricians and analysts newer to marketing mix modeling. Of course, the ScanmarQED team is always on hand in case you require further assistance.

To learn more about strataQED, speak to the ScanmarQED team directly. They can walk you through a live demonstration using your own data type and business questions, and advise on which modules would be most relevant for your organization's needs.