The Challenge: Measuring Marketing Effectiveness Across a Crowded Product Portfolio
A major consumer packaged goods (CPG) company was struggling with a problem that plagues nearly every multi-brand organization: how do you measure marketing effectiveness when your portfolio contains multiple products under the same umbrella brand?
The company needed to understand the impact of its marketing activities across 9 Stock Keeping Units (SKUs) that shared brand equity, media budgets, and distribution channels. The challenges were twofold.
First, the analytics team needed to display different model results on one single screen to compare media impact by Stock Keeping Unit. Without this unified view, each SKU existed in analytical isolation, and the team could not see how marketing spend on one product influenced another.
Second, the organization needed to measure media effectiveness at the category level, grouping related products together to understand portfolio-wide dynamics. In CPG, this is critical because consumers rarely think in terms of individual SKUs. They think in categories, and marketing activity at the brand level often creates spillover effects that standard single-product models miss entirely.
This is the media halo effect in action. When a parent brand runs a national TV campaign, the lift is rarely confined to the specific SKU featured in the ad. The awareness and preference generated spill across the entire product line, boosting sales of related items that never appeared in the creative. Without multi-product marketing mix modeling, this halo remains invisible, and marketing ROI is systematically undercounted at the portfolio level.
The Proposed Approach: A One-Stop Modeling Platform for Multi-SKU Analysis
To solve this, the CPG company adopted MassTer Studio, MASS Analytics’ end-to-end marketing analytics software. Unlike traditional MMM tools that force analysts to build each model in a separate project or file, MassTer Studio is designed as a one-stop marketing measurement platform where multiple models for different KPIs can coexist within the same project
Session Cloning: From 9 Hours to Seconds
The team created a dedicated project within MassTer Studio and estimated a separate model for each of the 9 SKUs. The workflow followed five precise steps:
1. Create the foundational session
The team built the first session, applying all necessary data transformation steps, variable categorization, and model specifications for the initial SKU.
2. Replicate across the portfolio
Using MassTer Studio’s session cloning capability, the team replicated the first session’s settings across 8 additional sessions. Rather than manually rebuilding each model from scratch, the cloning feature automatically duplicated the current session’s configuration, including preprocessing rules, variable mappings, and model parameters.
3. Auto-assign dependent variables
During cloning, MassTer Studio automatically set the dependent variables for each new session to match the corresponding SKU’s sales data. This eliminated the manual, error-prone task of reconfiguring target variables for every product.
4. Compare models in a single view
Once all 9 models were built, the team displayed them in a unified reporting screen. This full report contained metrics such as coefficients, T-stats, and contributions at multiple aggregation levels, including periods, regions, and channels.
5. Aggregate for category insights
Because several products fell under the same category, the team used MassTer Studio’s model aggregation capabilities to combine results and measure the media halo effect at the
The time savings were dramatic. In traditional marketing analytics software, creating 9 separate models would require approximately 1 hour per session for variable categorization and data transformation, totaling 9 hours of repetitive manual work. With MassTer Studio’s session cloning, the replication took only a few seconds. The analyst’s time shifted from mechanical replication to strategic interpretation, a shift that compounds with every future portfolio update.
Why Multi-Product Modeling Matters for CPG
CPG marketing is inherently portfolio-driven. A brand manager does not optimize a single SKU in a vacuum. They manage a product line where media spend, trade promotions, and distribution overlap. Multi-product marketing mix modeling acknowledges this reality by allowing organizations to:
- Measure marketing effectiveness at both the SKU and category level
- Detect cross-channel synergy and spillover effects between related products
- Avoid double-counting or undercounting media impact when multiple SKUs share the same advertising
- Build a unified marketing measurement framework that aligns finance, marketing, and sales planning
According to industry research, CPG brands that track portfolio-level metrics like market share, category penetration, and trade spend ROI are better positioned to make informed decisions about resource allocation and promotional strategy . Multi-product marketing mix modeling extends this principle into the statistical domain, ensuring that budget decisions are grounded in holistic rather than siloed evidence.
Results & Insights: From SKU Silos to Portfolio Intelligence
Time Reduction That Scales
The immediate benefit was operational. Session cloning reduced model replication time from 9 hours to seconds. But the strategic value went far deeper.
Media Halo Effect at the Category Level
The aggregated results revealed something no single-SKU model could have shown: the media halo effect was significant. Marketing activities targeted at individual products were generating measurable lift across the broader category. This meant that cutting spend on one SKU would not just hurt that product; it would dampen demand for its siblings.
Understanding the media halo effect is essential for marketing spend optimization in CPG. When a brand runs a masterbrand campaign, the creative may feature a flagship product, but the awareness generated benefits the entire line. If each SKU’s model is evaluated independently, the campaign appears overinvested in the featured product and underinvested in the halo recipients. Only model aggregation at the category level reveals the true portfolio return.
Granular and Aggregate Insights in One Platform
The single-screen comparison allowed stakeholders to toggle between views. They could drill into an individual SKU to see its specific coefficients and contributions, or zoom out to the category level to see aggregated sales decomposition and marketing ROI optimization opportunities.
This dual granularity is rare in marketing analytics software. Most platforms force a choice: either you get SKU-level detail or portfolio-level summaries, but not both in a unified workflow. MassTer Studio’s architecture was built specifically to bridge this gap, giving CPG teams the flexibility to answer tactical questions (How is my new SKU performing?) and strategic questions (Is my category marketing mix optimized?) within the same project.
Key Takeaways for CPG Marketing Leaders
This case study offers several lessons for organizations managing complex product portfolios:
Model at the portfolio level, not just the product level
SKU-level insights are necessary, but insufficient. Without category-level measurement, you will systematically misvalue masterbrand and umbrella campaigns that drive media halo effects across multiple products.
Invest in automation, not just analysis
The 9-hour to seconds improvement came from session cloning, not from hiring a faster analyst. In CPG, where portfolios change constantly with new launches, delists, and pack size changes, automation is the only sustainable path to recurring marketing measurement.
Demand unified reporting
If your marketing analytics software forces you to export 9 different Excel files and stitch them together manually, you are introducing error and delay. A single-screen comparison of multiple models is not a luxury; it is a requirement for portfolio governance.
Build for reuse
The cloned sessions created for this project became a reusable template. When the brand launches its next SKU, the team can clone again, preserving best-practice model architecture while adjusting only the product-specific variables.
Conclusion
For this CPG company, the shift from isolated SKU models to unified multi-product marketing mix modeling transformed how marketing effectiveness was understood. With MassTer Studio, the team could build 9 rigorous models in the time it once took to build one, compare them in a single view, and aggregate results to reveal the media halo effect hiding in their category data.
The message for portfolio marketers is clear: your products do not live in isolation, and your models should not either. Marketing spend optimization at the SKU level alone will leave portfolio value on the table. The brands that win in CPG are those that measure holistically, automate intelligently, and optimize at the category level.