How G-Star Measures Marketing ROI Across 14 Global Markets

The Challenge

G-Star is a major global fashion brand generating revenue through both retail stores and e-commerce. With marketing campaigns running across 14 markets and a wide mix of media channels and product categories, the organization faced a fundamental measurement problem.

Traditional attribution and last-click models couldn’t distinguish the incremental impact of media from baseline demand. Worse, they offered no way to separate short-term performance effects (direct response, promotions) from long-term brand growth driven by upper-funnel communication.

With media channels constantly evolving, G-Star needed high-frequency, real-time ROI insights, not quarterly reports that arrived too late to act on.

The Solution

Alviss AI built granular Marketing Mix Models across all 14 global markets. The implementation followed three key phases:

• Data structuring: Media spend, sales, and contextual data were organized at the right level of granularity relative to channel and product category splits
• Stakeholder alignment: Deep alignment sessions ensured shared understanding of business drivers, marketing channels, media ROIs, and the role of brand effects
• Automated pipelines: Data pipelines were connected directly into the Alviss platform, enabling frequent, automated ROI updates rather than one-off analyses

The Bayesian modeling approach provided confidence intervals on every ROI estimate, letting the team distinguish high-confidence findings from areas of genuine uncertainty.

The Results

• ROI estimates across all media channels in 14 markets, updated automatically as new data arrives
• Clear separation of short-term performance effects and long-term brand contribution for each channel
• Data-driven reallocation of spend across markets, channels, and product groups
• Improved balance between tactical direct-response campaigns and brand-building communication