The Challenge
An established online travel agency used a marketing mix model to set quarterly channel budgets across TV, radio, and a large digital programme. While this aggregate approach worked well for macro planning, it operated too slowly for day-to-day digital trading.
With dozens of digital campaigns running across search, social, and display, the trading team struggled to separate genuinely incremental performance from tactics simply coasting on credit earned by others, and within-quarter reallocations routinely fell back on platform reports that the team knew were inflated.
Ultimately, bridging this gap required a more tactical and agile view capable of guiding optimisations in between macro MMM refreshes,keeping unchanged a measurement framework that worked.
The Solution
With digital journeys already well tracked, an initial audit confirmed that clickstream events, campaign taxonomies, and booking records could be linked across the two-to-six-week research window typical of holiday purchases.
We built a data-driven attribution model on those journeys to distribute credit across every digital touchpoint at the campaign level, providing the tactical granularity the MMM could not carry. Calibrating the attribution output against the MMM’s channel totals kept the two views consistent and aligned, with the model apportioning each channel’s MMM-measured contribution across its campaigns.
Each measurement now covers its own ground, with the MMM setting channel quarterly budgets and attribution guiding week on week allocation, giving the digital team a more tactical support to distributes each channel’s budget across campaigns.
The Impact
The campaign-level view revealed that generic search campaigns aimed at early-stage research were earning far more credit than platform reporting had shown, while several retargeting lines had been living on conversions the journey data traced to other campaigns.
Reallocating spend within channels lifted digital efficiency by 12% across the following two quarters, all within budgets already set by the MMM.
The pairing also streamlined the weekly trading meeting by providing a single, consistent reference report to replace conflicting platform reported figures. The integration has now held up through multiple planning cycles, establishing a permanent operational bridge between quarterly planning and weekly execution.
Key Takeaways
- Complementary Measurement - The MMM continued to be used or long-term channel planning while using campaign-level attribution for short-term tactical trading decisions.
- Calibrated, Consistent - Attribution output scaled to the MMM’s channel totals, ensuring the two performance views never competed.
- Tactical Resolution - Campaign-level credit exposed over-funded retargeting and under-funded generic search helping to eliminate waste within healthy channels.
- Commercial Results - Digital efficiency lifted up 12% across two quarters inside unchanged channel budgets, and established a shared, single-report reference for weekly trading teams.
Tools and Techniques
- Multi touch attribution
- Tracking and campaign taxonomy audit - profiling clickstream journeys and validating logging consistency
- Data-driven attribution (Markov chain)- calculating removal effects calibrated to top-down econometric constraints
- Snowflake - data extraction and ingestion of multi-channel clickstream and booking data
- R - data handling and algorithmic attribution modelling
- Tableau - campaign-level credit visualization and deployment reporting
Related services
No matching items