EDION combined causal inference and MMM to assess the effectiveness of direct mail by region and reassess direct-mail investment region by region. The results indicated that reallocating funds to digital advertising could increase store visits in the relevant regions.
Given the characteristics of its nationwide store network, EDION recognised the need to optimise the combination of print direct mail and digital advertising by region. National averages did not sufficiently explain differences in regional responses, and the company faced the challenge of evaluating initiatives separately from external factors.
To address this challenge, the company used causal inference to estimate the underlying contribution of each initiative in each region after removing external effects. It also incorporated Meridian, the marketing-mix modelling system developed by Google, to simulate budget allocation.
The analysis concluded that the effectiveness of direct mail varied by region, with some regions benefiting from increased print distribution and others better suited to a shift to digital. The estimate indicated that revising the allocation while maintaining the current annual budget could increase store visits by 7%.
Going forward, the company plans to continue assessing the reallocation of funds to digital advertising, including YouTube, and repeatedly review the allocation in response to changing conditions. It also plans to expand the visualisation and optimisation of effectiveness by combining purchasing data from stores and e-commerce.