Transform marketing decision-making with next-generation Marketing Mix Modeling

How do you transform marketing decision-making with Marketing Mix Modeling (MMM)? For Ekimetrics, one solution is One.Vision – an advanced platform to optimize performance and secure marketing decisions. This solution is part of a strategy of continuous innovation that recently earned Ekimetrics a ranking among MMM leaders in the Forrester Wave^TM Marketing Measurement and Optimization, Q3 2023.
One.Vision is constantly evolving to adjust to market dynamics. Maximilien Variot, Senior Manager at Ekimetrics, discusses the platform’s new features that are already improving the experience for all its users.

‍

How do you adapt One.Vision to market changes to guarantee optimal marketing performance and meet the expectations of the clients who use it?

Maximilien Variot: The idea is to always adapt to business constraints. We update One.Vision several times a year to adapt the technology and the service to our clients’ business issues. By working in close collaboration with them, we can identify and resolve their pain points more efficiently. This capacity for continuous innovation is essential to offer a more and more fluid user experience. Our technology progresses, stays “alive” and modulable, and allows our clients to stay proactive to market disruptions.
This year, we’ve made three major innovations in One.Vision: deploying certain complex use cases at scale; putting our models into production without disturbing the activity of our internal users and clients; and making continuous UX/UI improvements.

What is the main challenge of deploying large-scale use cases?

M.V.: The American market in particular has guided us: since it’s an immense territory, our clients have to deal with marketing activity that’s highly fragmented, local, and spread across more than 200 data marketing areas. This leads to an explosion of marketing arbitrage and more complex calculations to achieve the desired level of granularity. This particularity slows the delivery of complex use cases (cascade modeling, for example).
To meet this challenge, our team has worked for six months on various technical aspects (algorithmic, functional, data engineering, or IT…) that now allow us to deliver in the platform voluminous hierarchical models, nested model networks, and even both at the same time (nested hierarchical, in our jargon). In parallel, we’ve reinforced the power and flexibility of our cloud resources, which are essential to treat these volumes of data.
We’ve also paid particular attention to improving the granularity of arbitrage, acrucial element when our clients dispose of multiple local activation options.
We’ve integrated these use cases directly into our clients’ environments by accompanying our approach with an adoption program. This allows them to fully exploit One.Vision’s capacities and to optimize their marketing strategies dynamically.
With this nested hierarchical approach, one of our clients has obtained impressive results: 180,000 new business opportunities, 50 million dollars of additional revenue, and a 13% improvement in marketing effectiveness. 

You mentioned another innovation: deploying models without slowing down the client’s environment. Can you tell us more about that?

M.V.: There’s a real before and after thanks to this feature we call “Preview Mode”: Putting our models into production for our clients logically implies regular controls before deployment; previously, this test phase led to slowdowns and disruptions in the client’s environment. Now, with this new deployment module, we can continuously test the results of the models in the background, in “preview” mode, and update them without exposing them to our client, making the process painless. In other words, we’ve simplified complicated business-specific processes to increase efficiency and fluidity. This meets our objective of creating real capacities and continuing to improve the decision-making process.

How have you improved the global user experience?

As I mentioned, we work closely with our internal and external users. We’ve put inplace quarterly and general feedback workshops to improve One.Vision. We take feedback into account to adjust the platform to the reality of business constraints. Our clients report, for example, the need to personalize nomenclature or fiscal periods or to provide different screen resolutions.
To further improve everything UX/UI, we’ve introduced more modularity for consulting results. For example, a country manager can easily consolidate the results of several products in the countries within their perimeter; a product manager can consolidate the results of their product in several countries.
With these constant improvements and the stabilization of our solution’s technological foundation, we can guarantee a solution that is always adaptable to the realities of business. 

What One.Vision upgrades are on the horizon?

The more we advance on One.Vision, the more new subjects arrive! Our short-term priority is to improve the setting of optimization constraints at scale. Simplifying the regular evolution of the taxonomies of our clients’ levers is another major challenge. We've planned another particularly promising project for 2025: integrating an MMM assistant into One.Vision, thanks to generative AI! The idea is kind of to have your MMM consultant in the interface, that can make personalized qualitative or “talk to data” recommendations based on project results, a module that provides encrypted responses, all in real-time. We’ve performed a conclusive Proof of Concept (POC) and have been preparing for its launch for several weeks.
What’s more, with growing demand from our clients for more and more collaborative solutions, we’re developing features like scenario management in public or private groups, for example.
All these innovations are part of our engagement to supply new-generation solutions that help our clients create resilient and effective marketing plans.