The Multilocal Challenge

How to insure a local visibility on the internet for each shop?

From national to multilocal

Each store is different and has its own local identity: their customers, with their favorite products, their catchment area, etc. National campaigns don’t take these parameters into account and show ads in a homogeneous way across all territories. Some stores will thus randomly be allocated more budgets than others, a pertinent message to the North of the territory will be diffused in the South, and so on.

So the stake is to know how to manage a local media planning: a campaign for each shop. This local media planning is very complex to be executed “manually” because it requires the creation and management of as many campaigns as there are stores in the network. We named this technical problem “multilocal challenge”.

Towards a more adapted and better targeted local digital advertising

Une campagne nationale homogène pour tout le territoire

A homogeneous national campaign for the whole territory

Une campagne spécifique pour chaque magasin

A unique campaign for each shop

ARMIS solved this multilocal problem by developing an artificial intelligence called FLAI (Fast Learning AI)


AI steps up to this challenge by automatically creating campaigns and advertising for each store by optimising investment on each digital platform in order to respect the budget given and that store by store. This results in an optimized and targeted advertising message at the local level, while also ensuring sufficient advertising pressure.

ARMIS has developed Machine Learning algorithms that bring complementary intelligence to digital platforms (Google, Facebook, Xandr, DV360, Waze, etc.) and calibrate these platforms for multilocal contraints.

La puissance de l’ia au service des magasins