EUROFOOD, a growing company offering a large range of products 

Eurofood has 90 employees and generates 15 million euros in revenue. The company imports products from Europe and Asia and distributes them exclusively on the Thai market. It manages about 600 products from brands such as Beyond Meat, Bonne Maman, Andros, Paysan Breton, and Tramier. Eurofood sells these products through two channels: Retail (food stores) and Food Service (hotels and restaurants).

Long planning horizons and strong seasonality means necessity for reliable forecasts.

Most Eurofood products come from Europe, which creates long lead times. “Products are available for sale three months after the order,” explains Etienne Devictor, Demand Manager at Eurofood.

One order usually covers two months of stock, so the overall forecasting horizon reaches five months. This long cycle increases inertia and makes stock replenishment harder to manage. Eurofood distributes products with shelf lives ranging from 25 days to three years, so forecast quality is essential for efficient inventory management.

This long horizon also reduces forecast reliability. Sales representatives, who were responsible for the forecasts at the time, struggled to plan five months ahead.

Seasonality adds more complexity, especially on the Food Service channel, which depends heavily on tourism flows. “We noticed that even with our monitoring tools, we reacted too late,” Etienne says. A structural bias made forecasts too optimistic. This bias led to high stock levels and created significant financial pressure. With recurring overstocks, the company often had to run sales operations to avoid product expirations. “We decided to change our process and adopt an appropriate tool,” he adds.

A change of process and the centralization of forecasts 

Until then, the demand plan was an aggregation of forecasts made on Excel by the sales representative. Due to their activity and optimism, the actors of the process had a tendency to over-plan and preferred to have more stocks to be sure to avoid shortages.

« We decided to change the roles and appoint one central person in charge of the demand generation to address the entire market” explains Etienne.

« Moreover, the data management was complicated due to Excel, so we decided to implement a forecasting tool more reliable in order to better size our stock levels and reduce the associated financial costs” continues Etienne.

The sales representatives have now a role of contributor instead of generator. For many products the statistical forecast is enough, allowing to reallocate efforts on what matters.

The choice of Colibri

Among all the solutions available on the market, Eurofood chose Colibri for the following reasons: 

    • The SaaS mode of the solution and the short engagement period implying a win-win relation, and an optimal service for the entire contract duration;
    • The simple process and the pre-packaged solution for a quick set up and ROI.
    • The upgradability of the solution, “today we are using the VISION module for the forecasts, but the possibility to evolve later to a more global solution by adding new modules such as FLOW for the replenishments is very interesting” explains Etienne.
    • The cost of Colibri, competitive, which of course participate in the quick ROI.
    • The localization of the teams: “to have a team in Singapore with almost no time difference from us is important, we feel supported”.
    • The possibility to have a prototype: “We were able to use the solution with our data before taking any decision, it allowed us to project ourselves in the solution and to convinced us in our choice”.

Encouraging first results

After 2 months of project, Eurofood is now live on the solution since the beginning of February. The prototype was used for 2 to 3 months before that, and the first results are encouraging.

« We recorded that Colibri allows a greater forecasts reliability, and this without even using corrected historics » states Etienne 

For example, Etienne recorded: +11 points for the class A products, +12 points for class B and +26 points for class C on the forecasts reliability in values on all products.

« The prototype allowed us a reliability at 63%, our objective is to continue to work on the forecasts quality to reach 76% in the upcoming months. Colibri will also allow a better cross-functional activity between the teams and improve and ease the decision making » concludes Etienne Devictor.