"Algorithmic knowledge and its flexible methodology have been key to creating a tool tailored to a sector as specific as ours."
FEATURES
Lobster sales forecast for the week ahead.
IMPACT
Reduction in waste by 25%, increase in sales by 1%, 50% reduction in staff dedication to the process.
ACHIEVEMENTS
Automatic prediction that improves the previous manual model by 5 points. Automated, protocolized and documented process.
BUSINESS PROBLEM
Maresmar is one of the main wholesalers of sea bass. One of its star products is fresh lobster, imported mainly from Canada. Lobster, as a perishable fresh product, requires high precision in predicting demand. In this case, due to the complex intercontinental air logistics, we needed predictions a week in advance in order to place orders in sufficient time. A prediction that was too ambitious resulted in losses with significant shrinkage. At the same time, an order that was too conservative could result in the loss of potential sales. Demand in this case also has a strong dependence on logistical problems in the sector as well as on the price of the competition. The previous method was completely manual and dependent on a few actors who had a lot of accumulated, but undocumented, knowledge, making the whole process not very robust and fragile to possible staff turnover.
DRIBIA’S SOLUTION
Together with Maresmar, we have built a custom sales prediction algorithm that, based on sales history, generates a daily sales prediction one week in advance. This prediction is distributed for each day of the week and for each sales warehouse.
At the same time, we have built a prediction review platform, which processes all the data to facilitate the consultation of sales history and so that the analyst can consult and modify the model’s prediction according to expert business knowledge. At the same time, the analyst can add incidence and reasons to the modifications that will be consultable in the future in order to make the sales history more informative. Finally, the platform calculates weekly performance indicators of the tool and the team enabling continuous improvement.
The platform is accompanied by documentation of the entire process that has allowed the internalization of knowledge not only in the current sales analyst, making Maresmar much more resilient to possible staff turnover.
