Future-proof retail supply chains


Challenge identifier: DPC1-2017


Proposed by

Sonae Center Servicos II, S.A., a Portuguese retail company, with around 40,000 employees and offices in more than 60 countries.


The McKinsey 2014 report ‘Global flows in a digital age’ lists supply chain management as one of the four core factors for the success of the digitisation strategy of any business.  With supply chain data innovation, improved data services, control, and reporting mechanisms can be created, enabling more advanced reporting and planning. Not only does this support key business activities, it also creates better resilience, data governance, and increasing transparency to partners. Supply chain optimization is vital to run a business with lower stock holdings and to minimize the lost sales that are due to stock outs (not having products available in the shelves); to Sonae these are critical business metrics. Supply chains of perishables to multiple locations are particularly sensitive to a variety of internal (transport, storage) and external (weather, tastes) factors.The challenge proposed here is to precisely design the supply chain graphs considering future promotional activity, in order to not only avoid stock outs but also minimize logistic costs.


Sonae will share data about product flows between locations, stored as a single, large denormalized table with the challenge winners. The data is created using Sonae’s operational systems and collected on their premises in a data warehouse. It is made available to third parties via Amazon AWS S3/Redshift. For the purpose of the experiments, the data cannot be stored outside the EU, although locations such as AWS Ireland and Germany are acceptable. More detailed information about the data can be found in our data catalogue.   

Expected outcomes

  • Algorithms and tools for data analysis, reporting, and visualisation, which help Sonae identify areas of improvement within their supply chain.
  • Prediction algorithms that help decrease total stock holdings and lost sales.
  • Supply chain optimisation algorithms, including algorithms that help decrease product waste and lead to better carbon footprints.

Expected impacts

  • Decrease total stock holdings by 10% and stock outs (lost sales) by 20% by better predicting and optimizing the supply chain needs.
  • Achieving these KPIs will have a direct impact of product waste decrease and less CO2 emissions.