Data Scientist

Job Description Purpose of the role – mission The Data Scientist role is responsible for designing and leading universe studies, setting up the extrapolation for Retail Panels, discussing and designing rawdata modelling for non-coop retailers, and supporting root-cause analysis in case of client’s complaints when related to these activities. This role requires strong data fluency, analytical skills with the ability to assess data relevance, and a deep understanding of retail panels and their value. Key activities: 1. Universe Studies: • Design and lead universe studies, acting as a “project manager” for these studies. • Document methodologies for transparency and provide clear explanations of selected methods. • Validate the universe estimation. • Conduct impact analysis of universe studies where relevant. 2. Extrapolation: • Design and simulate extrapolation setups for new panels. • In case of major change in the sample of a channel, redesign its extrapolation matrix, conduct impact analysis where relevant. • Support StatOps team in case of minor events (how to select a donor shop for a copy shop or a create shop, how to slightly update the extrapolation in case of e.g. empty cell, discuss quality issues / risks in case of weak / weaker sample). • Assess the compliance of the existing design of extrapolation matrix and eventually recommend improvement (in the design, the sample size = recommend target for recruitment, etc.) • Specify sample checks per country_channel and ensure that they are properly executed by Operation teams. • Jointly with Operation, review the quality of the existing extrapolation and recommend / redesign if necessary (feedback loop from MDQC teams). • Update the Target Sample File every year and in case of significant change in the universe or the sample. 3. Modelling: • Assess the feasibility of non-coop retailer rawdata modelling outside of the system: business interest (in collaboration with Product), availability of data sources, usability of modelling components • Design the solution, test the development, review the quality once implemented, review every year the relevancy of the solution • Once implemented, handover to Operation for ongoing maintenance and execution. • Support Commercial teams and Client Operation to explain the methodology of a Rawdata modelling. 4. Quality and Projects • Implement new country-channel setups, including reviewing channel definitions, designing universes and samples, and providing setup instructions. • Resolve Level 2 methodology clarification issues. • Support Major Data Change required by a redesign recommended by Data Science (major change in the universe, major redesign of an extrapolation, major redesign of a rawdata modelling): support the definition of the action plan, support the client communication. • Participate to Data Science regional and global projects