Project details

Artificial intelligence for large-scale retail trade.

Published by: CONSULTING COMPANY

Status: READY FOR PUBLISHING

Category: DATA INTELLIGENCE

Application domain: Marketing

Budget (EUR): UP TO 15000

Project description

A company operating in the large-scale retail trade is searching for innovative artificial intelligence based approaches to support commercial strategies to increase stores sales volumes.

Typical challenges in this area are:
- evaluate the impact of the opening of a new store, in terms of customer acquisition and / or loss due to competitors near the new site (even of the same brand of the company)
- predict the sales volumes of product group (or at least those whose sales volumes are the most important), isolating the most influencing variables
- evaluate the impact of specific marketing initiatives.

At present, the strategy definition is transferred to a dedicated team within the commercial department, and based on the experience of the previous years and on the workers ability. Though the internal procedures are well-established, and partially implemented in traditional instruments (e.g. Excel), now the company wants to develop innovative approaches to increment the efficacy of the strategies and to exploit artificial intelligence to reduce time consuming manual operations.

Ideally, the final target is to relieve the commercial team from manual time-expensive operations and focus on smarter tasks (e.g. scenario analysis rather than the definition of one scenario).


The solution should be delivered as a stand alone software to be run in parallel with the existing support tools.

Data are typically available as time histories, but further data may be retrieved from the company database upon request.

Project goal

The project goal is to develop a mathematical based approach to support the commercial strategy in defining marketing initiative to increase sales volume. The approach should combine models to forecast the sales volumes and the impact that specific strategy may bring, according to traditional (time histories, customer records, planned events, season and calendar effects etc) and non-traditional data (text-mining).