Development of automated customer feedback monitoring systems for retailers

Our customer: Popular clothing brand
One of the largest hypermarkets of building supplies
Context: Casual, simplistic analysis of customer reviews

Poor understanding of customer problems and attitudes towards stores and services

Lack of benchmarking against competitors
Industry: Retail

Objectives

To provide automatic classification and systematization of reviews
To create a mechanism for analyzing consumer reviews of stores, products and services
To identify problem areas and growth points

Tasks

Solution

Result

Develop an algorithm to automatically classify and process reviews
Create an analytical tool to study customer attitudes towards stores and products
Identify customer attitudes towards stores and services compared to competitors

Tasks

Solution

Result

Natural Language Processing (NLP) tools and the Random Forest machine learning algorithm, which combines a set of independent models, were used to solve the tasks. Internal store data and comments from open sources were used as a source of information
An algorithm for classifying, processing and analyzing customer reviews has been created
The attitudes and preferences of buyers have been analyzed, problem areas have been identified
An analysis of attitudes towards stores and services in comparison with competitors has been fulfilled

Tasks

Solution

Result

Provided store ratings based on customer satisfaction
Identified key issues and suggested ways to address them
Assessed the speed and quality of staff responses to reviews and identified areas for growth
Provided an analysis of store ratings against competitors