fbpx

Code: EDI-2020-17-SONAE

Domain: Retail

Summary

The goal is to improve efficiency levels to handle customer feedback (e.g. complaints, suggestions) in the form of structured and non-structured data.

Proposed by

 

Sonae is a multinational company managing a diversified portfolio of businesses.

Sonae MC is the market leader in food retail in Portugal, with a track record of over 30 years of consistent growth on the back of expanding the store footprint, managing a highly diversified portfolio of banners, formats and channels and maintaining a culture of continuous improvement and efficiency.

Description

The goal is to improve efficiency levels to handle customer feedback (e.g. complaints, suggestions) in the form of structured and non-structured data.

The amount of customer service requests received by email and web forms (growing online operation) has been increasing considerably but the approach to handle them is difficult to change. In the current operation, customer’s requests are handled manually by the order they are received. Prioritization is not in place and currently, SONAE cannot setup different SLAs for different types of requests and measure that effectively.

SONAE seeks a solution to solve this problem by tackling inefficiencies. It must report all types of requests SONAE receives and handles, since agents spend a significant amount of time manually classifying requests after a big effort of learning their classification taxonomy.

Artificial intelligence (AI) & Analytics can help to solve this challenge. This approach is critical in the context of all web (unstructured) communications from clients (e.g. emails, webforms, social networks). Also, it needs an unsupervised learning approach in terms of Machine Learning.

Data

The challenge has the following sample datasets available for download

Expected outcomes

At the end of the project, SONAE expects the following (AI and Data Science) deliverables and results:

  1. Taxonomy review – SONAE would like to take this opportunity to review its taxonomy, making sure they have the foundation in place to then triage requests. The team should propose a specific taxonomy derived from the data analysis performed.
  2. Automatic categorization – classifying incoming requests automatically according to the taxonomy defined will allow them to then work on workflow optimization.
  3. Workflows definition and efficiency optimization – prioritize requests, define specific SLAs for critical cases and route certain requests to certain teams automatically should be enabled. This should be integrated with Salesforce (preferably).

The following business goals should be measured:

  • Agent Productivity will be increased by, at least 5%
  • Average Handle time Per Ticket should be reduced by, at least 5%
  • Cases automatically solved should by increased by, at least 1%

 

How do we apply?

Follow by Email
Facebook
Twitter
YouTube
SOCIALICON