mercari AI

Improving Customer Support Productivity through Machine Learning

Last Updated :

Summary

We also provide information in the Mercari Guide to help customers solve problems on their own before contacting us. We are currently working to improve access to information so that customers can find the information they are looking for on their own, which improves the self-resolution rate.

Purpose

Machine learning can be used to automate and streamline part of the contact process and encourage self-resolution by giving customers access to appropriate information when they have problems. This leads to an increase in customer support productivity.

Description

At Mercari, we are continuously improving our app to ensure that customers have no problems during use. That said, it is not possible to completely eliminate problems. It is therefore necessary to provide some level of support when a customer encounters a problem. In such situations, a customer in need can take the following steps:

  1. Find information that may help them solve their problem
  2. Contact Mercari when they cannot find such information

The first step is called “self-service,” and we have collected information in the Mercari Guide to help customers solve problems on their own before contacting us. We are currently working to improve access to information so that customers can find the appropriate information on their own, which improves the self-resolution rate. With respect to the second step, the Customer Support department handles countless customer inquiries every day. In such situations, we are aiming to improve the customer experience by automating and streamlining parts of the contact process. As mentioned earlier, there is a need to encourage customers to access appropriate information when they have problems and resolve them themselves, and to automate and streamline parts of the contact process. Using machine learning is one way of achieving this.

Mercari has in-house tools such as the Mercari Guide and the Contact Tool, and machine learning-based functions can be added thereto in a flexible manner. For example, the Contact Tool can recommend a template for operators to use when replying, and give suggestions to enable quicker replies. Another proposal to improve the searchability of the Mercari Guide is to sort search results according to the relevance of the query to articles. We are working together with the Contact Tool Development Team to make these new ideas a reality.