AI Case Study
Honda reduced the time required to understand customer feedback by 80% using AI
Honda leverages IBM Watson Explorer to analyse big amounts of unstructured customer feedback. The company's quality assurance division uses the Explorer to understand received messages
IBM Watson Explorer helps Honda’s quality assurance division understand received messages and make sense of the enormous amount of customer feedback, which are mainly unstructured.
The content analytics capabilities were key to Honda's requirement to be able to understand and extract the key phrases and concepts from their customer feedback and build that into a set of reports and dashboards that highlighted specific problems rather than having QA field personnel to manually read through thousands of documents.
Industry
Consumer Goods And Services
Automobiles And Parts
Project Overview
"IBM GBS selected Tokyo System Research Corp. to help it implement an IBM Watson Explorer Advanced Edition solution. To do this effectively, a detailed dictionary needed to be developed to help the system identify key information such as car parts, problem symptoms, and other automobile-related information.
IBM and its partner chose to use IBM Watson Explorer Advanced Edition, which provides enterprisewide information access and unified information applications capabilities across internal and external data sources and services. In addition, Advanced Edition includes advanced content analytics capabilities to enable organizations to adapt their information access solution to specific domains and to extract insights from unstructured information to help identify trends, patterns, and relationships in their data. The content analytics capabilities were key to Honda's requirement to be able to understand and extract the key phrases and concepts from their customer feedback and build that into a set of reports and dashboards that highlighted specific problems rather than having QA field personnel to manually read through thousands of documents.
The solution included IBM Cognos Business Intelligence graphs and reports to summarize topics, categorize issues, and highlight trends. The comprehensive feedback analysis is now contributing to finding issues early, dealing with issues promptly, and providing feedback to parts analysts and designers."
Reported Results
"The solution reduced time required to interpret and classify incoming feedback by 80%. For example, reports about "abnormal noise from front damper" could be extracted and automatically associated with the phrase "front damper" from the parts dictionary, the "abnormal noise" identified with a symptom dictionary be used to find similar cases that connect the two."
Technology
IBM GBS selected Tokyo System Research Corp. to help it implement an IBM Watson Explorer Advanced Edition solution. To do this effectively, a detailed dictionary needed to be developed to help the system identify key information such as car parts, problem symptoms, and other automobile-related information.
Function
Customer Service
Technical And Product Support
Background
"Honda's process for gathering customer feedback about issues and classifying this information (e.g., whether similar symptoms have already been reported in the other data sources, whether the issues had already been solved by Honda, whether the issue is warranty related) was extremely time consuming as individuals had to read and classify each message, which averages about 310,000 message per month in Japan alone. Moreover, Honda's regional organizational structure meant each region was responsible for dealing with customer feedback from that region, which made it more difficult to aggregate input from multiple regions and search for global quality or part failure patterns. According to Honda, QA employees spent up to three hours per day reading and reviewing customer feedback, including warranty claims. Honda needed a way to automate all this manual work and effort."
Benefits
Data