August 30, 2017 | Written by: Amit Kumar
Categorized: Content Analytics
Share this post:
Today, unstructured information represents more than 90% of the information within organizations. This IDC Case Study estimates that the digital universe will grow 40% per year over the next decade and, by 2020 it will reach an astounding 44ZB or 44 trillion gigabytes. Most of the unstructured information exists in the form of patent filings, research papers, blog posts, news articles and social media data. Data that is proliferated across public forums and poses a significant challenge for organizations to understand, digest and use this unstructured information quickly to identify business challenges early. Cognitive search solutions are augmenting the capacity for organizations, like Honda, to extract, index, analyze and derive insights from unstructured information without adding large amounts of manual effort and resources to the mix.
Making sense of customer feedback
Honda is a $200 billion global automotive manufacturer and one of the largest automotive manufacturers in the world. Its global customer base exceeds 20 million. Honda’s Quality Assurance division gathers customer feedback through its vast network of suppliers, dealers, and service centers around the globe. Customer feedback data is used to identify quality issues and improve understanding customer sentiment. Analyzing customer feedback quickly is critical for to Honda’s business success. It helps Honda put in place measures and policies to mitigate quality issues and increase part safety before it impacts customers and revenue. It also helps in improving the quality continuously by sharing the insights from customer feedbacks with R&D and production departments.
The process of gathering and classifying customer feedback was extremely time consuming. On an average, Honda received 310,000 messages per month only from Japan. Most of the customer feedback came in free-form text in many languages. It required QA division employees to read and classify each piece of customer feedback, spending over three hours each day to do so. Being a global automotive manufacturer, each region managed customer feedback in its own way, making it even more challenging to aggregate feedback from different regions and search for global patterns in quality issues. These challenges constrained Honda’s QA division from identifying part defect patterns and quality issues to take early corrective actions.
Cognitive Search streamlined Honda’s QA division feedback analysis by 80%
Honda’s QA division needed to accelerate the process of understanding the vast amount of unstructured information coming from customer feedbacks. Honda collaborated with IBM to implement a cognitive solution to extract and classify vast amounts of feedback coming from all over the world. IBM Watson Explorer helped Honda read, extract, and organize key pieces of information in the customer feedback. It also provided visual dashboards and reports to highlight key patterns and insights uncovered from analyzing thousands of feedback submissions. A detailed dictionary was developed to help the system identify key information such as car parts, problem symptoms, and other automobile-related information.
Watson Explorer’s Cognitive Search and Natural Language Querying capabilities helped Honda’s QA division reduce the time required to understand customer feedback by 80%. The cognitive insights provided by Watson Explorer also enabled employees to respond more quickly to quality issues and discover complaint patterns. Honda now has a much clearer understanding of the issues faced by its customer and accelerated responses. Overall, the results have greatly exceeded expectations, reduced costs, improved performance, quality, and productivity.
Download the IDC Case Study: Honda Is Using IBM Watson Explorer to Drive Real Changes in Quality Assurance to learn more.