-->

Keep up with all of the essential KM news with a FREE subscription to KMWorld magazine. Find out more and subscribe today!

New tools provide deep insights from customer feedback

Article Featured Image

Efficiency and ROI

Banking has changed dramatically since the mid-1990s, when major banks began offering online services to their cus- tomers, and later, apps for mobile devices. Many banks are now internet-only. In 2014, Atom Bank offered the first “app-only” bank in the U.K. The company is very proactive in collecting feedback from multiple sources, including mortgage, saving, and deposit products; reviews; customer complaints; call cen- ter notes; and surveys. However, the insights gained from each channel remained siloed, so a full view of the customer was not possible. With a rapidly growing customer base that doubled in just 1 year, Atom Bank wanted to ensure continuous improvement that incorporated customer feedback.

The company selected Thematic’s customer feedback analysis solution to integrate and analyze feedback from seven channels, including the App Store, Trustpilot, Salesforce, and its CX platform. Thematic analyzes the data to reveal themes from all these sources. Atom Bank analysts can edit the themes based on their knowledge of the company and its customers. The results are used to improve operations, product quality, customer experience, and complaints-handling. Atom Bank experienced a 69% reduction in calls about rejected mortgage requests, a 43% reduction in calls about savings products, and a 40% reduction in device-related calls. Atom Bank is able to distinguish the verbatims that impact metrics from those that do not affect it and can also differentiate among customer segments.

In some rapidly changing industries, a flexible text analytics product is important. “Traditional NLPs focused on training large amounts of data on standard categories,” said Alyona Medelyan, founder and CEO of Thematic. “This works well for industries where things don’t change much and where everyone is talking about the same thing.” However, comments relating to novel industries or technologies such as new software do not fall into existing categories. “Thematic is most useful for companies that have unique products or those that change a lot,” she noted.

Transparency has been built into Thematic. “Many NLPs are black boxes—you don’t know why they returned a result and cannot edit the results,” remarked Medelyan. “When we spoke to researchers, they complained that these products did not allow them to incorporate their expertise. Ours presents an intermediate output and shows how phrases are organized into categories after being analyzed. Also, users can edit themes and the hierarchy based on their own knowledge.”

Text analytics provides clear value for understanding customer feedback, but some barriers remain. “Some companies veer away from it because of getting burned by products that over-promised and under-delivered,” said Kucera. Stuart concurred; “In this field, some companies have PTSD from having been told that it was easy, and then finding it was as laborious as analyzing the data manually,” he commented. “There is a new wave of text analytics, and if people get educated on its capabilities and match the prod- uct to their needs, they will receive significant benefits.” Despite these barriers, text analytics holds great promise for providing the deep insights desired by business entities.

KMWorld Covers
Free
for qualified subscribers
Subscribe Now Current Issue Past Issues