Schwan’s Home Service, Inc., a business unit of the Schwan Food Company, markets and distributes approximately 400 frozen foods under the Schwan’s® brand through home-delivery and mail order services. Featured product lines include the company’s signature ice cream, pizza, choice meats, seafood, ethnic specialties, breakfast items, and desserts.

Schwan’s sends out weekly national surveys to their customer base for feedback on products, service, quality of goods, and other criteria. However, with thousands of survey responses coming in per month, there isn’t enough manpower to process them all in the old school way. Until recently, Schwan’s had no efficient way to monitor the qualitative data coming in. This was a big shortcoming because the most valuable survey information was in the comments section. They did not have a way to get reliable, real-time insights into direct customer feedback. Initially, Schwan’s team read each comment one by one, manually coded them, and tried to determine trending themes. Most of the information they extracted seemed to fall under broad categories, such as “this product is great!” or “I’m very pleased with the service.” Nice compliments, but terrible in terms of actionable data. As such, the insights remained uncovered, and still they had no way of hearing the voice of the customer. To remedy this, Schwan’s turned to Semantria’s Excel Add-In for a solution. Currently, the Schwan’s team exports their openended survey data from their survey systems, and then opens it in Microsoft Excel with the Semantria Add-In installed. With a few clicks, the Add-In processes Schwan’s data, categorizes all of the comments according to a custom taxonomy (set of categories), singles out the main themes and ideas therein, and detects the sentiment signals around trending customer issues. Semantria turns its unstructured, open-ended responses into actionable data in a matter of seconds.

Schwan’s originally used Semantria to analyze their content that had low scores, with the intention of extracting the most common customer complaints, tracking these complaints to their source, and then remedying the issue at the point of origin. With the assistance of Semantria’s Data Analysis team, Schwan’s built a set of categories customized for their needs in a matter of hours. Using this taxonomy, all of Schwan’s open-ended responses are now quickly categorized and grouped. Furthermore, using Excel’s basic VLOOKUP function in conjunction with Semantria’s Queries, Schwan’s is able to pull in sentiment scores, important keywords, and other text analytics data, group it by problem type, and associate it back to the original customer and delivery ID within minutes.

Questions

1. Why was Schwan not using results to open-ended questions?

2. How did Semantria help them deal with this limitation?

3. What are they getting out of open-ended questions now that they didn’t get before? How hard is it?