Sergey Nivens - Fotolia

Text analysis tool helps Lenovo zero in on the customer

At Lenovo, textual analysis software enables analysts and line-of-business workers to have a more informed dialog about customers and their ever-changing preferences.

IT workers and business analysts have always had a hard time explaining the benefits of traditional business intelligence to line-of-business workers, even when the discussion centered on easily quantifiable topics. But now, the same discussions are moving to less obviously quantifiable terrain.

At computer manufacturer Lenovo, sales and marketing teams recently started using a new text analysis tool to measure customers' experiences. Leaders of the project found, at least initially, that they were having many of the same conversations BI teams have had for years with business teams.

"It's difficult to have a discussion between a business team and a data team," Daniel Bieber, global analytics architect at Lenovo, said. "The decision makers in the business are kind of operating off their gut. But whenever I sit the information down in front of the end users, their reaction is amazed."

For Bieber, putting useful information in front of business teams in this way is the best route to build support for using text analytics results. The sales and customer service teams at Lenovo started using the text analysis tool from Taste Analytics in June 2015 to monitor data from product surveys, purchase experience surveys and technical support chats. They also use the tool to scrape text from online product forums and social media. The goal is to track the complete customer experience from purchase through use.

For example, Bieber recently met with a merchandiser on the e-commerce team to talk about customers' product concerns. Bieber was able to show that one of the most common questions customers have is about battery life. The merchandiser had a feeling that this was the case, but the numbers made it more objective. Now the e-commerce team can justify spending time developing how it communicates with customers about battery life in different products.

"The conversations are changed now," Bieber said.

But it wasn't always this easy to convince people. Before implementing the tool from Taste Analytics, Bieber's team used a more traditional BI system to track customer interactions. He said this tool was used mainly for retrospective reporting. It was also was limited in the number of data sources it could pull in.

The main problem with the previous tool was that, when applied to text analytics, Bieber and his team had to spend a lot of time defining words and phrases that they wanted to track. But the Taste Analytics text analysis tool defines taxonomies itself, automatically identifying common themes in a collection of texts. It then builds reports on these self-defined taxonomies, identifying trends and common themes in customers' comments.

All of this means Bieber's team is able to deliver deeper insights more quickly to business teams, which helps the sales and customer service teams stop relying on their gut and start having more discussions based on quantified data.

"When I show them data, they're either surprised, or it's affirming their beliefs," Bieber said. "It makes things more objective."

Next Steps

NASA boosts aviation safety with help of text analytics software

Everything you need to know about textual analysis tools

How text analytics could influence future of records management

Dig Deeper on Text analytics and text mining