AI-Powered Customer Feedback Evaluation

10.7.2020

In an increasingly busy work environment, we are often not able to work with data in the intensity we would like to. We ask for feedback to improve our tools and processes but don‘t find the time to get the maximum insights. This is why we have come up with an AI-Powered solution.

The analysis feature works for recurring or one-time surveys where you collect lots of customer reviews including text fields. Thanks to the AI-based platform we can cluster and highlight customer feedback in free text fields. No need to read through all that jungle of comments any longer. We can focus on the most important reviews and find out what customers propose to us concretely to improve the tools or services.

What is the Customer Feedback Survey about?

We have started testing this application within the Procure to Pay Service Line and analyzed the free text fields that the service line receives regularly for the internal shopping system. Since the launch of the system, colleagues had received on average 500 comments per month. The customer had something to say, wanted to be able to read quickly the feedback to get the most out of suggestions and feelings.

How did Artificial Intelligence help?

With the help of Natural Language Processing (NLP), we let machine read and prioritize all that feedback for us. Hence, we were able:

  • to cluster feedback according to defined categories
  • to highlight the most positive and most negative comments with sentiment analyses
  • to identify linguistically, which feedback contains a concrete suggestion for improvement

and all this automatically!

Once all ideas and suggestions for improvements in customer feedback are identified, we show them in a structured and visual platform, like Celonis or Tableau.

What’s the impact on the service and team?

With our AI-powered analysis for customer feedback, we can gain deeper insights into customer opinion and boost the efficiency of the team so that the colleagues know which tasks to prioritize to sustainably improve the tool or service. AI supports us to find out the pain points of our users quickly so that we can continuously improve our customer satisfaction. We already realized ‘QuickWins’, like optimization of the product descriptions or implemented catalogue updates, both of which were proposed by our customers directly.