How can text analysis improve your customer service? Part 3 – Customer feedback

In the previous two parts (Part I; Part II) I have already talked about how to improve customer service by handling your emails and phone calls smarter. In this post, I will focus on customer feedback and share few use cases how analysing customer feedback wisely can help you improve customer experience in your company.

Customer feedback, on one hand, is very important knowledge to get, but on the other hand, if very many customers leave feedback, it gets difficult to manage all this new information. Luckily there are some options that help companies to manage and understand better their customer feedback.

Use case I

Net Promoter Score (NPS) is used for customer perception research. Even though the survey process itself is automated, the analysis part is time-consuming. Customer feedback questionnaires usually include both, giving a score for the service and a free text option for comment. Since people are different and feel things differently, there might happen discrepancy between NPS score given and sentiment in textual comments. Therefore to really understand how customers feel about your company and services, the comments need to be analysed more deeply.

Feelingstream has built a solution that helps companies easily do that. Our custom trained machine learning models have high accuracy to automatically detect sentiment and topics in customer feedback comments. This means instead of only taking into account given NPS scores, from text analysis it might be seen that the customer was actually very happy with the service, but just chose the wrong NPS score. The software also gives you an automatic callback report for customers with a negative experience to contact with them.

By analysing NPS that way, your company spends less time and effort on customer given data analysis and would get higher accuracy in categorization, while no biased categorization is included. As well it gives you improved capabilities to listen to your customers and react to the changes in the market.

Use case II

Instead of asking for feedback from customers, there is also another option to know what they think about you and your service. Customer opinion can also be found in their calls, emails or chats. Why would this be better than asking for it? Only 15-20% of customers respond to questionnaires. As well asking feedback annoys customers.

With Feelingstream solution companies can detect from all email and chat texts or phone calls the content, sentiment and search if and when the most common words for feedback were used.

As a result, the process of giving feedback is easier for customers (as it does not exist in the traditional way by asking questions), and you would get the most honest feedback as the customers don’t know that you are catching signals. Also, by easily systematising all data then all your customer perception is evaluated.  

Customer feedback is valuable to every company and it is even more valuable if understood right. Feelingstream can help you do that.

Do you systematically ask for customer feedback and would need some clarity in it? Contact us and let’s find a solution!

Related Posts