Everyone who uses surveys knows that not all questions are objective, there are open ended questions too. Questions like - What did you like the most about our product/service? Please tell us how we can improve? Why did you give us this rating? are open ended that need to be answered and analyzed properly.

As you would all know, a survey rating (An NPS score, CSAT score) is a number that can help you understand your rating, but for a more complete understanding, the actual value comes when you receive these open ended responses.

If you receive these open ended responses, as a brand, it can be about your product, service, marketing experience or about anything else. They can also be about a particular context depending on the question asked by you. And once respondents start replying, you need to be sure you make sense of this qualitative data.

Making sense of 4-5 responses, or even 100 such responses manually is still possible. But, what do you do if you have thousands and millions of such responses coming in everyday?

A lot of organizations have to spend hours of hard work and human resources in reading all such responses and in making sense of all the data through the means of manual tagging. What the brand ends up doing is reading the entire comment and then making sense of the same. Another matter of concern is that a repetitive task like this paves way for errors and also delays the organizations in acting upon the concerns of the users.

However, this does not mean that understanding the Voice of Customers (VoC) is not important. In fact, it can sometimes become the most essential part of any analysis and open ended questions often provide the most valuable and out of the box insights. Having said that, it’s not always easy to assess words. For organizations that do this kind of analysis manually, it takes a lot of time, effort and money.

This is where the Conversational Analysis feature launched by SurveySensum comes into play. Through this feature, SurveySensum helps its clients in automatically analyzing this data. This helps organizations in acting upon the problems quickly and in prioritizing the time of the company to actually solving the problems rather than spending long days and nights in detecting the problems.

What you start doing with SurveySensum is that you start tagging these responses. You can categorize different comments. For example, you can start tagging responses relevant for customer support into customer support, responses relevant to pricing into pricing, or under whichever category you deem necessary. And once you start tagging these responses, in some time, SurveySensum will start predicting the categorization of the responses automatically, i.e. the SurveySensum machine will learn the tagging process for you saving you hours of resources.

Unlike the long spreadsheets and excels where you always have to tag the responses and then make sense of the data received, SurveySensum will start predicting once you train SurveySensum. This training would be such where you tell SurveySenusm that x comments belong in a particular category and y comments belong in the other category and this training would be done once you tag at least 50-60 responses. Essentially, the more you tag, the more accurate the machine’s responses would get.

Once you are through with all this tagging and SurveySensum starts automatically predicting the responses for you, you can go to Tag Analysis and you can see the top topics that are coming out. You can look at the top topics and can share the same with the relevant teams so that they can take relevant action on the same as soon as possible. This would give you a clearer picture as to why your NPS rating might be low or what is affecting your customers the most. Once you have this information, you can prioritize and work on solving those issues.

For example, if you are an organization getting responses on your NPS survey and you get the information that the customers are facing maximum issues in logging in to your application, you can share the same with the product team who can further deep dive into the issue and resolve it at the earliest. After you have resolved the issue, you can monitor the same and understand, through the incoming responses, whether the issue has been resolved or not.

In a similar fashion, gradually, you can start fixing all the things that might be troubling your customers.

Note 1: This feature is currently available and limited to the surveys.