The “Topics” page shows you a birds-eye view of what your survey respondents talk about in their text feedback and their sentiment toward specific areas of your business.

Click here to learn more about topics and how they’re added to survey responses.

In this article, we will cover what widgets and reports you will find on the Topics page and will help you make sense of them.

The top section on the page will display a couple of widgets, each containing a list of the top five positive and negative topics. These are the topics that have been applied to most survey responses, and based on whether they’ve been assigned mostly positive or negative sentiments, they will appear in the corresponding widget.

Each widget will include the following elements:

This widget shows you a list of all the topics applied to at least one survey response. You can also change the graph’s style using the view toggle and switch to the bubble view.

By default, the widget will show you insights based on sentiments associated with topics, but you can switch the data source to survey responses. In the following paragraphs, we will cover the difference between these two modes and the metrics they offer.

When a topic is assigned to a survey response, it must have a sentiment associated (positive, neutral, or negative). Therefore, even though the topic might have been applied to a survey response that falls in the “unsatisfied” category, the topic’s sentiment can be a positive one.

For example, let’s assume that a survey respondent answered an NPS survey with a rating of 6 (unsatisfied) and left the following comment: “I have encountered a bug that made me lose all the progress. However, I love the new feature, and hopefully it’ll become more reliable”.

Based on this example, the response would have the following topics and sentiments:

So, if you’re viewing the widget in the “by Sentiment” mode, then you would see the “Feature” topic in the list with a count of 1 positive sentiment. But if you were to switch to the “by Rating” mode, then you would see 1 unsatisfied response associated with this topic.

This widget’s main metric is the Impact score, which helps you understand how impactful a topic is based on a list of criteria, such as the topic’s relevance to your business, its severity in the context of a survey response, frequency, etc. Next to the score, you will also see its dynamic over time based on the selected comparison time range. Basically, it’s a percentual difference between a topic’s impact score at the beginning of the comparison period and today.

Check here to learn how the Impact score is calculated.

Finally, the colored bars show how often a topic was applied to responses while having positive, neutral, or negative sentiments associated.

Viewing data based on rating means that we’re looking directly at the survey response and the category it falls in: satisfied, neutral, or unsatisfied. To explain with an example, an NPS survey response would fall into these categories:

This widget’s main metric is a topic’s weight. The weight is a percentual representation of how many times a topic was applied to responses compared to the rest of your topics. Basically, the weight will show you a leaderboard of your topics and will help you immediately identify the most popular ones. The weight metric is followed by the dynamic percentage, which will show you if a topic has increased or decreased in popularity within the selected comparison date range; in other words, topics that people started mentioning more often in their text comments or, on the contrary, topics that people rarely mention anymore.

The colored bars will show you how many satisfied, neutral, or unsatisfied survey responses this topic was applied to.

Depending on your preference, you can also quickly switch between the table view and bubble view for the Overall Topic Analysis widget.

When needed, you can export the data displayed in the Overall Topic Analysis widget as a CSV file.

The file will include the following columns:

By rating:

By sentiment: