Comments, Topics an Opinions can be broken down by Metadata fields to reveal Volume and Sentiment trends.

Segmenting customers - Feedback & Demographic data:

Find out if different categories of customers value different things using demographic metadata: For example, male vs. female, different age groups, level of education.

Prioritizing customers - Feedback & Satisfaction rating:

Do you need to focus on making happy customers even happier or preventing the unhappy ones from churning? Prioritize where to take action based on the segment you're interested in. 

Ensure consistent service across all locations - Feedback & Location data:

Are certain locations doing a lot better or a lot worse? Learn from high-achieving locations and improve the ones lagging behind. 

Here's a step-by-step walkthrough. In this example we are using a dataset comprised of online reviews of a food delivery app.

Along with unstructured customer feedback, you'll need to upload accompanying metadata such as location, satisfaction rating, demographic data or structured data specific to the dataset such as product or brand data.

The text mining needs to be complete before the system can create the Correlations.

For each identified Comment, Topic or Opinion a Correlation is displayed individually.

In this example, we are interested in the Comment “customer service is terrible.'' Click on the row to reveal the Comment Details for that specific Comment. 

The Comment Details popup has tabs displaying Records, Correlations and the Level of Interest over time. Select the Correlation tab to reveal a bar graph that can be adjusted to whichever metadata field you choose. In this case we chose "Age" range. Most of the Comments meaning "terrible customer service" came from people over the age of 65, followed by those in the 25-39 age range.

The metadata field can be adjusted by choosing from the dropdown accessible by clicking the filter icon at the top-right of the graph. You can show up to 10 bars and calculate either by Record or Comment.

Note: When creating Correlations from the Comments view only one Sentiment will appear as Comments are bucketed into one of the sentiments during the text mining phase.

Here in addition to Volume we can see a breakdown by Sentiment. In this case we moved to the Topic view and chose the Topic "Driver" and the metadata field "City". We can see that customers from New York, Toronto, and Los Angeles had a lot more negativity toward their food deliver driver that those in Denver and Chicago. This is a good opportunity to see what issues customer is those cities have related to "diver" and get ahead of them to keep those clients happy!