Feelingstream Product Overview – What is the DETECT Module?

Previously we wrote about how activating your creative side will help you use the LISTEN Module. In that part, the user has the freedom to search for patterns. So let’s say you’ve reached the phase where these patterns are properly defined and you’d like to detect them.

At this stage, the DETECT Module – one of the three Modules of the Feelingstream Platform – will come under focus. In order to improve customer experience, it allows each company to implement business critical values within automated text analysis. Manual data review might at first seem like an effective way to get a good insight of customer-agent conversations, but it is very time-consuming.

Compare how quantity affects quality. In one minute, an automated approach can detect sentiment in about 300 texts – how many texts would you be able to review manually?

For example, over the course of several months, customers have been calling for the same reason. You know that these calls follow a similar pattern. Is it really reasonable to detect them by hand if you could train an AI-supported solution which does this automatically and gives a wholesome picture?

Think of it as a sorting process. Imagine you have a box full of LEGO bricks in various shapes and colours. You want to have this messy set of items systematically categorized so you’d know exactly how many different colours were in that box, how many of them were 4×4 blue bricks and whether any of the items weren’t bricks?

What is the DETECT Module?

Detection is the process of finding something more specific within the given data and it is beyond a simple keyword search. A set of keywords and/or phrases may formulate a broader category and these categories can be of great use for tracking topics and sentiments of each call. The categories can help us detect for example:

  • whether the call was positive, negative or neutral;
  • if there was a sales opportunity or churn risk;
  • why the client contacted in the first place.

Sentiment detection allows the user to get a good overview of how happy is the customer in parallel with how the agent’s performance affects this happiness. Or vice versa – track down the negative calls and try to understand what went wrong, because after all, more happy customers mean more profit for your business. Quite often you can see patterns why negative calls occur.

An important way of improving company sales is by looking if any of the calls had sales potential and whether the agent could notice this opportunity to sell a useful product to the customer. Similarly, you can detect churn risk, i.e. what causes customers to stop using your services. A good example in both cases is when a client mentions getting a good offer from a competitor – this means the agent can attempt to sell a better deal, otherwise, your clientele will become one member shorter.

A common scenario is that the customer has an issue and contacts the company’s customer service several times, but each time agents gave a different solution. As a Team Lead, you can fix this issue by improving agent trainings so that they would solve issues in an even manner.


(Photo source: LEGO.com)


Related Posts