Semantic analysis in contact centres


In order to be able to optimize its customer experience, a company must know what is being said about it and its environment. Dissatisfaction about price, the claims of a competitor; they must be aware of everything. Against this background, it has become essential to closely analyse the reasons behind a large number of phone calls received by its customer service.

To do this, a leader in the energy sector with very large call centres that handle several tens of millions of queries each year must have reliable, effective semantic analysis solutions. The collection, processing and analysis of massive amounts of data are all complicated, demanding operations.


Technical solution provided

A company can carry out the semantic analysis of all phone queries received by its customer service, both in order to meet the expectations of customers and prospects, and also to reduce its rate of customer attrition. The Nice Analytics suite is a powerful tool that allows the vital voice of the customer to be studied. It collects voice data flows in real-time on different IP PBXs (or IP telephone system) and studies them. Before implementing this solution, an audit of the company’s needs, objectives and priorities must be carried out.

In the case of our client who is a leader in the energy sector, this resulted in the deployment of a large system hosted in the CoverApps Data Centre. Using this, operational agents now benefit from the automatic display of an information sheet on their workstation for each inbound call. This allows their communication to be personalized and their actions to be better targeted. During the conversation, they can receive several graduated notifications up to the point where they are required to transfer the call to a dedicated “customer retention” unit.


Operational benefit

Nice Analytics is the ideal partner for contact centre agents. With its advanced semantic analysis functionalities, it is complementary to the work they do. Indeed, a human agent is unable to detect all of the underlying messages in a conversation.

A semantic analysis tool can, for example, interpret instantaneously in real-time, or at a later date, the changes in rhythm and volume of a conversation, and also modulations in frequency. It provides agents with the necessary information. Armed with this information, agents have increased autonomy and ability to take action, thereby increasing their productivity. As a consequence, they are more available to concentrate on calls with higher added-value.