Customer Relationship: How AI Gives Meaning to Data

Home AI in Business Customer Relationship: How AI Gives Meaning to Data

Applied to customer relations, artificial intelligence carries several promises, two mainly: to offer users an increasingly efficient self-service and to increase the number of contact center agents within the customer context. However, the available data useful for setting up the context are rarely used, while the cases of use are numerous.


Always more data to improve customer knowledge


According to MongoDB, the American database platform, 80 to 90% of the data generated and collected by organizations is unstructured. By definition, data is represented or stored without a predefined format. The fields of an e-mail message, for example, cannot be analyzed by traditional analysis tools (even if the metadata of the e-mails gives them a certain structure which may make them look like semi-structured data.)


Artificial intelligence, for its part, will make it possible to exploit this information, by detecting intentions in the messages and in the attachments of these messages. A need identified by the Davies Hickman – Odigo* study. This tells us that the business leaders surveyed rank first (82%) in the need to know their customers better (age, vulnerability, etc.), in order to offer personalized experiences. This need is complemented (74%) by a desire to better understand the emotions and intentions of customers.
These data will therefore bring tangible value, they enrich customer knowledge and build the context, on which the agent will rely to deliver a personalized and relevant conversation. However, care must be taken to select useful data from this ocean of available data.


It’s all about context


The context is made up of all the interactions (regardless of the media used) between a brand and its customer, supplemented by all the information already known and stored in the information system. The information ranges from purchase history to delivery address preferences, anniversary date, etc. The context will therefore bring together the information, authorized under the GDPR, which will allow a detailed knowledge of the customer’s needs and their expressed and detected needs (intentions). It is this context that will make it possible to augment the agent.
With the knowledge gathered, it is thus possible to support the agent by offering him (within his business interface or the CRM) a set of responses constituted according to the intentions detected. Respondents to the study* consider, at 89%, that AI should allow them to match and route customer requests and questions to the right person. The context, therefore, plays a preponderant role in the routing and information of the agent. It also plays a role after the client-agent exchange.


Toward an augmented supervisor


The role of AI does not stop at offering automation of request processing (self-care) or enriching the context for the agent. Indeed, the supervisor will also be able to use AI to analyze the conversations previously recorded and transcribed in text format, through a speech-to-text (STT) application.


Insurance companies must therefore ensure that conversations concerning commercial canvassing are “usable, which implies that these communications can be listened to, copied and exported without their original recording being modified or erased”. As in the banking world (MIF2) where AI will make it possible, on the same principle, to provide proof of compliance by facilitating research in transcribed conversations. AI reveals other use cases, such as sentiment analysis, which will identify in messages and conversations the weak signals emitted by dissatisfied customers. Improving agent training and the quality of exchanges by analyzing transcribed conversations are other areas of application of AI to customer relations.


Artificial intelligence is useful at each stage of the customer relationship, for the greater benefit of the customer who will have a memorable experience, and the agent/supervisor couple whose work will be facilitated.


The interest in access to context is such that some organizations, such as banks, are seeking to extend the functional richness provided by AI to all company employees, whether or not they are agents. Why should access to better customer knowledge be reserved for contact center employees?