We are already familiar with all the countless applications of speech recognition for the general public, especially through voice assistants. Companies, too, can benefit from this technology in various fields and areas of application. Today, we are going to adopt a more marketing than functional approach when speaking data, especially that contained in the voice. It is important to know that the voice is governed by prosody, the set of oral traits we give to our verbal expression, in order to make our emotions and intentions more intelligible to our interlocutors.
Rita Singh, research associate professor at the School of Computer Science at Carnegie Mellon University (CMU), leads a team that develops voice profiling. “It’s about deducing all sorts of parameters from a person’s voice. The relevant parameters include everything from physiological parameters to demographic and sociological parameters. The list is very, very long,” says Singh. Companies are able to take advantage of this information, but how and, more importantly, why?
Let’s talk about the how.
The most obvious thing at first glance, but not necessarily the easiest, is to integrate speech recognition technology into the customer journey. Of course, this can take many forms, as the technology is intended to be versatile. From a bot installed on a call center type call platform, to a full-fledged voice assistant integrated into a product or on a site/application, the principle itself is to capture the voice of the customer. The form that the solution will take obviously depends on your choices and specificities. It is essential that brands use voice to create meaningful interactions that enhance, not interrupt, people’s lives. It is within this same ideology that the voice experience must be thought of so that the idea of profiling shows its full potential, both for you and your customers.
Let’s come to the why (with examples).
Identify its customers more precisely.
Voice data, extracted from the prosody we were talking about at the beginning of the article, is, like behavioural data (type of purchase, recurrence, number of visits etc…), necessary information to better understand and identify its customers. As a real decision-making tool, voice data can translate many elements to help you: gender, age, emotion felt during an action are actions that are now feasible, in a more theoretical aspect: the estimation of the height and weight of the person, his ethnicity, his place of residence, his level of education, his level of fatigue etc…
Improve the customer experience.
As an essential component for many companies, the quality of the customer experience is increasingly becoming a source of competitive advantage in today’s markets. Indeed, where the offer is struggling to reinvent itself, companies have a wider scope of action on how they deliver it. Whether it be prospecting, after-sales service or any other customer contact, voice information complements customer knowledge. In doing so, it is possible to improve the customer’s journey and experience by analysing the information collected. For example, create an algorithm for automatically suggesting products or services, calibrated to the gender and age of the person, automatically via the voice request. As we know, this is already possible today, but much less accessible because it requires external data and/or an authentication system. Except that the customer experience is important before the potential customer is transformed into an actual customer.
Refine the customer relationship.
With this additional voice data, the entire customer/company relationship can be improved. We’ve seen the steps further upstream in this relationship, what next? Quite simply, let’s take an example. An operator, in charge of customer support services, can benefit from key real-time data on the customer, including his emotions, in order to adopt a speech adapted to the situation to guarantee the best possible service. If it is important to convince the customer with dynamic marketing actions according to his profile, it is even more important to maintain a sufficient level of satisfaction to keep him loyal, where data is still essential.