water swirls


Learning to hear each customer’s true voice

Deloitte on Cloud Blog

Your agents listen to your customers when they call your contact center, but how can you hear them too? How can you truly understand what customers need and how to help them? Do you really know why customers are calling, what makes them happy, or what’s really driving their complaints?

July 24, 2020

A blog post by Reny Vargis-Cheriyan, Partner, Deloitte UK

Learning to hear what’s said, and what’s not said

It’s not just what your customers say, but how they say it that matters. Agents get this. They hear not only the customer’s words but also the sighs, pauses and intonation and, hence, understand the real meaning. But how can you do the same without listening to every call yourself!?

With the right technology, that can change. TrueVoice, Deloitte’s cloud-based voice and speech analytics platform on AWS, uses machine learning (ML) to help you understand what your customers say, how they say it, and what it really means, so you can improve performance in your contact centers, increase customer satisfaction, and reduce cost.

Modeling emotional signals

TrueVoice ML models analyze transcribed audio streams together with associated acoustic measures for all customer conversations. Such acoustic-based emotional indicators—like pitch, intensity, pauses, and speech rate—capture non-verbal cues. In this way, TrueVoice is able to paint a fuller picture of what customers are really saying, so that your analysts and others don't have to rely on words alone to infer meaning. Further information can be found here.

How it works

TrueVoice turns each customer interaction into text. It then analyzes the properties of the interaction including speech amplitude, pitch, tempo and non-verbal communication like sighs or long pauses. Then, TrueVoice combines that data with additional customer data to provide a complete picture of the interaction.

TrueVoice further compares data for each interaction against model data, and the results can provide insight into potential problems or opportunities to generate additional revenue. For example, it can detect negative customer emotions or call-agent pressure—all early warning signs of complaints—to help you become predictive and proactive in forestalling customer dissatisfaction. It can also help you pinpoint successful interactions, learn from them, and incorporate that insight into agent training to replicate the success.

Key business outcomes are displayed, so you can review calls and take needed action quickly to improve call-center operations, improve agent performance, and—ultimately—improve your overall customer experience, reduce churn, and enhance revenues. Plus, enhanced ML enables TrueVoice to continuously learn and improve its performance, which means that it gets smarter and faster, and you do too.

The technology behind the scenes

TrueVoice has cloud-based infrastructure, which gives amongst others, the benefit of low-touch API integration, scalability on-demand and consumption-based pricing. A wide variety of cloud-native services may be used, for example, Amazon Web Services’ Elastic Container Service for orchestration and SageMaker for model training. The model training protocol itself follows typical ML best practices such as cross-validation, hyperparameter tuning, model diagnostics, and performance evaluation reports.

Example use cases

Although there are many scenarios in which TrueVoice can be used, two of the most popular are automated Quality Monitoring and identification of demand drivers.

Automated Quality Monitoring

Many companies want the capability to review their contact center interactions and identify outcomes of interest to better understand their customers, or to help mitigate risk.

However, many of these organizations review less than one percent of their calls and select them at random. This means they have no way of knowing what’s happening in the vast majority of their customer interactions.

Instead, TrueVoice enables up to 100 percent of calls (and other interactions) to be monitored and, using ML models, automatically determines a quality score for them all based on predetermined measures such as mandatory processes, agent efficiency and customer experience. An intuitive interactive dashboard provides users with an objective ranking of the quality of all their customer interactions and, thereby, allows them to take appropriate follow-up actions.

Identification of demand drivers

Another common scenario is the identification of demand drivers, which provides an unbiased view of why people are calling. The traditional means of doing this usually relies on agents recording a call reason–'disposition code'-at the end of each call. This is often fraught being prone to subjectivity, human error, laziness and, sometimes, abuse. Furthermore, it is often impossible to account for interactions with multiple call reasons or to spot newly emerging trends where a reason has not already been assigned.

TrueVoice addresses all these problems and automates the process. The resultant holistic and objective insight helps management to make informed decisions regarding, for example, the implementation of appropriate digital transformation strategies to optimize operations–helping to reduce costs while maintaining appropriate levels of customer service.

The bottom line

TrueVoice can deliver several key benefits by unearthing the rich actionable insight buried within customer interactions. It automatically monitors behavior and emotion on each and every call and is able to quickly highlight areas of potential risk and opportunities to improve outcomes. It can enable you to transform customer experience by helping you to optimize operations and deliver the most desirable outcomes.

You can also more effectively support your agents. Using a quality score system, agents that need support can learn from high-performers, and you can tailor your training to individual needs and develop customized best practices from actual learnings. A potential added benefit is that agent engagement can also be enhanced, increasing productivity and reducing stress and, hence, the likelihood of churn.

Bottom line, with TrueVoice, you can develop the insights to uncover what’s really driving the outcomes of your customer interactions, you can improve performance and deliver an excellent customer experience to enhance customer satisfaction. Satisfied customers become sticky customers, and sticky customers become valuable customers. Those satisfied, sticky, valuable customers can drive more revenue for you. That’s a win-win situation.

Support: TrueVoice@Deloitte.com

Website: www.deloitte.co.uk/truevoice

Additional information: TrueVoice videos

About Deloitte

Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as “Deloitte Global”) does not provide services to clients. In the United States, Deloitte refers to one or more of the US member firms of DTTL, their related entities that operate using the “Deloitte” name in the United States and their respective affiliates. Certain services may not be available to attest clients under the rules and regulations of public accounting. Please see www.deloitte.com/about to learn more about our global network of member firms.

water swirls

Interested in exploring more on cloud?

Site-within-site Navigation. Do not delete! This box/component contains JavaScript that is needed on this page. This message will not be visible when page is activated.