Using Speech Analytics to Improve Customer Relationships in the Call Center



By Jeff Gallino

Here are answers to some questions I often hear about call center speech analytics.

Question: What’s the biggest trend you’ve seen when it comes to managing customer relationships, and how has that evolved over the past five to ten years? 

Answer: With the rise of digital transformation, we’ve seen a major shift in consumer preferences, and customer experience (CX) has arguably replaced traditional marketing in terms of becoming the number-one brand differentiator. More companies are realigning their business strategy to focus on delivering exceptional CX. This is because customer demands for personalization, innovation, and consistency are on the rise. 

As a result, it’s never been more critical for companies to understand the voice of their customer—not only within the C-suite but across all levels of their organization. Sure, companies have more opportunities to connect using email, live chat, social media, and other channels. But we’re seeing many organizations spread themselves too thin, thereby actually doing more harm than good. While it’s critical in today’s digital and omnichannel world to stay agile, companies must remain focused on what is truly important: building and nurturing customer relationships. 

Q: What are the main factors that lead to customer churn? 

A: Resulting in an estimated deficit of 136 billion dollars per year, customer churn is a sore spot for most US businesses. The good news is that it’s entirely avoidable. 

Speech analytics technology provides a holistic view into 100 percent of customer interactions. Click To Tweet

From an agent standpoint, the answer is much more fundamental than one may think—40 percent of consumers just want someone to listen to them. However, many agents fail to deliver on that simple request, as only 23 percent of callers feel listened to. Another major factor when it comes to delivering an unsatisfactory experience is the agent’s inability to show empathy. 

The emotional state of the caller at the beginning of the call is a tough situation for agents—regardless of their level of emotional intelligence—as 42 percent of callers arrive annoyed, 28 percent arrive upset, 22 percent arrived confused, and 17 percent arrive angry. Yet despite these strong emotional tendencies, 43 percent still express the need for a more polite and caring representative.

As for as the role played by the call center itself, sticking to the status quo just isn’t working anymore. Whether outdated scripts, ill-advised metrics, or a lack of knowledge in general, legacy call center management methods aren’t up to speed with the demands of today’s consumer base. 

Customers not only expect solutions to their problems, they want solutions in a timely manner with minimal disruption. When management fails to arm agents with the tools, training, and resources necessary to do their job, they’re only contributing to the problem. 

Q: What kind of insights does speech analytics offer? 

A: There’s no one-size-fits all definition for speech or engagement analytics, but at its core, speech analytics is a way to generate insights from conversations. But it’s much more than a transcription tool. It automates the process of listening to customer interactions, regardless of the communication channel used. This allows agents to make accurate and informed decisions based on customers’ needs. 

Speech analytics takes the unstructured data directly from both ends of the conversation and turns it into structured data. Once in this format, sophisticated categorization and tagging methods allow for searching and analyzing of information. These are all built in accordance with an enterprise’s unique business objectives. While this sounds like a lot, it streamlines the process and offers speed to actionable insights, which random call sampling would have missed. 

Take sales effectiveness, for example, and think of the behaviors of your most successful sales agents. This technology can create scores leveraging those behaviors, transfer that into historically proven, interactive suggestions, and automatically provide feedback and guidance to agents, which eventually raises top-line revenue.

Q: How do agents benefit from this technology in real-time versus post-interaction? 

A: Whether it’s on the back end for management purposes and corrective action, or on the front lines of the conversation to improve accuracy and compliance the first time around, agents benefit from speech analytics by receiving timely guidance and accurate feedback to help them improve performance. This results in increased first-call resolution rates, shorter average handle time, and increased customer satisfaction. 

For example, say a customer is becoming increasingly frustrated with the agent’s responses, or lack thereof. Based on preconfigured verbal or acoustic measures, agents can receive alerts to take a suggested action, such as an offer to save a customer from churning, or be immediately redirected to a supervisor to correct the problem. Companies can also track the customer journey across channels, so consumers don’t have to repeat the problem. This allows for a more personalized interaction.

When used for post-interaction analysis, speech analytics technology can track trends and make root cause discoveries to improve contact center performance and business intelligence across the enterprise. This is a direct result of agent’s receiving feedback on 100 percent of their calls. This gives them praise or coaching on performance to make them more equipped to address similar situations in the future.

Q: What is the difference between speech analytics versus traditional customer feedback methods? 

A: Surveys and net promoter scores (NPS) face limits by the number of responses they receive, accumulating a response rate of anywhere between 5 to 15 percent. Therefore, they’re limited regarding the story they tell. This is mostly because they will never be able to show you exactly what’s going wrong and how to fix it. Sure, they offer a glimpse into customer satisfaction, but they only skim the surface. 

Simply put, you can’t improve NPS and overall CX without first understanding the root cause of the problem. In addition, these feedback methods only offer a glimpse of the customer’s perception of your brand. You can’t rely on these methods alone to revamp your CX strategy. However, speech analytics technology provides a holistic view into 100 percent of customer interactions. This allows for a consistent, accurate analysis of a variety of performance indicators such as effort and empathy.

Jeff Gallino is the CTO and founder of CallMiner, a software company that develops speech analytics software. Founded in 2002, it’s headquartered in Waltham, Massachusetts.