Unlocking the Value of Voice Data


By Alex Fleming

Over the past few years, the contact center market has created more processes that have moved away from voice. The omnichannel direction has put a significant focus on using elements that are primarily text-based. This reflects the way end users, and we as individuals, now communicate. Instant messaging, email, and social media have grown in popularity, and contact centers have been quick to adopt these channels to engage with customers.

They are perfect for engaging in a relevant way, while keeping the highest level of quality text, where the data is easy to search, index, and store. This means that contact centers have a wealth of data to use to gain valuable insights into their business and develop more tools to further address end-user requests or requirements.

While omnichannel has enabled contact centers to remain relevant, the focus is shifting, with speech and voice calls providing the richest density of data. A recent report from Red Box highlighted that artificial intelligence and machine learning suddenly make voice data available in volume, when previously it would only have been accessible by listening to individual recordings. This presents organizations with an opportunity to tap a rich data set that can help drive true and measurable business outcomes.

How Do You Unlock Value? 

Due to the increased interaction of end users—disclosing personal information and even payment details—the contact center market is highly regulated. The regulation is in part why contact centers record calls to ensure that they remain on the right side of compliance. This means that contact centers are sitting on a pot of gold that has built up over the years regarding customer insight and engagement, as well as training information. They also have the constant influx of additional calls every minute, hour, week, and month, so the stockpile of data keeps growing. 

Automatic speech recognition (ASR) services enable this backlog of data to be transformed, and the processing of prerecorded calls means that historical data can be transformed into a text-based format. In addition, the ability to transcribe in real time means that agents are empowered with tools that can adapt and change depending on the conversation. This solution has the added benefit of elements such as sentiment, meaning, context, and nuances within the language that add a greater value to the text output.

The Right Data at the Right Time 

Contact centers are always looking to provide the highest levels of customer service to minimize customer churn and accelerate the resolution of issues. Speech-to-text technology allows them to conform to compliance and regulation easily, while also supporting their agents by increasing their knowledge and eliminating frustration.

The value and benefit of transcription is that the information can be available right away. Even when transcribing the call content after completion, the information is accessible soon after the conversation has ended. This enables situations such as issue resolution to be accelerated for quality assurance managers to use directly. This eliminates the time and effort required to review the call’s audio. Using text also means that the content of the call history can be searched more easily, with the added value of enhancing the customer experience even in a dispute situation. 

Cases Requiring Understanding and Resolution

When discussing customers and their issues, sometimes a simple tool can be the most powerful. Data gained through transcripts can be searched and used to generate trends of frequent issues. It might be that these issues can be solved through the addition of a new section to an FAQ page, a new response in a text bot, or even the creation of a self-service portal. 

The ability to empower customers in this way not only means that their issues can be solved, but it can be done with minimum impact on contact center staff. This frees them up to handle customer issues that need human interaction. However, this solution is only possible through the ability to make the content of calls available and make intelligent decisions from that.

At a deeper level, sentiment analysis can help at a one-to-one level for the agent to better understand the caller. If agents have methods on hand to gauge a customer’s mood, it can make the difference between a successful call outcome or losing that customer. 

Say the Right Things

The ability to transcribe the content of a call enables the identification and use of keywords to autonomously trigger actions within the agent’s dashboard in real time. Using ASR services in real time can significantly impact the service that new agents, even without experience, can offer to customers. 

Know If a Customer Had a Good Experience

Core elements for agents can be enhanced by what ASR can offer. These include the ability to capture conversations in a text-based format for faster review, triggering information to be at an agent’s fingertips and extracting insight for integration into analysis products so the agent can dedicate more time to issues that require human attention. Agents are then better equipped to provide higher quality service to customers and feel that they are doing their best job. With these tools, the agent feels that they have had a positive impact on customer interactions.

Alex Fleming is products marketing manager at speech-to-text technology specialist Speechmatics.