By Chaitanya Chokkareddy
Remember when you last lost your cool with a customer service department? How promptly did a manager rush in to rescue the conversation? Possibly, it took a few minutes while you waited (angrily) on hold, and a panicked agent managed to alert his boss. Or maybe you received a call back a few hours, or days later. Or quite as likely, no manager ever heard the angry snafu, and it took you multiple calls to get any redressal.
“This call shall be recorded for quality and training purposes.” We hear this line play before every call center interaction. But did you know that only 10% of these recordings are ever heard by quality assurance teams? Even fewer reach the ears of concerned managers and departments. Thanks to the new AI-based tool, this is set to change.
AI-based speech analytics is a new tool that can give managers real time data and alerts from their ongoing call center conversations. This data includes live tracking of customer sentiment, conversation quality or specific keyword usage. In this way, call center recordings can be used for much more than “quality and training purposes”.
• No angry conversation will go unnoticed. Managers will monitor all conversations via live dashboards and diffuse angry conversations faster.
• Agents’ speech rate or tone can be live monitored and this data can be used to improve performance.
• Marketing departments can access live feeds from their call centers on customer sentiment, customer complaints.
In short, it is an invaluable new tool to improve the quality of customer service and allow businesses to glean data from their call center.
So, what is Speech Analytics?
Speech Analytics is the process of analyzing any voice recording using AI-based tools. These tools or algorithms process information in a matter of seconds. As a result, ongoing calls can be analyzed in real-time. The analysis can be displayed on various live dashboards or sent as alerts to relevant teams.
What insights do Speech Analytics currently reveal?
Some of the common metrics that plug and play Speech Analytics Dashboards display include agents’ speech rate, speech volume, and customer sentiment on each call.
• Speech Rate: Clarity and enunciation play a key role in voice communication. According to studies, the ideal talking speed is between 120 to 150 words per minute (wpm). This rate varies across just across languages and dialects, and can also be affected by gender, culture, and purpose. A speech rate dashboard is a good training tool for customer service teams. Managers and agents can view their speech rates, and course correct if it becomes too fast or slow.
• Speech Volume: To best address a customer’s query, an agent must be audible. But more importantly, the volume at which an agent or customer is speaking can indicate how the call is going. For instance, a raised voice may indicate conflict. With speech analytics, businesses can set up alerts for supervisors to barge-in and take over calls whenever either an agent or a customer raises their voice.
• Sentiment: AI-based analytics can automatically gauge customer sentiment for the duration of a call. For example, some Speech Analytics dashboards categorize customer sentiment into happy, neutral or angry/dissatisfied. This helps managers keep a tab of overall call center performance, as well as individual agent performance.
Speech analytics are also used for keyword identification. For example, you can run a keyword search across all your call recordings for the usage of the word “sorry” or “not working” to identify service or even product-related issues with ease. Businesses can create customised dashboards for this type of keyword analysis.
How will this benefit businesses and customers?
1. Better Feedback. Traditional feedback methods are useful, but since all angry customers do not complete surveys, they give businesses an incomplete picture. By analyzing every call for anger or dissatisfaction, speech analytics gives a more accurate view of customer happiness.
2. Improved Service Quality. It is humanly impossible to go through every single call recording to analyze sentiment or agent performance. This means that traditional Quality Assessment cannot fully assess agent performance leading to poorer training outcomes. In contrast, a speech analytics tool can assess performance across every single call in seconds. This enables a more thorough assessment, faster interventions, and better training.
3. Faster Redressal. By alerting managers in real-time, Speech Analytics tools can help call centers keep a pulse on customer sentiment and address any dissatisfaction much faster.
In the future, businesses may use speech analytics to gamify call center representatives’ dashboards. The marketing department may use this data to keep a pulse on customer sentiment or to assess feedback on certain products or campaigns. The possibilities are endless. Overall the biggest benefit of Speech Analytics will be to tap into the vast data that call centers gather, leading to better services, and much happier customers.