The next big data source: speech analytics

Even though the speech analytics market is expected to see a compound annual growth rate of more than 18% through 2022, confusion still remains around its benefits and use cases – especially for South African organisations operating contact centres. 

Some of the confusion arises from a lack of education around the distinction between voice authentication and speech analytics. Voice authentication is predominantly used to reduce fraud risk and improve compliance. It is an automated method for identifying individuals based on measurable biological and behavioral characteristics, and you will see it most frequently used in the financial services industry. Physical traits (such as vocal tract dimensions) combined with harmonics and speech patterns create a unique digital pattern or voice print that is then used to identify a customer. Although there are other benefits (such as enhanced CX through quicker customer identification), this technology has been primarily used to combat the risk of identity theft through the contact centre.
Speech analytics, on the other hand, automatically identifies, groups, and organises words and phrases used during a voice call to reveal trends, opportunities, and areas for concern. In much the same way that humans can connect conceptually related words to create context, speech analytics automates this process to derive real insight from everyday conversations between customers and agents. While this can also reveal occurrences of non-compliance and fraud across phone interactions, the use of the technology extends beyond this application.

Considering that customers are speaking to businesses daily, speech analytics therefore represents a massive opportunity in the market. It empowers the organisation to have a better understanding of what the customer is telling them – and then uses that to improve engagement, operations and overall business strategies. 

Instant, relevant customer insight

Marketing departments spend a considerable amount of time and money on conducting primary market research to inform their strategies on pricing, products, advertising, branding and customer specific metrics such as satisfaction and loyalty. Unlike the lengthy data collection and analysis periods associated with these studies or the low completion rates of ad hoc customer surveys, speech analytics automatically collects and analyses this data to reveal instant, relevant and valuable customer insight.   

Insight could include how the marketplace perceives your offerings, competitive intelligence, identifying strengths and weaknesses with products and processes, and emerging trends or opportunities, amongst others. 

Employee performance and coaching

Speech analytics can also be used to identify the reasons for agent-related escalations, for example, and used as a foundation for employee coaching. By building a search on a phrase such as “want to talk to a supervisor” and letting concept-based understanding reveal the surrounding words and phrases, the analysis could reveal if the root cause is an agent-related issue such as a lack of knowledge, empathy or action. These calls can then be automatically classified with an “escalation” tag and flagged for follow up and coaching.

CX enhancements

First Call Resolution (FCR) can be difficult to track, but speech analytics can be used to monitor repeat callers by searching for related key phrases like “calling again”, “the previous agent” and “second time I’ve called”. Repeat callers are also more likely to be frustrated – using sentiment analysis can help isolate specific issues related to why the second call was required. By tracking issues related to FCR, a business can then identify areas for improvements such as scripting changes, knowledge base updates or even areas where more training is needed to improve service levels and the overall customer experience.  

Managing compliance

Another significant advantage of using the technology is that it can help ensure that businesses remain compliant. For example, in the case of FICA documents, there are certain phrases agents must use when talking to customers. Analytics can automatically check all these conversations in real-time and ensure the phrases are used. Traditionally, this was done manually by a QA team of people listening to conversations and checking for compliance – this technology can completely automate this process, significantly reducing costs and enhancing compliance.

Speech analytics can therefore be viewed as a highly flexible and powerful tool that has numerous applications and realises multiple benefits for businesses. 

Notably, however, the benefits of speech analytics do not get realised overnight. Instead, there is a steady progression in the maturity of how an organisation can understand, use and leverage it to derive maximum business value. From optimising day to day processes to enhancing CX journeys, the technology provides an ideal platform to proactively track and influence future business outcomes.