Research points to apprehension with speech analysis

Speech analysis can deliver real business benefits when correctly calibrated

Competitive pressures and increased emphasis on customer experience (CX) have transformed contact centre service providers into ‘’data hungry’’ organisms. Many have calibrated speech analysis tools to capture, internalise and gather value.

Speech analysis software examines conversations via phone, text, email, social media, and web chat. When used correctly, it can be a powerful business application in this data-driven world. However, there is apprehension among some contact centre operators regarding the effectiveness of speech analysis as a CX solution.

This post will investigate why speech analytics can be a beneficial business tool, why some operators have misgivings, and what needs to happen for it to deliver true value for those operators that are hesitant. It includes insights from a Connect iQ speech analysis survey. 

How is speech analysis beneficial?

Speech analysis is not new, emerging in the early 2000s, where its popularity has grown over the years as the software became more advanced and improved. It provides a multitude of benefits, particularly around operational optimisation and cost.

“Speech analysis identifies issues across all recorded communication events, not just a sample of calls or text, and thus identifies communication events that quality assurance (QA) would otherwise have missed,” says Janai Pillay, Operations Director, of Connect iQ. “Additionally, many of the manual QA and compliance processes are automated when the software is installed. Therefore, a higher volume of evaluations can be achieved.”

Pillay notes that more insights can be generated from the increased evaluations, exposing precise root causal factors for poor CX, customer churn, transfers, repeats and call backs. These insights can help improve and transform internal processes, resulting in improved average handle times and addressing issues before they are escalated to complaints. Some live calls can also be avoided through interactive voice response (IVR) technology. All these factors alleviate pressure off agents, as call volumes are controlled, boosting their productivity. As such, key performance indicators (KPIs) can be met, CX maximised and key metrics ticked, such as first call resolution (FCR) and improved net promoter scores.

These key delivery points are reiterated by industry players, including a respondent to the Connect iQ survey who said: “We use speech analytics to successfully identify and fish out complaints or calls that QA would not normally identify. We can then focus solely on dissatisfied client experiences and most times also identify calls where we had dropped the ball. We can close that gap before the client escalates it to complaints. It adds a “wow” factor to the client experience when you can identify these calls and improve the client relationship. Most of the time the application is not limiting your business but how you are using it, so find ways to make it work for you.”

An optimised internal process combined with the automation of manual tasks can cut costs significantly, as less is being spent on QA and monitoring. Issues can be nipped at the bud far sooner, reducing the number of required complaints handling specialists. Agent churn is minimised saving on training and recruitment costs. Also, speech analysis software has become affordable, quick and easy to deploy, saving on upfront installation costs.

Connect iQ’s survey results illustrates what enterprises use speech analysis for. The poll results below show that performance improvement (37%) is the top desired output of speech analysis among the surveyed operators. Customer experience (30%) and compliance (28%) are also among the top three motivators for speech analysis adoption.

Source: 2021 Connect iQ LinkedIn Speech Analysis Survey

Why is speech analysis not getting the nod from industry?

Despite these seemingly glaring positives, there still exists a reservation towards speech analysis among some contact centre operators. Connect iQ’s survey revealed that most operators are not satisfied by their speech analysis solution, pointing out the failure to effectively link with unique internal processes and a lack of human support as the main reasons. The graph below shows that most respondents (67%) indicated that speech analysis fell below their expectations. 

Source: 2021 Connect iQ LinkedIn Speech Analysis Survey

This has caused hesitancy in the industry to implement speech analysis. Although 43% of the surveyed contact centre operators stated a positive inclination towards using the technology, there is still a significant number that are sitting on the fence (13%) or have no intention of adopting speech analysis at all (17%). This is represented in the graph below.

Source: 2021 Connect iQ LinkedIn Speech Analysis Survey

How can operators make speech analysis work for them?

‘’Speech analysis software on its own is not the ‘silver bullet’ to optimising CX, compliance, and usable insights,” affirms Pillay. “It is thus crucial that speech analysis is complemented with structured, professional applications for it to be a success in a contact centre operation.”

A full-time analyst will be required to integrate the system to the internal processes and then manage the tool on an ongoing basis. Trends and reports can then be effectively derived, providing meaning and relevance to the data and adding value to the decision-making process. An action committee is needed to review the insights and then fix what is broken.

Pillay places emphasis on consistent changes of the speech analysis system: ‘’Efforts do not end there, as speech analysis must be constantly recalibrated parallel to changing customer preferences and operational structures,” he says. “New consumer trends impact consumer expectations affecting KPIs and thus measurements of quality. Morphing operating models also change internal processes. The speech analysis system needs to account for these developing elements to continue delivering optimal value to contact centre service providers.”

The business case is still clear

There is a clear business case for speech analysis solutions in the way it can optimise the QA and monitoring process, increase agent productivity and reduce costs. However, failure to acknowledge that the technology is not an end-all, one-size-fits-all, solution will guarantee a disappointing experience. Therefore, it is recommended to augment speech analysis technology with professional human resources who can integrate it into each unique operational structure and yield tangible value from the insights. Constant calibration is required to fit the ever-changing business environment.