When you research speech analytics features, chances are you’ll run into the phrase “real-time analytics.” As an AI-based speech analytics solution, these analytics are capable of analyzing conversations in real-time. When a call center agent is conversing with a customer, this technology has the ability to keep pace with the interaction.
How it works
Oftentimes, the use case required for such a swift response is screen pops. A screen pop is when a system flashes a recommendation to a representative while they are speaking with their customer. These prompts can encourage the rep to offer promos, remind them to thank the customer for their loyalty, suggest offering a specific service option or prompt them to use a more upbeat, positive tone.
With real-time speech analytics, the purpose is to drive customer conversations toward better outcomes. Depending on the interaction and the company’s personal focuses, this might be a lower effort experience, a higher sales conversion, a stronger FCR (first-time call resolution) rate or shorter calls.
Screen pops are enticing, but weak
Screen pops might appear to solve issues related to service, sales and compliance effectiveness, but do they work? Agents often fail to pick up on customer cues that end up costing businesses a sale or lowering the quality of the customer experience. They also sometimes neglect to read the company-mandated disclosure statements. It would seem that real-time screen pops would address these problems swimmingly.
However, flashing things in front of agents is not a sustainable way to encourage behavior change. The assistance from these screen pops by reps can be initially distracting and eventually tuned out as white noise. A strong coaching program is more effective than screen pops on any day.
The problem with real-time analytics is that they can only offer agents and organizations a band-aid solution to a bigger problem.
Real-time analytics don’t keep things real
The best agents are empowered to go with their instincts. In order to develop a rep to their full potential, they must be encouraged to think on their feet. This type of coaching inspires agent confidence and decreases powerless to help behaviors that incite high-effort experiences.
Conventional wisdom suggests that the goal of a rep is to immediately address and correct a problem as quickly as possible. This leads to the very temporary solution of placating customers who will likely be displeased later on as agents who weren’t thorough only focused on the issue at hand.
Increase initial effort for a low-effort experience
Tethr has found that the best low-effort calls are where an agent actually asks a lot of questions before fixing the problem. Initially, there might be more effort during the call, but since confusions would already be addressed early on, the effort levels would go down, creating a “U” shape.
Consider this, you walk a customer through an issue, resolve their problem and wrap up the call. While you might think the call is done, it might be far from over.
Some organizations are often under the impression they’re performing well because they have strong FCR scores. However, repeat calls often indicate downstream issues related to the problem that prompted the original call, even if that problem appeared to be adequately handled the first time around. Companies must anticipate and forward-resolve these issues.
Mining customer interaction data enhances CX
By mining customer interaction data, companies can better recognize the relationships among various customer issues. These valuable insights can help agents to resolve the customer’s primary issues, while anticipating downstream issues. This is called Next Issue Avoidance (NIA).
According to Tethr’s Chief Product & Research Officer Matt Dixon’s “The Effortless Experience,” when a customer has to recontact the company to chase down the resolution of their issue, it increases the likelihood that the customer will be disloyal at the end of the service experience 2.5X.
Near real-time speech analytics dives deeper than a quick fix
While in areas like compliance, real-time fraud detection can be very valuable, alerting a rep to recover an already frustrated customer is not as valuable as using AI to determine the root cause of customer aggravation and eliminating those at the source via targeted training and coaching. For some businesses, near real-time means within a business day. For others, it might mean minutes or hours later.
Customers can use Tethr to find out the root cause between a spike of customer calls within an hour. Although real-time speech analytics technology isn’t applied here, correcting the problem within the hour is the most effective route toward overall customer effort reduction.