Is your platform acting as big brother, collecting dust, or something more robust?

In today’s compliant centric credit and collection world, more and more agencies are turning to speech analytics platforms to help solve their compliance needs. These robust software offerings help agencies find and track issues much quicker than any internally staffed quality assurance team. Many agencies are deploying the software to act as the proverbial big brother as it is constantly mining for compliance infractions. Unfortunately, in many cases, this is the extent in which the system is utilized. Additionally, in some instances after initial set up, many think the system follows a ‘set it and forget it’ model. In this case, the investment is merely collecting dust. Speech analytics users should consider the following three areas so that your platform is so much more robust than just playing the role of Quality Assurance: understand the pitfalls, solve for best practices and make data driven decisions.

First and foremost, it is imperative to understand the pitfalls of the system you purchased. An off the shelf system, complete with canned searches, reports and scorecards, can provide an immediate helping hand to an already taxed quality assurance department. However, these systems have a fair amount of inherent inaccuracies and false positives when it comes to the mined calls and subsequent agent scorecards. These quality issues stem from poor audio quality, systemic issues of identifying tone and sarcasm, and recalibrations due to accents or rate of speech. In speaking with Kyle Carter, VP of Creative Solutions for Zenylitics, he suggested that while these pitfalls are very real, it is imperative that the issues be understood and accepted by the entire organization to resolve. Some solutions to these findings can be as simple as ensuring dual audio recording ability. However, determining false positives and tonality issues can be much more difficult. Random audits and constant recalibrations are a must when managing these platforms. Mr. Carter also mentioned that “because of the power of speech analytics, it is very easy for SA teams to get in the weeds,” and “that scope creep is very real and dangerous to the success of SA initiatives.”

Outside of the need to understand the pitfalls of a speech platform, it is also imperative that users reach out to learn best practices in the industry. Again, it is quite simple to make the SA software play the role of big brother. However, fully grasping the possibilities can lead to not only enhanced QA and compliance deliverables, but also become a viable tool to drive revenue in a collection shop. For example, an obvious lift comes from a decrease in compliance violations. If the end users, operators and management are aligned and scorecard build reduce compliance errors, by default one would think that the expense relative to violations will be less. However, imagine the possibilities of scorecards or mined data that can replicate the top agent’s best practices. What about the ability to look for specific language that is universal in driving performance? A few years back, an agency was struggling with expected production during tax season. Speech analytics software was used to create a scorecard to see how many agents and at what consistency where they are utilizing tax time verbiage in their demands. The results were surprising; less than 40% of the agent base were using tax return language when demanding money. With this script, adherence drastically increased, and subsequently, so did collections. If SA platforms are used as big brother only, agents and end users won’t be engaged. If the system is used to decrease training time and increase revenue, correct behaviors will be consistently reinforced and best practices will become more universal across the business.

With the amount of data the industry now has available, the ‘gut’ feeling can now actually be displayed numerically and subsequent decisions can be made not only effectively but with measurable and exact results. A speech analytics platform increases the data ten-fold, and can turn “anectodal” into “actionable”. A typical user may consider the software so that complaint reporting can be achieved for compliance needs. However, taking that a step further and the how many can turn into a why are they and a root cause can be performed at scale on all calls. By identifying the why relative to complaints and dissatisfaction drivers, an organization can build models to ‘look’ for behaviors of either agents or callers that can help eliminate the complaints and not just report on them. Couple that information with a strong marketing or mail campaign and potentially can proactively engage the consumers likely to complain before they do so. With today’s platforms, data is virtually unlimited and with that amount of information, companies are only bound by their creativity.

In speaking with Mr. Carter he said something that should resonate with anyone that has already is looking to purchase speech analytics software. “Speech analytics can be transformative to an organization if deployed appropriately. Understanding the interactions between customer and agent can provide unparalleled insight into process inefficiencies and consumer pain points, and can also provide the roadmap for solutions to those problems. Today’s consumer is searching for a frictionless experience, and analyzing our interactions can help companies understand how to create that for them.” It is imperative that owners of SA software avoid pitfalls that can cause the system to be a waste of time and money, determine best practices that can reduce costs and drive revenue, and use data to enhance the customer experience while increasing the effectiveness of the company.

Josh Seuberling is the Chief Operating Officer for WayBeyondRewards, a company that provides customers with nationwide access to thousands of dollars in benefits and savings, all in one place.

Kyle Carter is the Vice President of Creative Solutions for Zenylitics, a company founded to solve the most significant challenges with quality assurance programs by using a combination of AI powered speech analytics with professional quality assurance agents.