5 Speech Analytics To-Dos to Maximize Efficiency and Effectiveness of The Solution

Today, speech analytics (SA) is a critical part of a successful collection-focused contact center.  It helps ensure compliance with rules and regulations and empowers you to train (and coach) collectors to be more effective. In the long term, it can even help you keep your clients happy . This explains why (as per Account Recovery’s survey on speech analytics), two-thirds of the companies in the ARM space appear to be using speech analytics in some shape or form.

Having said that, speech analytics is only as good as how you use it. You cannot just implement it and assume it will address all of your problems in the first go. However, if you invest quality time in its implementation and ongoing hygiene check, speech analytics can help you yield the results you are looking for. To discuss this and more, Mike Gibb, CEO, and Editor of AccountsRecovery.net hosted a webinar. This webinar, which was also sponsored by Provana, discussed how speech analytics can act as the glue to illuminate key performance indicators (KPIs) across modern collection operations and set them up for success. If you missed attending the webinar, here are key takeaways.

1. Stay laser-focused on your biggest pain point

The people and the companies who are least effective at implementing SA are the ones that ignore stepwise implementation. Ideally, a speech platform from day one should make your QA process effective. And that can only happen when you follow a consultative approach to its adoption in your business. Sure, every business wants to improve their QA effectiveness, reduce headcount in the long run, or increase collections output. However, your implementation process should be based on your key pain point at the time of tech adoption. Compliance is the core reason why most companies must start to adopt speech in the first place and build other objectives along the way.

Fun fact: As per Account Recovery’s recent survey on speech analytics, SA adoption is proportional to the company size. The mid-sized companies’ adoption rates of speech analytics hover in the 30 to 35 percent range as per the survey. When you look at companies with a hundred seats, the adoption rates drop off significantly.

2. Build a wholesome speech team in house

To make a speech program successful, a lot of different skill sets have to come together. Hence, when you implement the solution, ask for inputs from stakeholders from operations, compliance, training, etc. You need an Ops person paying attention to what to do with data. A data scientist can do a deeper dive into the second, third, or even fourth layer of data – something extremely important for a long-run, successful speech program.

Pro-tip: To avoid false positives or calls being flagged due to poor definition, make sure you clean your data from the get-go and validate/re-validate it periodically.

3. Think beyond conventional sentiment analysis

It is always easy to label a call good or bad based on certain phrases or words. However, assessing the tone to conclude a call as a good or bad consumer experience is always tricky. This is where the actual context of the call can help you.

When your QA team assesses or analyzes the tone of the call, ask them to look for agitation. You can also check for common phrases that are used in expressing dissatisfaction, or escalation. At present, no system can catch the sarcasm. The actual context of the conversation and what you’re flagging the call for (coaching opportunities or compliance violations) can help you classify such calls properly. 

4. Put calibration through constant iteration

Calibration is not only a painful but painstaking process that you need to do right to get the ball rolling.  A small percentage of calibration and iteration is a continuous process that’s always going to happen. It will only help you make sure that you don’t have any drift between your expectations and how the system or your QA team is scoring your calls. Whenever you bring new QA agents into the system, as a rule of thumb, recalibrate and understand how they’re scoring their calls.

 5. Set ROI benchmarks to assess the performance of your speech program

If you are wondering, how speech analytics have saved you X amount of dollars, some metrics can help you identify the effectiveness of the implementation. SA helps in the reduction of compliance fines that could otherwise cost you a fortune. It can help you with agent productivity improvements or better collections on each RPC. All of this can be quantified and traced back to ROI improvement after implementing speech analytics into your business. Comparing your NPS before and after the SA implementation can also give you a sense of the effectiveness of the technology.

You can now watch the full webinar recording here.

Want to automatically monitor and objectively score your calls?

ICAP is an advanced speech platform that gives collection players control over their process-intensive operations, including 100% call monitoring, quality assurance, and agent performance management. It is built on Provana’s decades of experience in the ARM industry and human-like machine-learning capabilities. It helps collection businesses manage agent-consumer interactions effortlessly, converts unstructured data into meaningful insights, and ensures on-call compliance. Keen to learn how our technology can transform your call center? Click here to learn more.