(844) 207-6663
August 4, 2020

Reducing risk with speech analytics: Five easy things you can do right away

Speech analytics is no longer a new thing. 

Having come on to the scene in around 2012, the varied use cases have since been well-documented. From improving customer experience and agent performance to providing business insights and managing risk, it can be a powerful and unique tool for those looking to take their contact center to the next level.

However, perceived difficulties and costs associated with using this technology have been a significant blocker to uptake. Often, vendors selling speech analytics solutions recommend employing an entire team to manage the tool, including data analysts, subject matter experts, and more.

This has clear negative implications for total cost of ownership considerations and thoughts surrounding the agility and speed with which businesses can use a speech analytics tool once they get it. This combination of factors has been off-putting for many contact center decision-makers, particularly for those with a relatively small headcount, or those that are already grappling with complex system integrations.

However, this picture is changing as some speech analytics vendors have taken the time to invest more in data accessibility, UI, and UX. Usability is the order of the day. 

Solutions are now available that are easy to use and can enable contact center managers to begin answering essential questions about their business and operations from day one, without the need for in-house technical expertise. 

Choose the right tool, and with minimal heavy lifting, you can start to reap the benefits of speech analytics from day one.

But where to start? The applications are almost limitless. But, with the number of TCPA lawsuits filed up 1,279% in the last decade and customer experience playing a more important role than ever before, risk mitigation might be the right place to start.

Here are five easy ways you can start using speech analytics on day one to help you avoid some prevalent regulatory and business risks.

1. Bake compliance terms and critical business requirements into the first scorecards that you create

In the long run, speech analytics will radically improve your quality management capabilities. The ability to create scorecards that are then automatically filled by AI for every interaction taking place in your contact center is transformative in terms of both scale and objectivity.

However, you won’t be able to create the perfect scorecards overnight. It will be an iterative process that requires ongoing conversational, performance, and operational analyses, and will change as your business requirements do.

So start simple, with the compliance and business-critical stuff. It should be relatively easy to do as much of this is grounded in the use of specific language, like compulsory legal statements, so lends itself well to speech recognition technology.

By setting up targeted keyword lists, you can ensure agents are:

  • Delivering legal statements 
    • Mini-miranda
    • Recording disclaimer
    • FDCPA/CFPB and FCA requirements for credit discussions.
  • Adhering to policy 
    • Script adherence
    • PCI conformity.
  • Following brand standards
    • Proper greetings and call closing
    • Polite language.

Next steps: 

Once you have essential business requirements integrated into your automated scorecards, start to think about agent performance. Using the reporting available in your speech analytics tool, you can analyze what conversational factors (words, behaviors, techniques etc.) are driving call success, whether that be conversions, resolutions, or whatever the most relevant metric is. You can coach agents on these things, and include them in your scorecards. Download this case study to see how BPO, The Office Gurus, did this to drive up the conversion rate on a sales campaign they ran for a major retail partner.

2. Set up notifications for the highest risk behaviors

Once you have the most critical requirements fully baked into your automated scorecards, you will have an accurate and full picture of how these are being adhered to by your agents. You’ll also be able to monitor trends over time, and see how coaching, or any other initiatives you implement to make improvements, are having an effect.

But, for some things the potential legal and financial implications are too significant for you not to be notified as soon as possible. 

For this reason, good speech analytics solutions will allow you to set up custom notifications. It works in the same way as quality management automation, but instead of specific keywords (or lack thereof) filling in a scorecard, it triggers a notification. This could be an email or a text and will automatically distribute to all relevant parties in your organization so the appropriate party can take action.

Next steps:

Think about using contextual analysis to get to try to understand what is causing high-risk incidents. This means using speech analytics to analyze the words and phrases surrounding or leading up to problematic agent behavior. This will allow you to go from being purely reactive to being able to do something proactively to improve future standards.

3. Flag bad customer experiences fast

Customer experience is the critical differentiator in today’s business environment and your  contact center, one of the main touchpoints a consumer has with your business, plays a crucial role.

Without speech analytics, it’s estimated that 98% of contact center interactions go unmonitored and unreviewed. Therefore, the vast majority of incidents of bad customer experience are likely to fall through the cracks. 

As soon as you get a speech analytics tool, start filtering these occurrences out – and then do something about it.

There are certainly some keywords that will help to indicate a bad experience (many of them not repeatable here!) but sentiment analysis is also going to be important. This technology uses advanced AI to assign a sentiment score to every interaction in your contact center – split by agent and customer. Combining these will exponentially increase your ability to catch and follow up on these bad experiences.

Who knows? With the right response, you might even be able to turn these situations around to create an overall positive experience where initially there was a negative one.

Next steps:

Take your understanding of customer experience a step further by using speech analytics to improve the content and targeting of your CSAT (customer satisfaction) surveys. Knowing the topic of conversation and content of every call will allow you to send targeted feedback surveys to specific customers. These can reduce customer effort by only asking relevant questions, increasing completion rates, and uncovering more in-depth information than generic one-size-fits-all approaches.

4. Identify urgent coaching needs

Compliance breaches, escalations, and bad customer experiences – all of these incidents ultimately tie back to a conversation that was happening between the customer and one of your agents. 

Speech analytics allows you to get the full picture of agent performance because it assesses 100% of interactions entirely objectively. This means that in just a short amount of time, you’ll have a significant amount of data that you can use to assess agent performance from an individual agent level to a whole team or your entire contact center. 

So first off, prioritize the highest risk agent performance outcomes (which we have already identified in points 1-3), look at trends, and determine which agents require specific coaching in certain areas.

Next Steps:

Use speech analytics data directly, to coach your agents on best practices. Our partner, TheOfficeGurus, has found great success in using aggregated conversational data to show agents what language and techniques drive the most success in a range of things from driving down average handle time to increasing sales conversion rates. They’ve found it to be a very potent tool with which to encourage certain agent behaviors.

5. Go beyond voice

Our final piece of advice is to drive home the point that speech analytics is not just about speech. In fact, the name could distract people from the full potential of the technology.

Speech analytics works by using AI to transcribe speech to text, and then perform sophisticated analysis on said transcribed text.

So the second part, the analysis, can be applied to any text. That means your SMS, webchat, and email interactions can all be better understood as well. 

As the importance of these channels continues to grow, so does the importance of accompanying risk mitigation efforts. So make sure that you are uploading these transcripts to your speech analytics tool from day one. They will likely require their own automated scorecards. So, as we outlined in point one, start simple and build.

If you want to avoid a lot of unnecessary effort, try to make sure that any chosen partner will support an API to allow you to upload these, and your voice recordings, automatically each day.

Next steps:

Once you are scoring all of your omnichannel interactions you are very quickly going to gather a lot of valuable data. Even on its own, this is going to be transformative for your quality management efforts. However, even greater benefits can be attained if you are able to combine this with your wider reporting and analytics processes. Think about whether your current reporting and analytics allow you to effectively overlay different types of data from around your business, like QM scores, operational data, and business outcomes. This is what is going to give you a truly 360-degree view of your operation.

So that’s all from us for now. We hope we have demonstrated that adopting speech analytics doesn’t have to be another lengthy integration project that takes many months to reap any rewards. If you start simple and choose a user-friendly tool, there’s no reason not to hit the ground running and start reducing risk from day one. And then, there are a plethora of different paths you can take as you look to flex your muscles and get more out of a tool.

About LiveVox

LiveVox is a next-generation contact center platform that powers more than 14 Billion interactions a year. We seamlessly integrate omnichannel communications, CRM, and WFO capabilities to deliver an exceptional agent and customer experience, while reducing compliance risk.  Our reliable, easy-to-use technology enables effective engagement strategies on communication channels of choice to drive performance in your contact center. Our battle-tested risk mitigation and security tools help clients maximize their potential in an ever-changing business environment.  With 20 years of pure cloud expertise LiveVox is at the forefront of cloud contact center innovation. Our more than 450 global employees are headquartered in San Francisco; with offices in Atlanta, Denver, New York City, St. Louis, Medellin, Colombia, and Bangalore, India.

Share this story

Keep Up-To-Date on the Latest Contact Center News

Subscribe to our newsletter and stay current on all the latest technological advances in the contact center space.

About LiveVox

LiveVox is the only one-stop-shop for true omnichannel engagement that unifies modern channels, CRM, and WFO functionality into a single cloud customer engagement platform. Facilitating over 14B interactions annually, LiveVox makes omnichannel easy by unifying all conversations and interactions in one place. Founded in 2000, LiveVox is headquartered in San Francisco with offices in Atlanta, Denver, St. Louis, Colombia, and Bangalore.

To learn more, visit www.livevox.com or call one of our specialists at (844) 207-6663.

You May Also Like