Contact centers have been slowly restructuring aging and outdated voice channels to adjust for the mobile device/smartphone era and advances in artificial intelligence (AI), machine learning (ML), natural language understanding, and other enabling technologies.
While many enterprise IT departments already had plans in place to do this, only a few were executing it as a priority prior to the COVID-19 pandemic.
This has changed now. With the investments heavyweights like Amazon, Google, and Apple have made in AI, ML, and speech technologies, along with much improved acceptance and adoption rates for speech among the general public, it may soon be time for the self-service voice channel to finally fulfill its long-awaited promise.
To Start: Baselining What You Have
Planning and executing a digital transformation strategy that gets your enterprise to where it needs to be in the 21st century can be quite daunting.
Navigating away from traditional business models that are inflexible and outdated takes time - and a great deal of caution - if you are to reach that wonderful, flexible, agile nirvana that is ready to take on whatever the digital revolution throws your way in the years and decades ahead.
But it’s a journey well worth the effort, as anyone already in that adaptable, scalable, and future-ready environment will quickly tell you.
Whether you are already well on your way with a well-thought-out transformation strategy, or just beginning your journey, it’s never a bad time to take your bearings and carefully consider your next step...
To illustrate the importance of getting this right, consider that in 1958, according to the American Enterprise Foundation, U.S. corporations remained on the S&P index for an average of 61 years. By 2011, that number was down to only 18 years.Today, companies are being replaced on the S&P about every two weeks.
For the most part, technology has driven this shift. Companies that want to succeed in today’s business climate need to understand how to apply a solid strategy to harness the rapid pace of technological change.
...it may soon be time for the self-service voice channel to finally fulfill its long-awaited promise.
Unfortunately, for many bold new explorations and challenges, we are often keen to “just get on with it.” We tend to fear we are going to be left to wither on the vine like so many other corporate giants that did not see the writing on the wall until it was too late.
But before you rush off into the digital abyss with unbound enthusiasm and respect for all of those shiny new gadgets and services just dying to sign you up (or shake you down), take a moment to consider those early steps more carefully. I think it will serve you well in the long run.
Growth of Voice Self-Service
Many organizations, including those supporting the financial and government sectors, retail shopping, utilities, and healthcare, to name just a few, are now dealing with massive increases in call volumes, website visitors, and mobile app and chat sessions.
And as digital transformation accelerates and moving to cloud, or at least hybrid computing environments, becomes less of a future plan and more of a matter of necessity for many contact centers, the future of the voice channel as a means of efficient, cost-effective self-service looks brighter.
Some of the vendors supplying the contact center market and infrastructure needed to support these increases have already seen record increases in technology sales and the simultaneous need to innovate even faster than in the pre-pandemic era.
Common innovation themes and strategies include:
- AI-powered conversational interfaces.
- AI/ML informed workforce and advanced robotic process automation (RPA).
- Replacing/supplementing traditional IVR with intelligent voice assistants (IVAs) and chatbots.
- Rapid deployment of cloud-based contact centers.
And the pandemic has accelerated the need to adapt and adopt much faster: from five to seven years down to months and even days in many cases.
According to Statista, the IVA and virtual digital assistant (VDA) markets were already projected to grow substantially prior to the COVID crisis. This rate of growth will undoubtedly accelerate in the wake of the crisis and changing consumer habits.
A look at the pre-COVID projections for application growth in this area shows a similar trend (See Chart 1):
Consumer and user habits are already very well positioned to take off from the levels seen before the pandemic. We are way past the early adopter phases here and the technology is now well situated to become the norm, as opposed to the next new thing (See Image 1)...
Consider a Medical Analogy…
So we’ve established that, even in the face of text-based digital self-service applications and the continued strong popularity of web self-service, there is considerable demand for voice self-service applications.
But that doesn’t mean that the organizations that use it can rest on their laurels. Voice self-service technologies, even the best ones, can also have issues, such as, but not limited to:
- Poorly worded or ambiguously worded voice prompts.
- Unintelligible words in the recorded audio.
- Functions supported in the IVR like address updates and payments/transferring funds that would be better handled by text or web.
- Prompts that place too much of a cognitive burden on the callers.
- Prompts that request DTMF input when speech would be far more effective (and vice versa).
- Response timeouts that need to be longer.
And many other sources of caller frustration.
Let’s look at our health as an analogy. Physicians know well that an effective treatment of what ails us depends in very large part on a thorough diagnosis of the problem. So they test, analyze, measure, weigh, gauge, and interrogate.
And when it comes to the vital organs in particular, a good physician will always insist on accurate measurements of heart rate, blood pressure, and other vital signs.
In the same way, you need to understand what is happening with your customers, who are your lifeblood, and with the technologies – like voice self-service – that enable their engagements with you.
Voice self-service technologies, even the best ones, can also have issues...
This kind of thorough analysis is also required when it comes to finding the best way forward for your voice self-service channels as you work your way through your contact center modernization initiatives.
And it is in those very same voice channels - the ones that continue to handle thousands of calls on a daily basis from your customers - that you likely have a potential gold mine of information you can harvest to help direct how you move forward.
For if a picture tells a thousand words, then a conversation tells a thousand stories from what is said, and unsaid, tone and inflection, the order of the words, and from the flow of the exchange.
Understanding the Conversations
When you think about it, for years, if not decades, your customers have been calling in to your contact centers, working their various ways through the options supported via your call scripts, talking to your agents when necessary, or possibly even a bit of both. They’ve even demonstrated their skill and understanding of the flow and logic of those scripts as they’ve used them over time.
Additionally, the flows themselves represent years of careful design, trial-and-error runs, A/B testing, grammar tuning, and perhaps even some live usability testing. This is indeed fertile ground for harvesting insights into what your customers not only want, but also what they are good at finding and what they want to know over and over again.
And the same principle applies in reverse for what they don’t want to know or care about so much, and what they struggle to navigate their way around in your call flows.
...you likely have a potential gold mine of information you can harvest to help direct how you move forward.
So, before getting too far into replacing your old-school directed dialogs with nice, flexible natural language models or a branch of your DTMF-driven functionality with an ML-powered voicebot, or any other part of your voice self-service design strategy, look to do what the good physician does. Namely gather as much data as you can about the current “patient” state before making significant changes.
After all, to follow the analogy, no responsible cardiologist would have you undergo heart surgery without thoroughly finding out what has happened and is happening with your heart.
So, why should you make such critical, costly, and potentially disruptive decisions for the life of your business through your customer service without that information?
Voice analytics for voice self-service can come in many forms. There are traditional IVR reports that can tell you things like AHT in the IVR, how many caller input errors occurred in a given time period, how many callers transferred to an agent, and so forth.
Speech analytics generally refers to an analysis of any speech recognition technologies you have deployed in your contact center. This can include the number of speech utterances a caller or group of callers make in response to a given voice prompt, how many utterances were “out of grammar,” how many were recognized, and with what accuracy, etc.
Ultra-High Resolution (UHR) analytics shows you the operating efficiency and optimization levels your IVR and voicebots are achieving. When combined with CX technology, UHR analytics allow you to address pain and friction points for the caller: areas of your call flows that would otherwise go unnoticed.
The UHR analytics show you exactly how your customers perform at every single conversation turn (node) in your voice application. It also tracks individual caller frustration levels as they progress through each node and gets them help before they zero out or hang up.
Learn/discover which new technologies are truly ready for prime time and fit for your specific purposes.
You can use this as a guide as you change your call flows and add digital channels to accommodate new services, seasonal changes and so forth. It will do the work of your traditional number-crunching analysts and usability studies for you in real time, continually finding opportunities for cost savings and efficiency.
And just like a cardiologist checks up on their patients, and if needed has them undergo follow-up diagnostic tests to find out how well their hearts are working, so too you need to employ voice analytics to check out how well your new changes and investments continue to perform over time.
These methods, along with a well-informed and well-thought-out strategy for executing your digital transformation initiatives, will help ensure the customer traffic that is the lifeblood of your business keeps flowing nicely while you build out the new “highways.”
In summary, baseline what you already know and have, and in particular:
- Take a detailed inventory of what technology resources you own today and how adaptable these are likely to be going forward. Segregate end-of-life products and services from “still has use and potential”: just like those old DTMF days.
- Understand what worked well in the past for your customers and what still works well for them today. Really try to see this as if you yourself are the customer.
- Understand how varying age demographics use technology in different ways today. If you are a Boomer, ask your millennial kids or nieces how they themselves use technology these days. The same (and vice-versa) goes for Gen X, Greatest Gen, and any demographic. No harm in going straight to the audience after all of your online research.
- Learn what will likely never work and needs a completely new approach: or simply is not there yet (read as speech technology or AI in the early days). Research respected white papers and other sources of information whose authors don’t have skin in the game. Learn how to quickly discern what looks like product hype and what looks like the real deal.
- Learn/discover which new technologies are truly ready for prime time and fit for your specific purposes. Embrace them and add them to your digital transformation strategy toolkit as quickly as possible.
- Learn/discover which ones are not and walk away. You can always revisit.
While the debate over the responsible use of AI and services like ChatGPT will, and should continue with urgency, there is little doubt that niches like customer service and contact center inquiries with their well-defined rules and data domains will benefit from these advances.
That, coupled with significantly increased user acceptance of speech and the much-improved performance the technology providers promise, is likely to result in voice self-service finally fulfilling its long-awaited promise.