Amid the hype around artificial intelligence (AI) and machine learning (ML), it is easy for businesses to fall into the trap that technology is a panacea to solve complex business problems. The truth is more nuanced, say two leading executives in the unified communications industry, which is on the verge of evolution from reactive to proactive customer service.
Unified communication platform service providers are “flying in the tail of a comet of research and development by industry giants Google, Amazon, and Microsoft, picking up the building blocks of advanced technology and applying it to real-world customer experience (CX) scenarios”, says Telviva COO Rob Lith.
“As these tech giants spend a massive amount of money on R&D for commercial consumer products, we are able to pick up these AI and ML building blocks and apply them to specific uses in a contact centre environment for example,” says Lith. “However, things such as AI and ML are tools in the war chest.
There is a saying that AI is not magic, but maths. It can be coded to be very clever and to achieve the goal we set out to achieve, but it still needs the human touch.
“Human business analysts use these tools and the insights they bring to the fore to get an enriched understanding that is useful to them as they optimise businesses. As contact centres evolve to use a multichannel approach, AI allows for greater scale, but it is always an augmentation, not the total solution.”
Genii Analytics CEO Kobus van der Westhuizen agrees, saying when people understand what ML and AI are, the concept of augmentation makes sense. “At its simplest, ML is a data-driven platform that accumulates multiple sets of data, with algorithms in place that pitches it on the parameters you want to teach it on.
Van der Westhuizen says over the past five years, the focus has been on how to augment human-driven conversations and communications, in other words, how to make the process faster, easier, and more effective – both inbound and outbound.
“The next wave we are working on in the industry is autonomous customer engagement – you now accumulate data and historical perspectives on customers and start doing predictive engagements, without a human in the loop,” he says.
He adds that AI is seeing an uptake in five principal areas, namely natural language understanding (NLU) and natural language processing (NLP), data, robotics, cognitive use cases, processes, and tech-enabled AI.
“The principal areas that service providers working in the unified communications industry are NLU and NLP, data and processes. With the vast amounts of data that organisations collect we can implement hyper-personalisation where you can start servicing customers from the position of knowing who they are, where they are, their demographic, their preferred channels, what they have bought and what they are thinking of buying, among much more. This empowers a customer-facing employee massively.”
“With processes, we look at process automation where we use AI and data to automate mundane processes that a human would previously have actioned,” he says.
Lith points out that beyond the convenience of this type of capability, it frees the human to invest time and energy on more strategic and complicated matters that require human empathy and nuance. “Ultimately,” says Lith, “All these tools are intended to be implemented to help businesses service customers better.”
Van der Westhuizen agrees by saying that by integrating this level of intelligence and analytics into a platform, businesses can target real pain points, such as customer churn. “If you have access to this level of intelligence, you are able to start predicting what a customer will contact you with before they do, shifting you into the realm of predictive customer service, which is powerful when seeking to boost customer retention.”
Both Lith and Van der Westhuizen say that the marketplace has been struggling to prove return on investment (ROI) on AI. “This is the Achilles heel of AI, but can be avoided,” explains Van der Westhuizen.
He says that throwing technology at a problem without a proper business case has resulted in disappointment, and this will continue. “When approaching any business process – not just in contact centres – there must be a design-thinking exercise,” he says. “The business and their chosen partner must understand the problem from all angles before breaking it down into bite-sized chunks.”
By doing this, he says, a case can be made that AI and ML may or may not be able to address the pain points. “Once you decide it can, you are able to assign a hard deliverable, such as a percentage improvement in customer retention, for example, which gives you what we like to call a ‘true north’ – somewhere towards which you can calibrate your compass. This is measurable and you can demonstrate ROI.”
Lith agrees, adding that vague promises that AI is a panacea to solve all problems not only gives technology a bad name, but it results in wasted money, unfulfilled promises, and bad blood. “Approached correctly,” he says, “the power of technology holds immense promise to augment and bring out the best in humans.”