How far do you go with AI?
Does the basic chatbot respond to your current requirements? Will Machine Learning enhance your customer experience? Or are you ready for the full-fat generative iteration?
As this latest – and most topical – of technologies continues to fuel the next phase of the world’s digital transformation journey, these are the questions facing enterprises of all kinds and sizes.
The truth, of course, is that AI of all kinds is powering the constant evolution of how organizations provide their services, sell their products, and empower their workforces.
However, there is perhaps another, slightly controversial, set of truths: AI cannot (yet) fully replicate human inputs; AI cannot (yet) guarantee commercial success; and AI is definitely not (yet) a case of one size fits all.
That said, to deny its obvious and huge potential to redefine for a generation how the world works is, of course, recklessly naive.
Instead, organizations must assess where they are on their own individual roadmap and deploy AI to an extent which delivers added value whilst simultaneously avoiding unnecessary (and expensive) complexity.
Taking the guiding hand of an expert provider is a wise move.
“Many organisations are already benefitting from AI but aren’t really aware of it,” says Chris Bardon, Chief Software Architect at global enterprise-class contact center provider ComputerTalk, which is helping thousands to take the next step.
“Chat bots have been a thing for a long time, as has natural language understanding. Even predictive text on a phone keyboard is a form of AI. Generating new language – in the way ChatGPT is doing – is the flashy new element but it’s important to see through the hype and not be distracted.
“Understand the AI you already deploy, understand what new AI deployments can do for your business, and, most importantly, understand that it is actually the combining of multiple AI tools together that is really the way to achieve value-adding transformation.”
Of course, for most organisations, it’s all about improving the customer communication experience – smoothing their entry into the business and then getting them to where they want to be quickly and with minimum friction.
In the case of one ComputerTalk customer, they were able to leverage AI-powered natural language speech recognition to refine and improve its Interactive Voice Response (IVR).
“They had 25 different ways a call could be routed, and multiple menus that were failing to triage calls very effectively,” says Bardon.
“Lots of callers were ending up in the wrong place or timing out. Putting a broad, natural language question at the top – ‘what do you need help with?’ – made a huge positive difference. It’s not necessarily reinventing the wheel, or building an entire generative AI chat bot, but it’s saying where and how can I increase efficiency or make a customer journey better.”
Not that AI is a one-hit fix. A chatbot, for example, must be continually curated and evaluated if it is to deliver real value.
“Whether it’s using generated language, natural language, or if it’s just a simple IVR and text, you can’t just design it and put it out there into the wild,” says Bardon.
“You need be constantly monitoring abandoned rates, completions, whether customers are getting to the place they need or are they being deflected to human agents, how much dialogue is occurring in the process, is the conversation history maintained or do customers have to repeat themselves? Once people have had a bad experience with a bot, will they go back?”
Also, AI is way off from the point where it can fully replace the human agent.
During customer interactions via any channel, organisations are rightly concerned with what the quality of those interactions – as well as the content – says about their brand. They rightly want to know what is being communicated and how.
That means supplementing the human agent’s toolkit with AI-powered tools is where the opportunities lie.
“Among other things, we are prototyping a virtual coach-type application that leverages AI to assist an agent,” says Bardon.
“It can look for specific keywords and suggest a specific response. But it can be quite dynamic too, such as knowing whether the customer has called about a particular ticket before, what other tickets to this customer have been opened, and why.
“Those are things an agent could find in a CRM, but they have to go and look them up rather than have a virtual assistant anticipate their relevance and provide them in real time. That’s where I think AI gets really cool.”
One thing is certain for now – AI is not about to replace human ingenuity. It is a continuous process of evolution which will absolutely get more and more clever. Consequently, the organizations likely to get the most from it will be those willing to take an iterative approach.
To learn more about how ComputerTalk can help your and your customers’ businesses leverage the power of AI, click here.