Author: Helen Murray, Chief Customer Solutions Officer at Webhelp UK

Webhelp recently conducted research with the polling experts YouGov to ask over 2,000 British adults what they think about Artificial Intelligence (AI) and how they believe it might change the way that brands offer customer service to them. Webhelp UK’s Chief Customer Solutions Officer, Helen Murray, takes a closer look at the findings…

From the research findings, it seems that most people still believe in the importance of the human touch in their interactions with brands. In our survey, 44% said that they do not think AI will impact them positively in any way and 52% said that it would make dealing with brands more impersonal.

You can take a look at this recent blog by my colleague Dave Pattman, Innovation Director at Webhelp, for a deeper dive into the research results, but as I read the results it made me think about some of the complexities of using AI in the customer service environment. When designing any customer service solution the ultimate objective is to deliver a fantastic customer experience (CX), so AI should really just be one more tool or option – why has it recently been under so much scrutiny?

I believe there are at least five key issues that executives planning CX need to consider more seriously:

  1. Customer journey planning; providing customer service using AI tools can create a more positive experience for customers, but in some cases it can dramatically reduce the experience. It is necessary to completely revise the customer journey so you can understand where there is the potential to trip up. It’s not good enough to think of AI just as one additional service channel that can be bolted on to an existing customer service operation.
  2. Going too far; some organisations jump in at the deep end and go for an approach to CX that relies heavily on AI. This can partly be driven by how ‘cool’ it appears to be using cutting edge technologies, or perhaps because of a desire to reduce the cost of human agents in the contact centre. Whatever the strategy, it is never usually advisable to dive in headfirst, try some pilots and see where your customers will accept AI. It’s possible to use AI to support your human agents rather than on the frontline with customers, so think carefully about where it might be most effective for the first test projects.
  3. Not implementing enough AI; though it seems to be just the opposite of issue 2, what I mean is that AI and automated service powered by AI really does have some benefits – despite the misgivings expressed in our research. Cleverly introducing a first line of support that is automated can dramatically increase your reaction time to customers (and availability), so there is a danger that executives may not achieve these gains if they are apprehensive about running some pilots or test implementations.
  4. Forcing the customer to use human or digital service; this is a big problem in many customer service implementations that have introduced AI, especially with chatbots. If you make it impossible for the customer to switch to human service when they are frustrated with the answers from the bot then you will dramatically reduce the customer experience. Use your automated systems in the right way and for the right type of engagement and customers should be happy to engage with an automated system. But always allow them the option to switch because sometimes their problem will venture beyond what the bot can cope with!
  5. Planning based on technology not customer needs; As I mentioned in the first point, your AI implementation should be driven by the customer journey and the contact drivers – do you really understand what is causing the customer to get in touch? With simple transactions customers can actually prefer to engage with a bot because it is easy, fast, and there is absolutely no waiting to be served. But nobody wants to engage with a bot when the discussion requires a conversation that is emotional or highly personal – perhaps a life insurance claim for example. Understand what drives your customer to get in touch and you can far more effectively plan the best way to respond. In short, none of this journey planning should be driven by what the technology is capable of – the AI itself is just a tool.

It’s clear to me that we have been barrelling along what the industry analyst firm Gartner calls ‘the hype cycle’ for the past couple of years with AI. However, there are now live implementations and good case studies out there. This is now becoming a reality, but in some cases the possibilities that the technology offers is driving strategy rather than a cool calm focus on what the customer really needs.

Consumers in general are getting familiar with many AI-powered systems. Millions of homes now have Amazon’s Echo or Google Home installed. People ask Siri or Cortana for help rather than typing questions into a search engine. In many cases I believe that people are being exposed to AI far more than they probably realise.

Our research with YouGov confirms that managers with a responsibility to plan how they can deliver the best possible customer experience need to strike the right balance between the advanced technology available and real people – humans helping humans. Consumers are getting more familiar with AI, but a poorly implemented customer service system using AI can create some very negative customer experiences.

How do we really want to interact with brands? What do we really think about AI and Automation? How important is it to strike the right balance between human talent and AI and Automation for CX? Click here to download the complete Webhelp and YouGov study

In case you haven’t registered yet, Sign up to receive fresh insights and invitations to exec events with our Webhelp Disruptor Series campaign: Watch our video to know more!

Get in touch if you’re facing disruption – or want to disrupt – and want to talk through the implications for CX. E: