AI - Is it a customer or organisational need?
There is currently a clear trend in Customer Experience circles. It seems it has become almost impossible to attend a CX event without hearing people talking about or focusing whole presentations on the topic of Artificial Intelligence (AI) and chatbots.
Yet, is AI only the fancy gadget in fashion, or is it the solution to a real customer and organisational need?
Let’s look at the organisation first
A few weeks ago, I was talking to the CEO of a financial institution, who was proudly boasting to his audience of eager listeners, “I told my guys to get me a Bot by Easter!” I could actually picture him, whisky glass in hand at the local ‘CEO Club’, talking to his fellow CEOs and proclaiming, “Do you have a Bot? No? Oh well, I have a brand new one!” Then, in reply to any positive answers, eagerly declaring, “I bet my Bot is better than yours!”
His ‘guys’ were actually very happy to have secured this unexpected investment so easily, and excited to be able to accelerate their digital transformation and get involved in this initiative on the technological vanguard. However, they were also still trying to work out how to use the AI to best effect. They had a solution to customer needs which were still not clear. It was like having a brand new shiny toy and not knowing what to do with it.
I am also aware that other organisations are testing AI and chatbot solutions to reduce operating costs by replacing agents in the contact centre. And yes, obviously they also want to improve the customer experience, but their primary driver is cost efficiency. This is a very topical area of debate at the moment, with conservative predictions saying that AI will automate many processes and replace up to 50% of people’s jobs by 2050.
Now, let’s take a look at customers
I doubt they would say, “I really want to use a chatbot in my customer journey”. On the contrary, they are likely to be suspicious of the quality and relevance of service provided by a ‘machine’ that is interpreting and answering their questions. After all, this is a moment when customers need either help or answers they cannot find alone in their online self-service journey. Deeming themselves alien to the uncool group of old-fashioned people who pick up the phone to the contact centre, they are progressing their ‘I can work this out by myself’ journey online with determination… until they acknowledge they are really stuck. And here comes the dilemma.
Let’s be honest – when they are online, humans behave in a very strange way. They want to ‘connect’ with poeple, but without physical contact. They want to ‘communicate’, as long as there is no direct interaction with the audience. They want to ‘chat’, but without talking. They want lots of ‘friends’ and ‘followers’, yet they avoid engaging with them too much. And things are happy and rosy until they need something they cannot find by themselves in the self-sufficient and self-centred world of the World Wide Web. This is the critical moment when Dr Jekyll suddenly transforms into Mr Hyde and starts screaming, “I need somebody to help me, NOW!”. At this stage, any length of queue or waiting time separating him from that human help suddenly feels utterly unbearable. “You are number 12 in the queue, please hold. Your call is important to us”, while the ‘All by myself’ soundtrack by Céline Dion fills the never-ending holding time.
Where the customer need meets the organisational need
So how can we transform from the ‘I need nobody’ position to this desperate urgency in a split second?
Webchat offered that reassuring compromise of talking to a human, while not really having to actually talk to one. You know what I mean. Dr Jekyll can continue to behave as his amiable and friendly online self, without unleashing the loud and aggressive Mr Hyde. Wait time online somehow feels shorter, or at least more acceptable… for the first few minutes! Then, again, Mr Hyde starts rumbling from within with impatience, tapping away at the keyboard with increasing force and irritation. “Mr Chat Agent, are you there, or are you chatting with too many other customers at the same time and not focusing on me?”
This is where Super Bot can come to the rescue. No waiting times, no answer delays, no longwinded sentences, clear replies and links to self-help videos or step-by-step prompts. Just perfect and flawless… as long as the answers are relevant and personalised to the individual circumstances.
But it is not that simple
A recent study by Nimbus Ninety highlights that 47% of digital leaders believe that machine learning will be the most disruptive technology of 2018, yet only 11% identify it as one of their areas for investment. Instead, Big Data Analytics remained the top area for investment for most organisations, with 86% admitting they needed ‘to transform their approach to data to compete effectively’.
This makes eminent sense, as I believe that the introduction of AI needs to be a carefully designed balancing act with Big Data and a customer-centric approach. Simply adding some generic AI technology to the contact centre will not deliver the right customer experience. Simply replacing agents with bots will not meet customer needs. Customer Needs Analysis and Big Data should be the foundation upon which a ‘specialised’ AI solution is built, focused on a very specific ‘vertical’ process area at a specific stage of the customer journey. Unless the underlying customer insight, data and algorithms are accurate and fully tested, the outputs will be too generic and annoying. Garbage in, garbage out.
But another essential element of AI development is striking the right partnership between human intelligence and artificial intelligence. The development of omnichannel capabilities has already shifted human channels towards complexity. Many research studies show that people are happy to self-serve for transactional tasks, but still prefer to talk to people on the phone, face to face or via webchat, when they are dealing with a problem, an issue or a crisis. These are the highly emotional situations that ignite Mr Hyde, and having to deal with complex automation would increase his level of intolerance even further. In addition, AI solutions are still not good enough at detecting emotional context and may exacerbate a crisis situation.
In conclusion, AI can materially improve the customer experience, but the omnichannel journey needs to be carefully designed with a progressive escalation from automation and self-service to assisted service, leveraging AI first, but seamlessly moving to webchat and human intervention should the situation require it. Artificial Intelligence should not be considered in isolation as a trendy development or a way to cut costs. Instead, the focus should be on building ‘Augmented’ Intelligence solutions, which leverage the strength of computers and data and augment it with the power of human emotional intelligence and contextual empathy.