The business world has gone crazy over AI. The leading manufacturer of AI chips Nvidia has seen its share price almost single handily drive the growth in the S&P500. In the year from June 2023 Nvidia shares have tripled in value and it accounted for 34.5% of the S&P 500’s market cap gains in the first half of 2024.
The Booming Market for AI Chatbots
The global chatbot market is predicted to grow at 23.3% (CAGR) between 2023 and 2030 with Gartner predicting that AI chatbots will be the primary customer service channel by 2027. They are expected to make massive cost savings and even drive sales revenue. The prize for getting it right is huge and companies are desperately investing to try and keep up, let alone get ahead of the trend.
The Customer Experience Challenge for AI Chatbots
The potential rewards for companies appear to be huge but there is one not insignificant problem – customers don’t like them.
According to Gartner, only 8% of customers used an AI chatbot in their most recent customer service interaction, and of that small proportion only 25% said they would do so again. Gartner went on to discover that where the need is relatively simple such Returns or Cancellations the chatbot resolved the query in 58% of cases. However, where it was more complex such as a billing dispute, it only resolved the query in 17% of cases.
And, research by Callvu revealed that 81% of customers would rather wait to speak to a human customer service rep, than use the immediately available chatbot.
As one customer said to us in a concept test conducted recently for a financial services client, “Please don’t give me another c**p chatbot.”
The Misalignment of AI Chatbots with Customer Experience
At Caffeine we believe that the critical reason for this mismatch between business efficiency and customer preference is because companies are not developing or managing the capabilities of the proposition in line with the expectations of customers.
The root of the problem is that there is a classic fascination with the technology as opposed to what customer want from the experience. This leads to over-hyping the capability of AI chatbots, and failing to design in some fundamentals that customers expect.
Critical Elements Missing from the AI Chatbot Experience
So, what are some of the critical elements that are missing from the AI chatbot experience at the moment?
Research by Reutlingen University showed that to be take seriously the solution needs to:
- lower customer effort
- deliver a quality solution
But critically, the experience must have:
- high procedural & interactional justice.
The Importance of High Procedural and Interactional Justice in AI Chatbots
OK but what exactly is high procedural and interactional justice? In translation, this means that the experience must feel empathetic and fair.
That’s what the research says but there are also some companies out there proving the theory. One example is Lemonade insurance, who has created two AI chatbot experiences, The first, called AI Maya, is described as “a playful onboarding and customer experience bot”. The second is AI Jim, which handles claims. Simply the naming of these two experiences shows the lengths that Lemonade went to in designing the interaction to feel more natural, empathetic and human.
Case Study: Lemonade Insurance’s AI Chatbots and the Customer Experience
AI Maya needs just two minutes of your time to create a personalised quote and be ready to take you through an easy payment process. Compare that to the common experience of getting home insurance quotes elsewhere, for us that is a huge reduction in time and effort.
What about AI Jim, can it make the claims experience feel fair? In December 2016, AI Jim received a claim from a Lemonade customer, describing the situation (a stolen coat) into their phone camera. In the next three seconds, AI Jim reviewed the claim, checked the policy, ran anti-fraud checks, approved the claim and send the instruction to payout the money.
“I was shocked by how easy the process has been with Lemonade…I already assumed there was no way that I’d recover my losses: other insurers either pile paperwork or deduct tons of charges that I don’t understand. But this time was different. I signed an honesty pledge, answered a few questions, and Lemonade reimbursed me in a matter of seconds! The service is amazing and I am so happy that I signed up!”
Lemonade goes on to say that AI Jim can take an entire claim to resolution on its own for almost 50% of all its claims.
Lower customer effort – tick, deliver a quality solution – tick, high procedural & interactional justice – tick.
Improving the Customer Experience with AI Chatbots
So, what’s the lesson? If you want to win over customers, save your company a lot of time and money, and surf the wave of AI chatbot hype, start with the customer not the technology. Understand customer expectations of the complete experience you want to re-invent, then take time to design an experience that brings together the human, digital and physical interventions that meet that expectation the best.
