ey by Genesys, 78 per cent of respondents cited “they understand and acknowledge my emotional state” as a top service criterion. Consumers want to be understood Machines are limited in how well they can understand human beings. Natural language processing (NLP) is improving, but most automated communication interfaces remain relatively crude.  Where humans continue to have an edge over machines is in understanding the context of communication, not just all possible semantic meanings of a word or phrase. This context is critical for successfully delivering a personalised customer experience, because it involves truly understanding what the customer wants or needs. Consider this: a doctor listens on average to a patient for just 11 seconds before interrupting them. Why? Because they are an expert, they have seen it all before, they want to give the right diagnosis and offer the right solution. So, they step in to expedite the process. Meanwhile, research suggests that consumers are increasingly frustrated with chatbots and automated phone menus. A survey by Simplr found that 80 per cent of people are much more willing to use a chatbot if they know they can easily and quickly transfer to a live person. This is both the risk and the opportunity for AI-driven experiences. Consider for a minute how we establish relationships with new people we meet. Let us say it is the first date. You ask about their family, what they do for work, for fun, and down the track get to know their dreams and goals. You don’t ask or assume upfront. CX leaders must prioritise seamless collaboration between bots and human agents.  Using data ethically and well When we personalise experiences, we have a wealth of data that can help us make assumptions to shortcut for convenience. To ensure the experiences are relevant and sit on the right side of being creepy, we need to make sure to allow space to build a relationship over time, without interruption. Moving toward explicit personalisation and data consent is not only critical in a post-cookie world, but also ethically and legally necessary. Customers are willing to offer more personal data in exchange for receiving value. Understanding your customers better and carefully planning and designing automated customer experiences will make them much more effective. There are design techniques such as progressive onboarding, where you allow a customer to set preferences and validate what you know about them over time.  As an example, the Simplr survey found that older demographics were much more “annoyed” by uninitiated chatbots that pop up on a website than younger consumers, who do not mind them nearly so much. An older consumer is more likely to prefer talking on the phone to a live agent, compared with younger people, who are more likely to prefer starting with a chatbot. Involve employees When digitalising a service that needs to mimic or replace a human one, we need to ensure that employees are involved in the design process. This includes involving them in training the AI to help it make predictions and ‘think’ and respond in the same way human staff would. We know that one size does not fit all – every person needs to feel understood – but most customers still feel treated like numbers. You need to design for one, before you extend to many. Anticipating needs means listening and getting to know your customer. The problem is, even humans are not great at active listening. And this is something we need to get better at if we want machines to be able to do it. Ultimately, the future is human. Every dollar and minute spent is a potential investment in building a connection with your customers or employees that brings an opportunity to engage with them at a deeper, aspirational level, using technology not as a tool, but as an advantage. Dig deeper to create those purposefully distinct experiences that address their functional and emotional needs and remember that, however you measure it, value always belongs to the perceiver.