We caught up with our CEO, Emma Robertson, to discuss AI as a tool and its impact on customer experience, the importance of getting your data right and when AI can go from being a help to a hindrance. Read on for Emma's valuable insights.
Q: Do you feel AI should only ever be used as a strategic tool and not a substitute for an effective and powerful customer experience?
Emma Robertson [ER]:“The reality is that we’ll see AI developing as a ubiquitous tool, supporting, augmenting or forming part of the overall mix, but never substituting.
If we see AI tooling as augmenting, both as it exists now and how it’ll develop, it's an exciting and empowering opportunity to focus on innovative ideas and creating the space to operate at the differentiated end of the customer and employee experience.
As soon as you say substitute, though, that suggests doing away with human intelligence, experience, and emotional connection. We should absolutely avoid doing that: if you can get your AI tool to replace everything you do, then so can everyone else and you end up with no competitive advantage when it comes to customer experience. We’ve spent the past 20 years trying to bring the focus back to the customer and automating out the brand to human connection will mean automating out your point of difference as well.”
Q:You touched on something there, which is the perfect segue. Customer centricity; where do you see AI fitting into the future of customer centricity?
ER: “I think a straightforward answer is everywhere.
You'll see very few customers saying, ‘I want some AI’, but you will see them saying, ‘I want faster resolution to my customer service issue’ or ‘I want to be able to try on that shade of lipstick without going to the store’.
AI might be part of the tool set that delivers on that capability, but it won’t be the entire experience. AI enabled experiences may be everywhere, but we need to balance it with the empathetic side of the experience, the one that talks to customer’s wants and needs, ambitions and aspirations, as well as the functional side of what we must deliver for them.
One of the key developments in customer centricity over the last 10 years has been to not think of CX as features and functions – but as experiences and connection; The top of Maslow’s hierarchy of needs. It's not about just having a dropdown box or having the right navigation, but it's about pairing that with an understanding of someone’s motivations. What else is going on in their world? How do we fit within their ecosystem rather than them fit into ours?
At the same time AI from a data point of view, has a huge opportunity to provide some of that– turning years and decades of data gathering into predictive insights that drive experiences. Personalisation at scale will become the norm – but again, it will only be differentiating when balanced with experiences and brands that connect at a human not functional level.
Q: And you touched on what I think is really important, is having that insight. When talking about data, what are first steps that you think businesses of all sizes can take to prepare for that smooth AI customer experience?
ER: “It's twofold. Firstly, you’ve got to get your data sorted, and you have to do it now. Current AI tools are largely trained on Large Language Models (LLMs) that summarise, analyse and produce very credible outputs but it’s something everyone can do: everybody is accessing the same data, with the same tools, to get the same outcomes.
Particularly when we talk about customer experience, if you really want AI to start being powerful, it has to be “of you”; your organisation, your brand, your tone of voice, your history. That’s when AI is going to amplify CX. For that to happen, you have to feed these amazing tools, not just with the LLMs that everybody has access to, but with variations that include your own data sets.
Secondly, it’s about education and awareness from a cultural point of view. For some people AI represents fear, for other it represents excitement, or it might be an absolute distraction, and there's many other variations of how people are dealing with AI internally. This is an opportunity to accelerate that culture change and not wait until you're using AI in the best possible way or when you've achieved peak outcomes. Start that internal education journey now; tap into the people who are feeling fearful and understand how to reassure them: work with people who are excited and understand what you can do to harness that excitement and everything in between.
AI lends itself to a culture of experimentation which suits the maturity of the market right now. Agile applied well provides a strong cultural framework for experimentation with AI that many businesses will want. A data driven culture will also be an important part of that and getting the data literacy of your organization up will automatically create a more fertile place for AI.
Finally, you shouldn’t underestimate the importance of creativity in this because AI remains a technology that needs inspiration, creativity and innovation for how best to apply it. It’s not about substituting creativity with AI but fostering and amplifying it.”
Q: And is it that creativity, or what exactly excites you about having powerful AI tools and what it can bring to your business solutions?
ER: “AI provides the ability to get some of the repetitive low-value work done much more efficiently. Whether that’s pulling data sets for research or getting customers to self-serve FAQs before connecting with an agent, there's a lot of repetitious activity that’s important, but isn't at the sharp end of adding value. The most exciting thing is being able to spend more time in that higher order CX space, whether is the strategic, creative or technology thinking that's pushing the boundaries more. It’s a simple ratio of, if you're spending 80% of your time in preparation, you only have 20% left to do advanced thinking. Whereas if you filter out some of the repetitive activity, you might have 50% / 60% / 70% of your time left for advanced thinking.
I think there’s opportunity to create space for the creative and strategic thinking that drives the difference. You can have software engineers who are creative thinkers as much as creative designers; it's about identifying in every job role and function what it is that’s foundational versus flow.
Q: The Customer Experience panel you’re speaking on in March is titled, in short, ‘AI as a help, not a hindrance’. When do you think AI does go from being helpful to being a hindrance?
ER: “I can think of two possible scenarios.
One is when the hype gets in the way of doing what's needed. The headline gabbing AI examples are all very visual and novel – from writing songs in the style of Nick Cave (lawsuit pending on that one) or Coke Making their first ever fully AI generated advert. But beyond the hype, other things need doing, even if they’re not the typical headline-grabbing topics. To enable AI to automate processes, first those processes need to be consistent; to use AI for front line customer service requires training, content and strategy; to support mass personalisation you need your CRM, single customer view, customer history in place. So don’t be distracted by the hype and keep what’s actively going to take your AI journey forward.
The other scenario is when you're using technology for technology's sake, when the solution is AI, rather than the customer outcome-whether that's AI or any variation of digital data, software engineering, or data engineering.
You need to consider if AI is the right tool or part of the right tool set to solve that. We should never ditch the things we do that underpins a customer need because the organisation feels as though it should have an AI application. It becomes a hindrance when the answer is AI, but you don’t actually know what the question is.
In customer experience terms this is the history of digital development – 'we need a website', 'an app will solve that', 'let’s use AI'. Customer Experience came into its own when organisations start asking 'why’ before designing 'what' and 'how' – successful Customer-centred implementations of AI will do the same.”
