The age of the large language model is here. AI agents are being launched left, right and centre. Like it or lump it, AI has changed the way we operate today, streamlining customer service operations, boosting productivity by as much as 40%, according to McKinsey. But, AI can appear to be a bit of black box, and hard to know where to start.

In an era where consumers are valuing authenticity now more than ever, the stakes are high. If a chatbot is not designed or executed well, customers can end up becoming more frustrated with your brand.

Video: DPD Customer Service Chatbot admits it is unable to connect customer with a human

When taking the leap into AI chat, the quality of experience matters more than anything. 

Here are a few tips to help:

• Start with the problem, not the technology
  • Investing in a chatbot shouldn’t be something you do because everyone else is doing it. What’s the problem that you’re trying to solve? What do your customers really want? If 80% of your customer FAQs revolve around resetting passwords, then sure, an AI agent can do the heavy lifting. But if your customers value the one to one service they receive from your staff above all else, then consider the role AI plays in facilitating that CX journey.
• What you put in = what you get out. 
  • By taking the leap into AI, you really need to lift the hood to unpack its parts.
  • Don’t underestimate the importance of designing good process operations that form the backbone of the language model. Allowing the model to refine the customer enquiry will help optimise the accuracy of the answer. The clearer and more understandable the query, the better the output.
  • Safety and security needs to be baked into the engine. I.e. how is the interaction data stored, utilised, managed across your business. Building out an ethical framework is a good practice too.
  • This is not a set and forget endeavour. Maintaining your knowledge hub and making sure you have people in place to do so is key.
• Build your brand experience into the model
  • With AI you can set the tone and language of the bot’s response so that it can speak like you.  If your brand’s tone of voice is friendly and down-to-earth rather than professional and polished, customise the bot to respond in the way that is consistent with your brand’s identity.
• Harness the power of real-time customer insights
  • The big advantage of generative AI is the advancement in data analytics where via an AI agent, we can now analyse a large portion of customer sentiment, segment customers, and ingest this data into the business in real-time.
  • The implications of this: businesses don’t need to keep asking that NPS survey question after every call, rather, they can collect NPS sentiment directly using AI. Without the need for the middleman in between, this expedites insight into our operations, and turns Voice of Customer programs on their head.
• Always offer customers the option to speak with an actual human.
  • For some organisations, a chatbot is the first and only ‘humanised’ contact a customer has with them.
  • The bot should be there to support and assist, but it shouldn’t lead to a dead-end. If it’s the only way for customers to contact you, then you’re putting up an artificial wall.
  • There are going to be times when the bot spits out the wrong answer or leaves a query unresolved, so regardless of channel or location, there should be a way for customers to reach a live agent at any time.

Building consumer trust and loyalty in your brand takes years to do, and can be wiped away in an instant. In a world racing to complete digitisation, consumers are questioning what is real and what is fake, and indeed trust is the number one currency for your business. Human interactions build trust. Use AI to complement and augment your human interactions, not replace them.

Sources:

AFR article: Davidson, J., Eyers, J., Tadros, E., Patten, S., Black, E., & Lenaghan, N. (2025, January 21). The top six business trends for 2025. Australian Financial Review.

Forbes article: Arruda, W. (2024, August 4). How the rise of the AI-enabled employee will impact career success. Forbes.
https://www.forbes.com/sites/williamarruda/2024/08/04/how-the-rise-of-the-ai-enabled-employee-will-impact-career-success/

Harvard Business Review article: Chui, M., & Manyika, J. (2023, August). AI won’t replace humans, but humans with AI will replace humans without AI. Harvard Business Review.
https://hbr.org/2023/08/ai-wont-replace-humans-but-humans-with-ai-will-replace-humans-without-ai

Intercom Pioneer: Intercom. (n.d.). Pioneer. Retrieved January 21, 2025, from
https://pioneer.intercom.com/

Kriss, S. (2023). The AI empowered customer experience: A CX practitioner’s guide to the possibilities and risks of AI. [Paperback].

Teneo. Chatbot examples gone wrong: Lessons and insights. Teneo AI.

https://www.teneo.ai/blog/chatbot-examples-gone-wrong-lessons-and-insights