Which customer interactions should we build AI experiences around?

Start with one or two things your business can do exceptionally well for its audience, rather than an interaction designed mainly to showcase AI. The Frameworks argues that access to generative AI tools does not produce quality output on its own: professional expertise and hands-on experience with the tools need to work together.

Use audience understanding, brand purpose and customer wants and needs to decide which task deserves attention. The Frameworks advises brands to keep those fundamentals central, even as AI capabilities change. The aim is not simply to work faster or produce more efficiently, but to create more meaningful, relevant experiences. A useful starting question is therefore where your expertise can help customers in a way that expresses your brand's purpose and strengthens the relationship.

AI earns its place when it makes a useful interaction possible or practical, rather than merely adding a conversational layer to an existing journey. The Frameworks describes its approach as using AI as the method, not the subject: building personalised, interactive experiences intended to withstand real use rather than serve only as demonstrations.

Assess the opportunity in terms of what customers can experience differently. The Frameworks reports launching a small AI project that delivered a previously unattainable level of customisation and personalisation for its client and the client's customers. That provides a practical criterion for choosing an interaction: whether AI enables a useful degree of personalisation that was previously impractical.

In a 2025 test comparing a fixed chatbot with LLM-generated savings guidance against website Q&A, the Financial Conduct Authority concluded that comprehension and engagement depend on content design, delivery and placement within the customer journey, not just the model.