Intelligent Experiences: the future of customer experience

If you are worried about AI and the impact on your business and your relationship with customers, you’re not alone. But chances are you are focused on the wrong thing: most of the discussion has been on showing up in LLM conversations. Or shifting your customer service from humans to agents. Or using AI to optimise internal processes and workflows. Or preparing for a future of agent-to-agent interaction.

All of these are important, and represent real opportunities. But if you allow big tech to own the customer experience your brand will soon be dead, and your other AI initiatives will amount to rearranging deckchairs on the Titanic. The solution is to supercharge the customer experience on your own channels, from web to apps to retail, with AI. We call these supercharged touchpoints: Intelligent Experiences.

What are Intelligent Experiences (IX)

Intelligent Experiences represent the future of customer experience. We can already see this in the way major consumer environments are evolving: Google is moving with AI Mode toward generative layouts, interactive tools and simulations; Amazon uses Rufus as an AI shopping assistant inside the shopping experience; Booking.com and Expedia are shifting travel planning towards AI-driven, personal and visual interactions; and OpenAI is making ChatGPT, through Apps, increasingly into a place where interactive brand and service interfaces can live.

AI is therefore becoming a native ingredient of websites, apps, platforms and all other digital experiences, including emerging channels like voice, XR and phygital spaces. This is not simply using AI to build static interfaces, it's AI building content and interfaces in real time for an audience of one.

Today, the most well-known example of AI being the interface is chat: the default UI of LLMs. However, LLMs themselves are becoming increasingly multimodal and able to render UI on the fly. Google now explicitly calls this "generative UI": AI that can create an appropriate visual, interactive experience for a specific question. In the near future, we expect a Cambrian explosion of adaptive, real-time rendered interfaces, not just in LLMs but everywhere on the internet and in all digital products.

Some of these will be very rich and motion-driven, including real-time video and visual browsing experiences. Some will be completely minimal, like a just-in-time AR interface projected onto smart glasses to help delivery drivers with the right info to be more efficient. Many will be somewhere in between: agentic commerce solutions that combine brand content with product content to deliver hyper-personalized recommendations, or intelligent experiences that manage life events from holiday booking to moving house end-to-end, mediated by multimodal AI through a combination of AI phone calls, email/CRM, and web interface.

Imagine if booking a stay in Paris at a luxury hotel chain combines an interactive video of your hotel with personalized recommendations for nearby things to do. A whole page of text and visual content based on a single prompt. Adding guests or changing dates updates the entire page to match your new requirements. You can book the trip, or save it until the best price matches your personal schedule.

There are many more opportunities to explore. We should think of IX as a design and development philosophy that changes not only how digital experiences will look and behave, but also how they are built and managed.

Does IX mean that every aspect of the experience is dynamic? Not necessarily. Even if only the content is AI-generated, or the UI is dynamically assembled based on existing components, we can still speak of IX. It’s the feeling for the user of having their applications behave in a delightfully personalized, responsive, cognified way that makes them intelligent experiences.

Why it matters

While foundation labs try to justify their valuations by convincing everyone that all B2C interaction collapses into agent-to-agent communication, the opposite is true: AI not only unlocks new ways for brands to connect to their customers, it also creates a strong urgency. Brands should leverage IX to claim buyers’ engagement and loyalty before they are commoditized by LLMs - competing for attention in an agentic feed without any kind of lasting relationship with their own customers.

This is not a zero-sum game. By creating the internal architecture necessary for IX, brands also make it possible to expose their brand and product content to external agents or LLMs. A retailer can invest in Intelligent Experiences by creating a rich and innovative shopping experience on their D2C channel. At the same time, they can expose part of their product catalogue, for instance refills or hero products, to be indexed and served inside LLMs or commerce agents, using much of the same data architecture.

So a brand can support some types of agent-to-agent commerce while also making sure they can connect directly with customers on their own channels. Just like they do today across a marketplace versus their own site, or a shop-in-shop format versus the flagship store.

There are many reasons to distrust the "no one will visit websites anymore, everything will be agent-to-agent" narrative: the unreliability and lack of transparency in agentic recommendations, the human need to validate important decisions, and the lack of engagement in these types of interactions. To put it simply: asking an agent to find you a pair of running shoes and pick one of three recommendations based on pictures, lowest price and a star rating might be fast and accurate, but it also boring. And for many product and experience categories, users don’t want boring.

But we don't have to get into a religious argument about it. IX is the future of digital interaction, whether through traditional or new digital channels. The more important point is that it requires a new kind of data and systems architecture.

The new architecture: Intelligent Experience Systems

One important implication of the emergence of IX is that it speeds up the convergence of different databases and systems into a strongly integrated, connected but modular tech and data architecture. This architecture supports a single, intelligent ecosystem that enables both a singular end-to-end customer experience and a singular end-to-end employee experience.

Concretely: a user gets a shortlist of options through AI search, is directed to an intelligent experience on the brand's website which is tailored to their profile and brings in rich brand experience content, continues their journey asynchronously in store where they are recognized by staff through their loyalty program ID, makes a purchase - potentially with an automatic replenishment subscription, and receives agentic or human customer support and future promotions based on their complete purchase, engagement and behavioral profile.

If this sounds like what today's CRM vendors already promise, it is in fact only possible by removing the existing siloes between sales data, CRM data, campaign engagement data, analytics and media insights. In addition, all of this data needs to be made interpretable by an agentic orchestration layer that generates content and UI in real time, then tracks the user response in a feedback loop that adapts and improves the experience from click to click.

All this is achieved by connecting different data sources into a unified customer profile: the customer golden record. The underlying data architecture represents the customer as a multidimensional set of relationships: identity, preferences, permissions, behavior, transactions, content engagement, service history and journey context. Vector databases enable agentic systems to interrogate this profile depending on context and journey phase.

Another requirement is for data sources like PIM and DAM systems to be enriched with additional information and formatted correctly. All marketing and brand data, whether it exists in content, in people's heads or on external channels, needs to become machine-readable and ordered into a single reliable, authentic and unified representation of the brand, so that all communication remains on-brand. And finally, IX Systems require carefully modelled and transparent governance structures, data privacy and digital autonomy infrastructure that gives people in the organization visibility over orchestration and control.

This is a lot. But Intelligent Experience Systems do not get built up in one big transformation project. Rather, they represent a set of architectural, design and development principles and practices that move the organisation from an old way of working to a new one, suitable for the age of AI. The job is to identify the areas where IX brings the most value to the organisation, build the infrastructure to make that happen, and then build out from there.