Today, a single bad experience can undo a hard-won customer relationship. Customers move quickly, compare constantly, and switch brands with less hesitation than ever.
Yet many customer experience strategies are still built around delayed reactions. Friction is identified only after it has already cost conversion or trust, whether through abandoned carts, negative reviews, or spikes in support tickets. By the time those signals appear, you’re managing refunds, discounts, support spikes and reputational damage rather than preventing them.
Leading retailers are changing that model. Instead of relying on a one-size-fits-all approach and reacting to support tickets, they’re designing experiences that are more context-aware and designed to reduce friction before or as it appears.
That evolution, from reactive support to proactive flows, is defining the next frontier of customer experience.
Move beyond scripted support
For years, automation in CX meant chatbots. They could answer FAQs, route tickets, or surface help articles, but they relied on predetermined scripts. When a customer stepped outside those paths, the experience stalled.
Today’s AI systems have moved beyond scripts. AI agents can interpret intent, analyse behavioural signals in real time and take action across systems, from updating order details to surfacing delivery options, without manual intervention.
That changes how quickly businesses can respond to emerging friction. Instead of waiting for a customer to submit a ticket or purchase elsewhere, AI can surface behavioural cues instantaneously, whether that’s extended time on a delivery page, repeated size comparisons, a stalled checkout. That gives businesses the opportunity to intervene with precision while the customer is still engaged.
At furniture and lifestyle brand Koala, the company is building what it describes as an AI “nervous system” that powers everything from cart recovery to operational forecasting. When a customer abandons a purchase, an AI agent can instantly generate a personalised recovery offer, balancing customer appeal with product margin and available inventory.
For most businesses, however, the starting point is smaller than that. Rather than deploying an AI “nervous system” immediately, teams can start by identifying one high-volume, rules-based use case where friction is predictable and repeatable, whether that’s delivery enquiries, order updates, return status checks. Automating those interactions builds confidence and frees capacity, while creating space to expand into more complex scenarios over time.
Design for friction before it surfaces
Once teams gain confidence responding to friction in real time, the next question becomes more strategic. How much of that friction could have been avoided altogether?
For many businesses, optimisation still begins after the launch. A brand introduces a new store design, a campaign goes live, and performance data determines what needs fixing. Even A/B testing relies on live traffic, meaning real customers absorb the downside of weak experiences.
AI is shifting that process upstream. Instead of waiting for a conversion drop-off and analysing what went wrong after the fact, AI can now surface experience issues before changes go live. Shopify’s SimGym, for example, uses AI-powered “shoppers” to move through product pages and checkout flows, simulating different personas, budgets and shopping intent to identify potential issues before a single customer can encounter it. That means fewer live customers encountering avoidable friction, and fewer downstream support interactions to manage.
Make your data AI-ready
If AI is helping teams respond faster and design more confidently, it’s also reshaping where experience happens.
Customer journeys are no longer confined to owned channels. Increasingly, discovery and decision-making take place inside AI-powered environments, from search assistants to generative chat tools. In fact, since early 2025, Shopify has seen a ninefold increase in AI-driven traffic to merchant storefronts, alongside a fifteen-fold rise in orders originating from AI searches.
That expands the remit of CX. Instead of designing solely for websites or apps, teams now have to consider how their brand shows up inside AI conversations. Product clarity, intent-focused descriptions, availability, delivery information and pricing all need to be structured so they can be understood and acted on by AI systems.
To do this effectively, product data needs to be treated as strategic infrastructure. Descriptions must be detailed and benefit-led, and variants need to be structured consistently. Formats such as Q&As, bullet points, and comparison tables also make information easier to parse, for both machines and customers.
As discovery and transaction converge into agentic commerce, the shopping experience becomes one fluid conversation. The brands that show up consistently will be those that ensure their product data is accurate, structured and easy to interpret.
Connected systems strengthen the impact
All of these advances become more powerful as systems become more connected.
AI can surface hesitation, simulate journeys and operate inside conversational environments. But it can only act on what it can see. If inventory isn’t updating in real time, AI might recommend something that’s already out of stock. If pricing differs between channels, customers receive conflicting information. If order history and behavioural data sit in separate systems, AI can only identify and act on trends based on partial context.
That’s why unified systems matter. At Mocka, for example, running on a single commerce platform means teams have a consistent view of product data, orders and customer behaviour across channels, giving tools like Sidekick, Shopify’s AI assistant, meaningful context to work with. Rather than piecing together insights from disconnected systems, the team can see behavioural shifts across their business as they happen, whether that means adjusting merchandising, refining messaging or responding to emerging demand patterns.
In practice, that means pausing before layering on more AI capability and asking a harder question: are your systems aligned enough to support it? When your foundations are connected, automation becomes more reliable, interventions more precise and experiences more coherent.
Build for anticipation
Customer experience is no longer something that sits at the end of the journey. It begins earlier, and increasingly, in places brands don’t fully control.
AI is giving brands new leverage. It helps surface hesitation while it’s happening, test ideas before customers feel the downside, and show up clearly inside AI-driven discovery. The retailers that pull ahead as AI-powered CX evolves are the ones taking the time to connect their systems, clean up their data and think more deliberately about where friction shows up. That’s what will define strong CX in the years ahead.