From Hype to High-Impact - AI in e-commerce: Key Takeaways

AI is rapidly moving from experimentation to practical application across enterprise commerce. Last week, fusefabric, Deloitte Digital and Shopify brought together enterprise leaders at Deloitte Digital’s London studio to discuss how organisations are beginning to operationalise AI in real commerce environments, and what teams need to put in place to prepare for what comes next.

The conversation deliberately moved beyond AI as a collection of features. Instead, it focused on how AI changes decision-making, operating models, and risk across enterprise commerce teams.

Below are the key themes from the discussion.

From AI support to agentic decision-making

AI is increasingly moving from supporting teams to influencing, and in some cases making, decisions across discovery, recommendation, and operations.

This shift will not happen uniformly. Some parts of the commerce journey, such as repeat or utility-driven purchases, are likely to become agent-led sooner. Other areas, particularly those driven by inspiration and brand experience, will continue to rely on richer, more human-led journeys.

For enterprise teams, this means planning for a mixed model rather than a single future state.

Foundations matter more than tools

A consistent theme throughout the discussion, and something many organisations are seeing in practice, is that technology is rarely the limiting factor.

More often, organisations struggle because:

• Product data is inconsistent or poorly structured

• Decision ownership is unclear across teams and regions

• Delivery processes are designed for slow, incremental change

AI does not fix these issues, but rather it exposes them.

Teams that are seeing real value from AI have invested first in fundamentals: clean data, shared standards, and operating models that allow decisions to be made and acted on quickly.

Commerce experiences will change shape

E-commerce experiences have gradually become more uniform over time, but now agentic systems challenge this pattern.

As AI becomes more aware of intent and context, experiences are likely to shift away from navigation-heavy journeys toward more guided, mission-focused interactions. At the same time, some discovery and transaction will increasingly happen outside owned channels, mediated by AI agents.

This raises the importance of clarity, both in how products are described and in how brands communicate value in ways that systems can interpret.

Differentiation needs to be clear and explicit

One concern raised during the panel was whether brands risk losing identity as AI plays a larger role in the buying journey.

The shared view across the panel was that differentiation remains critical, but it must be expressed clearly. Product value, service quality, loyalty mechanics, and trust signals need to be understandable not just to customers, but also to the systems acting on their behalf.

In an agent-led world, ambiguity becomes a disadvantage.

Where AI is already delivering value

While agentic commerce is still evolving, there are clear areas where AI is delivering measurable benefits today:

• Data and insight, enabling faster analysis and decision-making

• Content and merchandising, improving product information at scale

• Operations, reducing time to market and the cost of change

Organisations seeing success are focused. They apply AI where it removes friction from existing bottlenecks, rather than experimenting everywhere at once.

Software delivery is evolving too

The impact of AI extends beyond customer-facing experiences. It is also changing how software is built and delivered.

Traditional delivery models, often built around hand-offs and fixed roles, are increasingly being challenged as AI accelerates the pace at which teams can build, test and iterate. In their place, more continuous and integrated approaches are emerging, where teams can test, adapt, and deploy more quickly with AI support.

For many enterprises, this represents a deeper change than simply adopting new tools. It requires rethinking how teams collaborate, make decisions and deliver digital experiences.

Governance enables scale

AI governance was another key topic. The consensus was clear: governance should enable progress, not slow it down.

Leading organisations are approaching AI risk deliberately across data, legal, operational and people considerations, while putting structures in place that allow innovation to continue responsibly.

This level of discipline is essential if AI is to be trusted at scale.

Final reflections

Agentic commerce is not a distant concept. The capabilities that support it, strong data foundations, clear decision-making, and adaptive delivery models, are already shaping enterprise commerce today.

The strategic alliance between fusefabric and Deloitte Digital reflects a shared commitment to helping organisations navigate this shift in a practical and structured way, combining strategy, technology and delivery to operationalise AI in commerce.

The key takeaway from the discussion was clear: success with AI depends less on experimentation and more on organisational readiness. The organisations making the most progress are those investing early in the foundations that allow AI to scale across their commerce operations.