The Rise of the Retail AI Assistant: What to Expect by 2030

By 2028, generative and agentic AI will influence over $595 billion in global retail eCommerce sales. AI shopping assistants will guide shoppers toward that spend. Behind them sits AI for enterprise: the assistants your own teams use across commerce, CRM, ERP, and PIM.

Most of those assistants don’t talk to each other yet. That’s set to change.

Too many copilots, too little context

Nearly every system now ships its own copilot. In fact, almost every enterprise application has an embedded assistant, as predicted last year by Gartner. Each one knows its own system and nothing else. Ask the commerce copilot about stock, the CRM copilot about customers, and the ERP copilot about margin. Then stitch the answers together yourself.

The stack behind them explains why. 44% of retail executives say their company’s legacy systems are slowing down innovation. At the same time, 68% expect to deploy agentic AI for key operational and enterprise activities within 12 to 24 months. This shows that agentic AI is arriving faster than the architecture underneath it.

Fragmented assistants also limit context. Take one question: should we discount this jacket? With all the necessary signals firing from different copilots, an assistant that sees one of them only answers one quarter of the question.
Curious how AI can streamline enterprise fashion operations without adding more silos? Discover SCAYLE Retail AI.

The rise of the connected AI assistant

A connected assistant works across your systems and shares what it learns. Instead of four copilots, your team talks to one. It combines three kinds of context:

• Business context: products, prices, stock, margins, and merchandising rules
• Organizational context: roles, permissions, approvals, and workflows
• User context: who’s asking, what they own, and how they work

Gartner sees the path clearly: embedded assistants first, then task-specific agents, and multiagent ecosystems by 2029. Picture a merchandiser asking why linen shirts slowed down in Germany. The connected assistant checks live stock, compares demand by market, reads the markdown rules she’s allowed to change, and drafts a proposal for her approval. One question, one coordinated answer.

Organizational context keeps that answer safe to act on. Only one in five companies has a mature governance model for autonomous AI agents.
Want AI that works across your commerce operations? Explore SCAYLE Retail AI.

What will this mean for retailers by 2030?

Assistants will reach shoppers and teams at the same time. McKinsey estimates that agentic commerce could orchestrate up to $1 trillion in US B2C retail revenue by 2030. Your foundation decides how much of that your brand captures.

Expect these shifts between now and then:

• Assistants will move from answering questions to completing tasks.
• Shopper and team assistants will draw on the same product, stock, and customer data.
• Retailers whose systems share context will gain the most, because assistants can only act on what they can see.

Three building blocks prepare you for the move to autonomous agents:

Give AI fashion context

Fashion adds layers such as size and fit, seasons, trends, and brand rules. An assistant that understands how they relate can support commercial decisions, not just technical ones. The same context should serve the shopper-facing assistant and the team-facing one, so recommendations, merchandising, and operations follow one logic. Our guide Beyond AI Features shows how enterprise brands turn AI into retail-ready outcomes.

Connect your systems

Assistants can only use data they can reach. API-first architecture and real-time data pipelines give them live access to stock, prices, and customer profiles. Our article on preparing your tech stack for agentic AI covers the technical requirements.

Keep control

Autonomy should grow with trust. Define which decisions an assistant makes alone and which need approval, then expand as results prove reliable. Retail leaders at our eCom Leaders event described the same approach: fix the data foundation, align stakeholders, and test in contained areas first.

SCAYLE’s AI developer shows the principle in practice. It follows 13+ project-specific rules, so it works inside each store’s architecture and conventions.

Which assistant will your team want in the future?

Count the copilots in your stack today, then how many can see more than one system. The gap between those numbers is your roadmap. Retailers that connect context, systems, and control now will give assistants real work by 2030. Which assistant will your team trust to do it?

Ready to build the foundation for connected retail AI?