Agentic AI and the future of retail: Staying ahead in the age of intelligence

Now, we’re entering the third great disruption: agentic AI. According to Salesforce, 22 percent of shoppers are already using AI for inspiration and product discovery. That’s just the tip of the iceberg.

At Talon.One, we see agentic AI as a shift with the potential to reshape retail as profoundly as e-commerce or marketplaces did before it. Whether it reaches that scale will depend on how quickly both consumers and brands embrace it.

What feels certain, though, is that loyalty will need to evolve. The retailers best positioned for success will be those who move beyond loyalty as a simple "earn-and-burn", and instead build it into an ecosystem that creates lasting connections as AI becomes part of everyday shopping.

Most people are familiar with tools like ChatGPT. They’re great at answering questions, generating content, or helping brainstorm ideas. But what they don’t do, at least not on their own, is take action.

That’s where agentic AI comes in. Instead of just giving advice, agentic AI can actually do the task. It sets goals, makes decisions, and completes actions without constant prompting.

Here’s what that difference looks like in practice:

This shift is already here. Earlier this year, Shopify partnered with OpenAI so that when you ask ChatGPT to help you shop - say, “Find me a 4K TV under €1,000” - it doesn’t just give you a list. It shows real product cards, reviews, and product links.

Since then, Shopify has released its first agentic commerce functionalities - bringing together three platform capabilities so agents can help buyers find products, add them to cart, and check out, all without leaving the experience. That’s a huge leap from static recommendations like Google Shopping. 

If this is where we are today, what comes next for AI and e-commerce?

Many experts believe the next step is when AI agents move from assisting to acting independently. Instead of simply presenting choices, they’ll be able to handle the entire shopping journey: anticipating needs, scanning across retailers, weighing price against value, negotiating offers, and finalizing purchases, all on behalf of the customer.

We explored the idea during a fireside chat at this year’s INCENTIVIZE summit, with Jens Scharnetzki, Chief Product Owner at ENBW, and Stephan Ritter, Director at Deloitte Digital. As Stephan put it: 

Imagine a customer’s AI anticipating they’ll need new running shoes before they’ve even thought about it, finding the perfect pair at the best price, and completing the order without them ever opening a browser. That’s where we’re headed.

When this becomes the norm, beautifully designed product pages and checkout flows may matter far less. The real challenge becomes: How do you ensure that an AI, and its human, chooses your brand over every other option?

The good news is that retailers already hold a powerful advantage: first-party data. When used responsibly, this data can power hyper-personalized experiences that rival the recommendations of any external agent. The key challenge is knowing how to prepare for what’s coming next.

Here are three strategies brands can use to stay ahead in an agentic AI world.

If you’re concerned that AI might make it easier for customers to switch to alternatives, strengthening your loyalty program to create genuine, value-driven relationships can help safeguard those connections.

Winning customer preference requires more than points or discounts. It’s about inspiring direct connections built on:

Practical examples include free shipping for direct purchases, bonus gifts for loyal buyers, or exclusive events that reinforce community.

In a world where AI does the buying, visibility requires a new discipline: Generative Engine Optimization (GEO).

Just as brands once optimized for Google rankings, they’ll now need to optimize for AI-driven discovery. That means:

One of the most important shifts happening right now is the development of Model Context Protocol (MCP), an emerging standard designed for AI systems to securely access external data and tools. This means AI agents using MCP can pull in live, structured information like product catalogs, availability, or loyalty rewards.

For retailers, this is a big deal. It means promotions, offers, and loyalty program context can become directly visible to the AI agents making shopping decisions. Rather than an AI recommending a competitor simply because it can’t “see” an offer, MCP gives retailers a way to ensure their incentives are part of the equation.

Imagine a scenario where:

The takeaway: It’s not enough to have great promotions or loyalty mechanics, they need to be machine-readable and accessible to AI systems. MCP will likely become the bridge that makes this possible, helping brands surface the full context of their value at the very moment an AI is making the choice.

Agentic AI is still evolving, but its impact will scale quickly. Ark Invest projects that by 2030, agentic AI could drive $9 trillion in e-commerce sales, representing a quarter of global transactions.

At Talon.One, our advice is clear:

Here’s a quick checklist to evaluate your readiness:

If you can’t tick all these boxes yet, now is the perfect moment to start.

Agentic AI will change retail in ways we haven’t seen since the rise of e-commerce. But this shift isn’t something to be nervous about; it’s a chance to do things better.

The brands that lean into it will be the ones who: