Editorial
Research Report
Business Leaders CX Report: What’s Working and What’s Not in AI-Driven CX
The Gist
- Agentic commerce is closer than many think. Retail leaders describe a three-phase shift from AI-enhanced storefronts to buying inside chat interfaces, and eventually autonomous agent-to-agent purchasing.
- Trust and clean data will decide winners. Consumers remain cautious about AI making purchases for them, while brands need governed, enterprise-wide "certified data assets" before autonomous commerce can scale.
- Stores are still the main battleground. Even as agentic commerce grows, physical retail remains dominant, and many leading retailers are using AI right now to improve in-store service, inventory accuracy and associate productivity.
If you spend any time reading about the customer experience and AI technology space right now, it's easy to find lots of folks who are losing their minds over the promise of agentic commerce. You know the idea that our own autonomous AI agents will seamlessly run our lives and do all of our shopping for us.
Now, many people find this idea completely intoxicating.
Me. I'm not sure.
But I'm always willing to learn more about a subject so I can better understand what is going on. To help me with that, I was recently lucky enough to spend some time with Morgan Seybert, Head of Retail at Tredence, who works with 8 out of 100 of the world's largest retailers.
Seybert believes that we are witnessing three distinct phases of AI retail adoption:
-
Currently, we are in Phase one, where retailers are fundamentally re-envisioning their digital storefronts to be far more personalized and AI-enabled, utilizing generative AI assistants and dynamic pricing.
-
Phase two marks the move toward interacting with and purchasing directly through Large Language Models (LLMs) such as OpenAI's ChatGPT or Google's Gemini. In this phase, we will see ads integrated into the chat experience, and consumers will be able to transact directly through the LLM interface without ever visiting a brand's traditional digital property.
-
Phase three, which Seybert describes as the holy grail, will see fully automated, agent-driven purchasing. This is the era of agent-to-agent commerce. For example, in phase three, customers will build and deploy their own personal AI agents that deeply understand their lifestyles, preferences and desires.
Related Article: OpenAI's ChatGPT Instant Checkout: The Dawn of Conversational Commerce CX
Table of Contents
- When AI Shops the Super Bowl Party for You
- Agentic Commerce Challenges
- What Happens to Old-Fashioned Shopping?
- Agentic AI Inspires Better In-Store Experiences
When AI Shops the Super Bowl Party for You
To bring this final phase to life, Seybert recounted the following scenario to me: Imagine you are hosting a Super Bowl party for eight people. Your personal AI agent will already know exactly what food is in your kitchen cupboards and refrigerator. It will understand your and your guests' dietary needs, tastes and the recipes you like to make. Armed with this information, your personal agent will identify and independently approach a retailer's AI agent, converse with it, share information, negotiate and complete the purchase on your behalf with virtually no human clicks required.
Now, to some folks, they may sound like something out of a sci-fi movie, but the reality is that the capability and possibility are almost here.
And it could represent a significant market and opportunity.
I was talking about this recently with Katja Forbes, author of Machine Customers. She pointed me to a recent piece of research from WorldPay on agentic commerce. Quoting the research, she told me that WorldPay predicts that within the next five years, 9% of online purchases in the US consumer market will be made through AI agents. With the US ecommerce market projected to reach $2.9 trillion in 2030, that translates to $261 billion of AI-driven spend in the US alone.
Forbes then noted that these figures only represent the B2C market, and that the B2B market, while less visible, is likely to be "much, much larger."
Now, if we step back and reflect on these phases, it's easy to see the potentially significant upsides of this agentic future for many consumers, including greater ease and convenience, as well as reduced information overload. Consumers will no longer have to shop using keywords, they will shop based on descriptions of their life missions and intents.
And, they'll benefit from the fact that AI agents are likely to be pretty resilient to manipulation, and will stay focused on the trade-offs that need to be made across prices, products and availability in order for them to deliver value.
That sounds pretty good. But this type of future will also bring its own challenges and hurdles, particularly for brands.
Agentic Commerce Challenges
Will You Give AI Your Wallet?
The first challenge brands will need to navigate is trust. Handing over your wallet and personal data to an autonomous machine requires a profound leap of faith. Currently, research suggests that consumers are comfortable letting AI handle routine, low-risk tasks, but they are hesitant to let an AI agent make final purchasing decisions for them or manage sensitive personal information.
For example, CSG's 2026 State of Customer Experience report found that 56% of consumers are uncomfortable letting AI take actions on their behalf, and that number rises to 81% among consumers over the age of 62. These findings are mirrored by the aforementioned WorldPay agentic commerce report, which finds that younger consumers, who are more familiar with digital assistants and generative AI, are far more comfortable delegating purchasing decisions. Specifically, they found that 50% of those aged 18–34 are open to AI shopping, compared with just 27% of those over 55.
Are You Worried Over Data and Hallucinations?
The second big hurdle revolves around a brand's data.
Over the years, organizations have created conflicting data assets. For example, according to Seybert, a marketing team might define "sales" differently from the finance team. This could result in conflicting versions of the same metric.
Now, on the surface of things, that may not seem like such a big deal. But, for an AI agent, that's a problem. Why? Well, if the AI agent doesn't know which version of the truth to use, it could cause it to hallucinate.
Therefore, before you can deploy agentic commerce, Seybert recommends that brands build what he calls "certified data assets" that are strictly governed, quality-controlled and aligned across the entire enterprise. If you don't do this and try to deploy AI without them, Seybert believes that would be akin to trying to build the roof of a house before you've even poured the foundations.
Will Your Marketing Be Agent-Ready?
In addition, brands need to understand that marketing to AI agents differs significantly from marketing to humans. Modern websites are designed to attract and engage human visitors. However, to perform effectively in an agent-to-agent future, brands will need to adapt their infrastructure to make it machine-readable as well as human-readable.
What Happens to Old-Fashioned Shopping?
But, let's assume that brands get their data house in order and build a strong bond of trust with their customers.
In that scenario, I'm left wondering what happens to "shopping," which many consumers in many countries really enjoy. In fact, in some places, shopping is a national pastime.
Moreover, when you add to the mix evidence that suggests that some customers are tiring of digital and are seeking more personal, in-store and face-to-face experiences, things get complicated.
I asked Seybert about this, and he pointed out that the majority of retail transactions still take place in-store, and this is unlikely to change any time soon. This is supported by research from eMarketer, which finds that ecommerce currently accounts for approximately 20.5% of global retail sales, up from 19.9% in 2024 and is only projected to reach 22.5% by 2028.
Webinar
May
12
How Marketing Teams Can Take Control of the Web Experience
See how Veo and Kvalifik put marketing in the driver's seat, shipping faster and growing qualified bookings by 150%.
Register
Webinar
May
14
Two Audiences, One WordPress. Is Yours Ready?
The practical steps to making your WordPress portfolio ready for both human visitors and AI agents.
Register
Webinar
May
19
From Content Sprawl to Competitive Advantage: A Knowledge Operations Roadmap
How automation, AI and a centralized content platform turn knowledge fragmentation into a scalable competitive advantage.
Register
Webinar
May
20
From AI Investment to CX Results: What Enterprise Leaders Need to Know
Move beyond experiments. See how top enterprises scale AI for CX results.
Register
Webinar
On demand
Where Contact Center AI Is Actually Headed
A straight read on where the contact center platform market is going and what it means for your next technology decision.
Watch Now
Webinar
On demand
AI for Your DXP: Connect What You Have, Transform How You Work
Most AI strategies stop at the platform—but work happens elsewhere. Bring intelligence into Teams, email, tickets and CRM.
Watch Now
Agentic AI Inspires Better In-Store Experiences
However, he also pointed out that while agentic commerce is getting a lot of headlines right now, many progressive retailers are using AI not only to rethink how they will play in an agentic commerce future, but also to transform their in-store experience.
To give a flavor of what three leading retailers are doing and to illustrate the impact that AI is having on the in-store experience, Seybert shared three examples with me:
Matching Fabric for Sofa — in 2 Minutes
The first featured a retailer in the furniture and home goods sector that struggled when a customer brought a fabric swatch into the store and asked whether they could have a sofa built covered in the fabric that they had brought in. Previously, it would take a store associate around two hours to manually create a custom SKU to accommodate such a request. However, by embedding a generative AI assistant into the store associate's app, the retailer reduced this process to under two minutes, achieving a 98% efficiency gain. Customers can now see an image of their custom product in seconds and visualize how it will look alongside their previous purchases, which significantly reduces purchase friction and cuts down on expensive in-store labor.
Get the Golf Swing — and Customer Data — in Order
The second example featured a large sporting goods retailer that has been leveraging the vast amount of data generated by its in-store golf swing simulators. Rather than treating the initial club fitting as a one-time transaction, the retailer is using AI to analyze the golfer's swing speed, handicap level and specific golf style. When the customer returns to the store months later, store associates can access this AI-driven data to provide highly personalized "clienteling" recommendations, such as identifying the perfect golf ball or training aid tailored to that individual's mechanics.
Bots Finding Missing Store Items
The final example featured a major retailer operating over 10,000 stores, suffering from out-of-stock rates exceeding 12%, which was costing them up to 5% in lost sales every year. The problem was that their traditional inventory data showed items were in the store, but the products were actually sitting in the store room rather than on the shelves where customers could buy them. To solve this, the retailer implemented an AI tool that analyzes computer vision data and historical sales patterns to direct store associates exactly where to find missing items. The store assistant app now alerts the associate to the specific out-of-stock product and points them to the exact shipment and tote in the back room. This initiative reduced out-of-stocks by 30% and drove a 2% bump in top-line sales.
Those are some pretty impressive results.
While I think it will be fascinating to watch how agentic commerce evolves from phase one through to phase three, I’m just as excited to see how AI will enhance not only in-store retail experiences but also retail operations. After all, it still is the biggest playing field.
Learn how you can join our contributor community.