4 minutes
How small businesses can prepare for agentic commerce
Enterprise merchants are already building for AI agents. Small businesses must keep pace with smaller budgets and thinner margins for error. The answer is not to copy enterprise infrastructure, but to make agentic commerce safer and easier to adopt.
Amit Dommeti
Senior Product Manager, Agentic Commerce
Amit Dommeti is a senior product manager for agentic commerce at Worldpay, now Global Payments, working on AI-driven purchasing and merchant readiness tools.
Key points
- Agentic commerce means AI agents are part of the buying journey. An agent may help a shopper compare products, build a basket or complete checkout, even if the customer still confirms the final purchase.
- Small businesses need to prepare for AI-driven discovery and automated fraud risk. Agents need accurate product, price and availability information to select a merchant. Merchants, in turn, need ways to distinguish trusted shopping agents from suspicious automated activity.
- The best starting point is low-risk, measurable use cases. Begin by using AI in controlled, internal tasks before applying agentic tools to customer-facing transactions.
For many small businesses, getting online over the last fifteen years became possible because software got easier to use – a website builder, a payments button, a point-of-sale app that worked right out of the box. Agentic commerce is the next chapter of that story, except this time it isn't business owners just learning new tools. They are preparing for AI agents to become a part of how customers discover, compare and eventually buy.
This direction is becoming clear even though the technology, standards and models are still developing. The practical question for a small business is not whether to rebuild everything today but rather look at what getting ready means without an enterprise budget behind you.
What is agentic commerce, exactly?
Agentic commerce, defined simply, describes a purchase journey in which an AI agent helps discover, compare, select or buy products and services. It doesn't need to be entirely automated, with an agent doing everything from start to end. A shopper might ask an agent to compare a few products, or hand over the entire cart, checkout and delivery decision. Agents can plug into any part of the flow.
There are two ways this shows up right now. In a human-present journey, the shopper reviews or confirms the transaction; the shopper stays in the loop and authenticates within the AI platform to complete checkout. In a human-not-present transaction, the agent handles payment and authentication on its own – more autonomous models are still developing and are not yet a routine experience for most small businesses.
Why small merchants can't sit this out
Nearly half of shoppers around the world – 40% – say they'd let an AI bot browse on their behalf, according to Worldpay research. Global Payments' 2026 Commerce and Payment Trends Report names agentic commerce as one of six trends reshaping commerce broadly, not a niche experiment confined to a handful of large retailers.
For small merchants, discoverability is the whole game. If an AI agent doesn't find you, it won't buy from you. That's the same logic that has driven search visibility for two decades, applied to a new kind of search.
"Discoverability is the whole game. If an AI agent doesn't find you, it won't buy from you."
That shift starts with data, not technology. Agents rely on the same signals search engines have used for years: accurate product names, pricing, availability, delivery terms and structured data. The difference is that information now has to be readable by a machine deciding what to recommend, not just a person deciding what to click. A small business with a clean, current product feed has a reasonable chance of being surfaced by an agent. One with outdated listings or inconsistent pricing across channels won't be, no matter how good the product is.
Checkout automation for agents is still maturing. Discovery is not – it's already where merchants win or lose visibility. Getting the product feed right is the cheapest, highest-impact step any online business can take today.
Three barriers unique to small businesses
Most agentic commerce tools are built for enterprise merchants. Small businesses run into three problems specifically.
- Technical readiness: Most agentic solutions assume access to an MCP server or current APIs. Small merchants often rely on commerce platforms or software providers to provide these capabilities – many do not have these in-house.
- Budget for experimentation: Large retailers can test agentic tools without an immediate return. Small businesses generally can't spare that budget.
- The cost of failure: A small increase in fraud is an operating cost a large retailer can absorb. For a small merchant, it can have a disproportionate effect on cash flow and customer trust.
That risk isn't hypothetical. In a controlled demonstration, our fraud solutions engineering team built a working card-testing fraud bot in about an hour, using roughly 300 lines of code – a rough measure of how little it takes to target a merchant that isn't watching for it.
Small businesses do not need another technology stack
Small business owners have already invested in getting discovered and then running their operations – search, marketing, their product catalog, point-of-sale system. It is unrealistic to expect a merchant to fund separate integrations or vendors for agent verification, fraud, reporting, etc. For agentic commerce to work for small businesses, the complexity must be absorbed by the providers they already use and trust. Agentic commerce should not feel like a net new offering entirely but rather another medium in which they can still grow their order volume and customer base.
"For agentic commerce to work for small businesses, the complexity must be absorbed by the providers they already use and trust."
The winning model for small businesses should be bundled and outcome-oriented: faster onboarding, fewer integrations, transparent pricing and clear evidence that these new channels will produce incremental sales.
The case for optimism
I keep coming back to what happened with e-commerce itself. Software made it possible for nearly anyone to open an online store, not only large retailers with in-house development teams. Shopify, Canva and Instagram did that work. I expect agentic commerce to follow the same path, with small merchants gaining a shop on ChatGPT or Gemini as easily as they built a website a decade ago.
What has to change to make this fair for small merchants
Three things matter most for small and mid-sized business owners.
- Liability: Today, if a customer says an order wasn't what they wanted, the default liability tends to fall on the merchant. Those rules need to be revisited as agents come between merchants and shoppers, leaving merchants with less of the data they'd normally use to make their case. In the meantime, current attended transactions – where a human confirms the order before it completes – keep liability where it has always sat, with the shopper, until the industry agrees on new rules for other setups.
- Cost: Agentic tools built for enterprise merchants often carry enterprise pricing: per-seat licenses, custom integration fees, volume minimums. Small merchants need pricing that scales down with them, not a business case built for a nine-figure retailer.
- Set-up: In practice, implementing agentic tools can't take a tech team to set up, and they shouldn't require an in-house IT person just to keep running. This is a problem for the payments industry to solve broadly, not something any one company can solve alone.
What we're building for this
Our role at Worldpay is to help give small business owners time back. They should be able to focus on building great products, serving their customers and growing their businesses, rather than piecing together separate tools.
We’ve built agent verification and bot detection. It's not something we're bolting on from a separate vendor. Today, that means flagging agentic orders and surfacing an agent's name, provider and verification status – distinguishing a legitimate shopping agent from an unverified bot before an order completes.
Where to start
Start with one low-risk workflow. Use AI to improve product descriptions, summarise customer questions, organise internal knowledge or identify gaps in product information. Keep a person reviewing the output.
Make your store easier for agents to understand. Keep product names, prices, availability, delivery terms and return policies accurate and consistent across channels.
Ask your providers the right questions. Can your commerce platform identify AI-referred traffic? Can your fraud tools distinguish trusted agents from suspicious bots? Will agentic transactions appear separately in reporting and disputes?
Choose a contained commerce test. Start with a low-value category, a defined customer journey or a small product range. Measure conversion, order value, fraud, cancellations, returns and support contacts.
Agentic commerce FAQ
Do small businesses need to prepare for AI agents making purchases from their store?
Who is liable if an AI agent buys the wrong thing from a small merchant?
Can small businesses afford agentic commerce tools built for enterprise merchants?
What's the difference between agentic commerce and AI tools that help run a small business?
How can a small merchant tell a legitimate shopping agent from a fraud bot?
Related insights
