The shift in online shopping already happened. You just may not have seen it yet.

Yesterday, your customer clicked “Buy Now.”

Today, they say to an AI agent:

“Reorder my skincare and get it to my location by Friday. Use my saved card.”

Then they close the laptop.

No browsing product page. No adding products to cart or proceeding to checkout.

An AI agent makes the decision. The AI agent pays.

You either get the order or you don’t.

That is the reality of where e-commerce is heading. This Google examples explain what the industry is adapting to.

While most brands are still optimising landing pages and tweaking checkout buttons, the real battle is happening somewhere else entirely.

Right now, the two protocols shaping that future of agentic commerce are:

• ACP (Agentic Commerce Protocol)
• UCP (Universal Commerce Protocol)

They are not just technical standards. They are the infrastructure behind how AI agents discover, evaluate, and choose products.

Of course, the real question is not just what are they? It is: Where does your brand show up when these agentic commerce protocols make the decision?

What is Agentic Commerce?

Agentic commerce is when people stop shopping manually and start delegating the process to AI. Instead of browsing websites, comparing tabs, and reading reviews, they say what they want and the AI handles the rest. That includes:

• Finding products
• Comparing options
• Making decisions
• Completing purchases

In other words, the AI becomes the shopper.

This shift removes friction, but it also removes visibility. Your beautifully designed website matters less if no human ever lands on it.

And this is not a fringe trend. According to Gartner projections, agentic commerce could power up to 33% of enterprise transactions by 2028.

What is ACP (Agentic Commerce Protocol)?

ACP is designed for one thing: getting transactions done inside AI conversations. It is closely tied to systems like OpenAI and payment infrastructure from Stripe.

Think about the experience. A user is chatting with an AI assistant. They ask for a product. The AI suggests an option. The user agrees. The purchase is completed — all without leaving the conversation. That is ACP.

Why ACP Matters

ACP is built for speed and simplicity. It works best when:

• The user already knows what they want
• The AI has enough context to make a confident decision
• The purchase does not require heavy comparison

Think repeat purchases, subscriptions, or everyday items. If someone trusts the AI, they do not want to browse. They want the outcome.

The Limitation — and What Has Changed

ACP operates in a more controlled ecosystem. That means if your brand is not integrated within that environment, you are not even considered. There is no second page of results. There is no “maybe later.”

That limitation became more visible in early 2026. By March, OpenAI had begun revamping ChatGPT shopping away from Instant Checkout — the first major ACP product expression — toward merchant-controlled checkout experiences, citing a lack of flexibility. Analysts pointed to challenges around vendor onboarding, product data accuracy, multi-item carts, and loyalty programmes.

Walmart’s experience was pointed: after offering approximately 200,000 products through OpenAI’s Instant Checkout, purchases completed inside ChatGPT converted at one-third the rate of those redirected to Walmart’s own website. Walmart EVP Daniel Danker called the experience “unsatisfying.”

ACP remains active and may continue to evolve. But its first major product moment stumbled on real-world commerce complexity.

Stripe, which co-built ACP with OpenAI, has now joined the UCP Tech Council. That alone tells you something about where infrastructure players are hedging.

What is UCP (Universal Commerce Protocol)?

UCP is an open standard backed by Google and supported by major commerce players including Shopify, Walmart, Amazon, Microsoft, Meta, Salesforce, and Stripe. It connects deeply with systems like Google Gemini and is designed to cover the entire shopping journey — from discovery and comparison through to post-purchase support.

Here is how it feels: a user asks for a product. The AI pulls options from multiple sources, compares them, then helps complete the purchase.

Why UCP Matters

UCP is built for discovery and comparison. It works best when:

• The user is exploring options
• There are multiple competing products
• The decision requires evaluation

Think high-ticket items or first-time purchases. If ACP is about execution, UCP is about exploration and execution combined.

UCP’s design separates commerce into layers — core transaction primitives, capabilities such as checkout and order management, and extensions for domain-specific needs like loyalty programmes. A merchant publishes what it can do. An agent discovers that profile and negotiates the interaction. This means UCP does not try to force every merchant into a single checkout model.

The April 2026 Shift

On April 24, 2026, Amazon, Microsoft, Meta, Salesforce, and Stripe joined the UCP Tech Council, the governance body steering UCP as an open standard. They joined founding members Google, Shopify, Etsy, Target, and Wayfair, creating a ten-member council spanning search, payments, social commerce, enterprise software, and retail infrastructure.

The significance is not just the names. It is the breadth. The companies that process, route, host, and monetise the majority of global e-commerce have chosen to sit inside the same governance room. That is the clearest signal yet of where the open standard is heading.

The Limitation

Because UCP is broader, it is also more competitive. You are not just being chosen — you are being compared. If your data, pricing, or reputation is weaker, you lose the decision.

ACP vs UCP — Key Comparison

Which Should Brands Bet On?

This is not a decision to choose one or leave one behind. If you are thinking “Should we choose ACP or UCP?” you are already behind.

The right question is: where does each protocol fit in our customer journey?

Because your customers are not choosing one system. They move between both. They might ask an AI assistant first, get a recommendation, validate it through search, then let the system complete the purchase. If you only exist in one of those moments, you lose the other.

In this new environment, losing visibility means losing revenue completely.

Which Should Retailers Choose First?

If your audience is younger or AI-native, start with ACP. These users trust AI assistants, prefer speed over browsing, and are comfortable delegating decisions. They are already using conversational systems as their first step.

If your audience is broader or more traditional, start with UCP. These users still compare options, use search heavily, and want validation before purchase. UCP meets them where they already are.

If you want to win long-term, build for both. A user might start in one system and finish in another. They might trust AI for speed but rely on search for confirmation. The brands that win are not the ones who choose one channel — they are the ones who show up everywhere the decision happens.

Why This Matters More Than You Think

Before, you optimised for traffic, clicks, and conversions. Now you optimise for selection. And selection happens before the user ever sees your brand. That is a completely different game.

When W3C’s HTML and HTTP standards stabilised in the 1990s, businesses that built clean, structured, crawlable sites early gained durable search advantages that took competitors years to close.

The UCP governance consolidation may be setting up a similar dynamic.

Merchants or retailers whose systems are agent-readable from the start stand to capture a disproportionate share of AI-driven purchases before the broader market catches up.

How to Prepare Now

1. Structure Your Product Data

Agents do not read marketing copy. They read structured data. If your product is described as “BEST TEE EVER!!!” that means nothing to an AI. Instead, you need clear product names, standardised attributes, and machine-readable formats. If your specs are buried in PDFs or inconsistent across pages, you lose visibility.

2. Expose Real-Time Inventory and Pricing

Trust is everything in agentic commerce. If an AI agent completes a purchase and hits an “out of stock” error, that breaks the system. Agents expect real-time inventory, accurate pricing, and clear shipping costs and timelines. If you cannot provide that, the AI will simply choose another brand that can.

3. Build Entity Authority

AI agents evaluate reviews, mentions, brand authority, and consistency across platforms. If two products are similar, the agent chooses the one with stronger trust signals. Your reputation is no longer just about perception — it directly affects whether you get selected.

4. Run a Test Transaction

Do not assume your system works. Test it. Use an AI agent and ask it to buy from your site. Watch carefully: can it find your product? Does it understand your offering? Where does it get stuck? Can it complete the checkout? Every point of friction is a lost opportunity. Fix those now.

FAQs

1. Do I need to implement ACP and UCP immediately?

Not immediately, but you need a clear roadmap. Start where your audience is most active, then expand.

2. Will traditional e-commerce websites become irrelevant?

No, but they will become infrastructure rather than the primary experience. AI agents will handle most of the interaction.

3. How is this different from regular SEO?

SEO helps users find you. Agentic commerce ensures AI can choose and transact with you.

4. What happens if I ignore this shift?

You risk becoming invisible. If AI agents cannot access or understand your products, they will not recommend or purchase them.

5. Is this only relevant for large brands?

No. Smaller brands that adapt early can gain a significant advantage by becoming default choices in AI-driven decisions. The April 2026 governance consolidation makes early action more — not less — important.

FAQs

1. Do I need to implement ACP and UCP immediately?

Not immediately, but you need a clear roadmap. Start where your audience is most active, then expand.

2. Will traditional e-commerce websites become irrelevant?

No, but they will become infrastructure rather than the primary experience. AI agents will handle most of the interaction.

3. How is this different from regular SEO?

SEO helps users find you. Agentic commerce ensures AI can choose and transact with you.

4. What happens if I ignore this shift?

You risk becoming invisible. If AI agents cannot access or understand your products, they will not recommend or purchase them.

5. Is this only relevant for large brands?

No, in  fact, smaller brands that adapt early can gain a significant advantage by becoming default choices in AI-driven decisions.

Final Thought

A deep thought! Can AI agents buy from you at all?

Because when someone says:

“Order it for me.”

The AI will not browse ten websites.
It will not compare endlessly.

It will choose one brand.

And once that choice becomes a habit, it becomes the default.

If you are unsure whether your business is ready for this shift, do not wait until you start losing transactions.

At DMi Agency, we help brands prepare for exactly this.

From structuring your product data to optimizing for AI-driven discovery and transactions, we make sure your brand is not just visible but chosen.