XAgent is the open execution market for Agentic Commerce. It turns merchant products, pricing, inventory, order logic, settlement, and fulfillment capability into commerce interfaces that AI agents can read, call, and purchase from. Agents quote, authorize, settle, fulfill, and prove the purchase — machine-to-machine.

What is agentic commerce?

Agentic commerce is a model where AI agents — not humans clicking through web pages — discover services, place orders, and settle payments autonomously. XAgent makes this real through five execution pillars: Quotable, Authorizable, Settleable, Fulfillable, Provable.

How is XAgent different from Stripe or Coinbase Commerce?

Stripe was built for humans clicking checkout buttons. Coinbase Commerce is a crypto payment gateway. XAgent is the protocol-native execution market where AI agents transact: protocols standardize the request, XAgent operationalizes the transaction. The moat is merchant operations — catalog, quote, authorization, order, settlement, delivery, receipt, refund, reconciliation, fraud signal, jurisdiction metadata.

How do AI agents pay merchants on XAgent?

Agents pay using the x402 protocol — an HTTP 402 challenge/response standard. The agent receives a signed payment instruction, authorizes a stablecoin transfer (USDG or USDT) via EIP-3009, and the merchant is paid in seconds with zero gas fee to the buyer (OKX facilitator covers it).

Do I need to write code to list my store?

No. One-click OAuth in your Shopify or WooCommerce admin turns your store into a merchant capability that AI agents can quote, order, pay, settle against, and prove — zero code, zero contracts.

Agentic commerce numbers live in earnings calls, not filings

Shopify's tripled AI traffic is in no filing. Etsy's under-1% is a spoken aside. The one agentic number that reached an audited statement this quarter was a cost, and it belongs to a smaller company.

By XAgent Team · 2026-08-13

The most quoted proof point in agentic commerce right now is that AI-driven traffic and orders to Shopify stores tripled year over year. It is a striking number and it appears nowhere in Shopify's second-quarter press release — we searched the full text for "tripled", "3x", "AI-driven traffic", "AI-attributed" and "AI-referred" and got zero hits on every one. It was said on the call. That is the pattern across this earnings season, and it is worth naming before the next round of forecasts is built on top of it.

What is filed and what is spoken

A number in a press release or a 10-Q has been through disclosure controls. A number offered on an earnings call has been through a slide deck. Both can be true; they carry different weight, and right now almost every agentic figure in circulation is the second kind.

Shopify's tripling is call-reported. Etsy's chief executive said on Etsy's call that traffic from AI agent platforms remained under 1% of the company's total traffic — the only denominator any large marketplace has volunteered, and also spoken rather than filed. DoorDash's chief executive told analysts agentic order volume from AI partners remains low. None of these companies is doing anything improper. But an industry-wide adoption case is being assembled almost entirely out of remarks.

The asymmetry matters because of what these numbers omit. "Tripled" without a base is unfalsifiable: a tripling from a tenth of a percent and a tripling from ten percent are the same sentence. Etsy's under-1% is the only figure in the set that tells you the size of the thing being multiplied, and it is small.

We made this argument about the published volume figures for agent payments, where four widely-quoted numbers turned out to measure four different things. The equity story has the same defect from a different direction: not incompatible measurements, but no measurements at all — just growth rates over an undisclosed base.

The one agentic number that did reach a filing was a cost

On August 6, Commerce.com, Inc. (Nasdaq: CMRC) — the parent of BigCommerce and Feedonomics — filed its second-quarter results. The figures below are from the press release furnished as Exhibit 99.1 to its Form 8-K, so unlike the rest of this discussion they are filed, not spoken.

Revenue was $84,511 thousand against $84,433 thousand a year earlier: flat. Over the same period, cost of revenue rose from $17,739 thousand to $20,982 thousand. Gross profit fell from $66,694 thousand to $63,529 thousand. In the company's own summary, "GAAP gross margin was 75%, compared to 79% in the three months ended June 30, 2025."

Meanwhile "Gross Merchandise Volume (GMV) was $8.8 billion, up 14% compared to the three months ended June 30, 2025."

Read those together. Merchandise moving through the platform grew 14%. Revenue did not move. The cost of serving that flat revenue grew by more than three million dollars a quarter. That is a monetisation gap and a cost line widening at the same time, on a filed statement.

Be careful about the explanation

Here the discipline has to apply to us as well.

The press release does not say why cost of revenue rose. We searched it for "crawler", "bot traffic", "hosting", "infrastructure cost" and "margin compression": zero hits for each. The company's own filed document reports the compression and does not attribute it.

The attribution — that the increase came substantially from AI crawlers and agents retrieving product data from merchant storefronts, and that the company is choosing to keep storefronts broadly accessible anyway — comes from management's remarks on the earnings call, reported by trade press. It is plausible, it is consistent with the company's stated strategy of catalog syndication to agentic channels, and it may well be right. It is also, precisely, another spoken number about agents.

So the honest version is narrower and still useful: a public commerce platform reported flat revenue, 14% GMV growth, and four points of gross-margin compression in the same quarter, and the explanation on offer for that compression is agent traffic.

What a merchant should take from this

Three things, and none of them require picking a side on whether agentic commerce is early or overhyped.

Ask for the denominator. When a platform, a processor or a vendor quotes you a growth multiple on agent traffic, the useful question is what it grew from and over what period. Etsy answered it. Almost nobody else has.

Assume agent traffic has a cost before it has a revenue. Whatever the cause of Commerce.com's margin, the shape is one merchants should plan for: agents read a lot more than they buy. Crawling, catalog retrieval and repeated product lookups are served at your expense, and the order that justifies them arrives later, if at all. The gap between being discoverable and being executable has a bill attached.

Instrument your own side. The reason nobody can give a denominator is that agent-originated orders are hard to attribute after the fact — a session that started in a model and ended on your site looks like direct traffic. If you want a number you can defend, it has to be produced at transaction time: an identified agent, a quote issued to it, an order tied back to that quote. That is a property of how the transaction is executed, not something a analytics tool can reconstruct afterwards.

What's next

Watch for the first company to put an agent-originated GMV figure in a filed document rather than a spoken one. That will be a more informative event than any of this season's multiples, because it will come with a definition attached.

Until then, treat every agentic growth rate as a claim about an undisclosed base, and treat the cost side as the part already showing up in the accounts.

If you would rather your agent traffic arrived identified, quoted and attributable — so that the number you report is one you can reconstruct — list your store on XAgent and let the open execution market produce the record.