2026-08-09 · protocol governance · UCP

Ten seats at the table: what UCP governance means for an independent store

Companies that compete on everything else now share one governance body for agentic commerce. That convergence is real and it matters. It also does not answer the question an independent WooCommerce merchant actually has.

TL;DR — On 24 April 2026, Amazon, Meta, Microsoft, Salesforce and Stripe joined the Universal Commerce Protocol Tech Council, alongside founding members Google, Shopify, Etsy, Target and Wayfair. Ten seats, of a body reported to hold up to sixteen. The consolidation is genuine. But UCP governance is a table for platforms, and WooCommerce was never a UCP launch partner — independent stores implement it themselves, or not at all. For a self-hosted merchant the practical question is unchanged: is your catalog readable and verifiable, on your own domain, before any protocol arrives?

What actually happened in April

UCP was announced by Google at NRF in January 2026, co-developed with Shopify and a group of large retailers, as an open standard covering the full shopping journey: product discovery, cart building, checkout and post-purchase, across platforms and payment processors. It is transport-flexible by design — REST, A2A and MCP are all described as ways in. The Tech Council is the body that steers its technical direction.

In April that body doubled. Amazon, Meta, Microsoft, Salesforce and Stripe took seats next to Google, Shopify, Etsy, Target and Wayfair. Read the list by function rather than by brand and the shape is striking: search, the largest marketplace, social surfaces, enterprise commerce tooling, cloud and assistant agents, payments, and major retail. Companies that fight each other daily, sitting at one governance table.

Amazon's presence is the most surprising line item. Through 2025 Amazon was actively restricting AI crawlers from its platform. Joining the council that governs how agents transact is a different posture than blocking them at robots.txt.

The part everyone reads wrong

Ten days before that announcement, WooCommerce 10.7 shipped an agentic_commerce payment gateway flag, first used by the Stripe implementation of ACP — the rival protocol maintained by OpenAI and Stripe. Much of the coverage framed this as "WooCommerce backs ACP" and then the April news as "and then ACP lost".

Neither framing survives contact with the detail. The gateway flag is protocol-agnostic: it is a surface any agentic protocol can plug into, and ACP happened to be first. And Stripe now sits on both sides — co-maintainer of ACP, council member of UCP — for a reason that has nothing to do with loyalty. Stripe earns when an order completes, regardless of which protocol delivered it.

That is the correct merchant posture too. You earn when the order completes. Protocol allegiance is a question for platforms with a stake in the standard; a store's stake is in being reachable by whatever arrives.

Where independent stores actually sit

Here is the fact that gets omitted from the enthusiasm. WooCommerce was not a UCP launch partner. Shopify co-developed the protocol and, as a platform, can switch it on for millions of merchants at once — its merchants inherit agentic capability as a platform feature and are left with a data-quality problem rather than an integration problem. A self-hosted WooCommerce store inherits nothing. It implements, or it stays outside.

This is the structural asymmetry of platform-governed standards, and it is not malice — it is arithmetic. A council of ten large platforms optimizes for what large platforms can deploy. Roughly a third of the world's online stores run on software with no seat at that table.

The layer no council can hand you

Say UCP wins outright, or ACP does, or they coexist. In every branch, an agent still has to answer the same prior questions: does this store exist, what does it sell, is this item in stock, at what price, in which variant, and can that claim be trusted. Protocols standardize the transaction. They assume the catalog.

We made this argument at length in UCP and ARC: two layers, not two rivals, and nothing in the April news changes it — if anything the ten-seat council sharpens it, because a governance body converging on checkout leaves the catalog layer exactly where it was: the merchant's problem, on the merchant's domain.

That is the layer KaliCart Bridge builds, and it is deployable today by a store with no seat anywhere: discovery documents, bounded read-only catalog search, honest price and stock blocks, live WooCommerce data rather than a copy. It is also the layer that survives whichever way the protocol contest resolves, because it is not competing in that contest. See also our reading of the checkout-side standard in ACP for WooCommerce.

What we would watch next

Two things, without predicting either. First, seats: reporting suggests the council has room for up to sixteen, so the composition is not final and who joins next says more than any press release. Second, the gap between governance and adoption — a protocol can be universally governed and still thinly implemented, and merchant-side reality has consistently lagged the announcements by more than the announcements suggest. We will believe the convergence when independent stores can act on it, not when the table gets longer.

Sources