Merchant Participation Framework
The committability ladder
Agentic commerce is not one step. It is five: discovery, validation, commitment, execution, and settlement. Networks, payment service providers, and open protocols are advancing the surfaces around the merchant. The stage in the middle, where a local service authors the commitment it can make and keep, is still the merchant's own work.
Published June 28, 2026
The one-line takeaway
Identity, intent, cart, and payment execution are being built fast by surfaces and rails. None of them author the commitment a local service must stand behind. Committability is the merchant-controlled stage of the ladder, and it does not arrive from upstream.
- Published
- June 28, 2026
- Format
- Framework
- Sources
- 5 public sources
- Model
- Evergreen, not a news note
OpenAI, Stripe, Visa, Google, and any other company named here are public source subjects. Obenan has no partnership, integration, or endorsement with any of them.
The 60-second read
What the ladder is, and why the stages must stay separate
The committability ladder is a way to read agentic commerce as five distinct stages rather than one event. Most public progress sits at the ends of the ladder. The merchant-controlled middle is where local-service commerce is decided.
What are the five stages?
Discovery, validation, commitment, execution, and settlement. An agent first finds a business, then checks whether it can act, then needs a commitment it can rely on, then executes payment, then settles. Each stage is a different question with a different owner.
Why does collapsing the stages mislead operators?
When discovery, payment execution, and settlement are treated as the whole picture, the commitment stage disappears from view. That is the stage a restaurant, clinic, salon, or hotel cannot skip: whether a specific deposit, hold, cancellation window, or guarantee can actually be made and kept.
Which stage is the missing operator layer?
Commitment. The surfaces above and the rails below are advancing in public. Merchant-authored committability is the stage no surface or network can author on the merchant's behalf, and it is the stage Obenan helps a merchant own.
The model
The five stages of the committability ladder
Each stage asks a different question and has a different owner. Read top to bottom, the ladder shows where public infrastructure is strong and where the merchant-controlled gap sits.
Discovery
An agent finds the business and what it offers. Feed distribution, catalog exposure, and category expansion live here. This stage is increasingly well served by AI surfaces and shopping protocols, and it is owned mostly by the surfaces.
Validation
An agent checks whether it can act: is the agent legitimate, is the merchant a real participant, is the site navigable. Agent readiness scoring, agent and merchant directories, and identity signals live here. This stage is owned mostly by the networks and platforms.
Commitment
Before money moves, a commitment must exist: can this deposit be taken, this slot held, this booking cancelled inside this window, this service guaranteed at this time. This is a merchant-authored acceptance contract. It is the stage no surface and no network authors, and it is the missing operator layer.
Execution
Payment runs: authentication, tokenization, payment-method portability, and preservation of the merchant as merchant of record. Card networks, payment service providers, and open checkout protocols are building this stage quickly, and they own it.
Settlement
Funds settle and the obligation closes: fulfillment, reconciliation, refunds, and dispute handling. Whether settlement is clean depends on whether the commitment at stage three was real. A clean rail cannot settle a promise the merchant never authored.
Stages one, two, four, and five are being built in public by surfaces and rails. Stage three is the merchant's to author. The ladder makes the gap legible instead of letting it hide between discovery and payment.
What stage three requires
Six things merchant-authored committability needs
Committability is not a feeling or a marketing claim. It is a set of merchant-controlled facts and rules that must be true, current, and expressible to an agent before a commitment is safe to make.
These six are the substance of stage three. None of them is delivered by feed distribution, readiness scoring, or payment execution.
Catalog and service truth
Accurate, structured facts about services, hours, availability, and location scope that an agent can read and act on without guessing.
Eligibility
Whether a specific request is eligible right now: can this party size, time, service, or party type actually be accepted at this location.
Freshness
The facts must be current at the moment of the agent's request, not a stale snapshot, because a commitment made on stale truth is a commitment that breaks.
Bounded commitment object
An explicit, bounded statement of what is being committed: the deposit amount, the hold duration, the cancellation window, the guarantee, and its limits.
Acceptance and policy terms
The merchant's acceptance-side policy, including no-show, late, and deposit rules, expressed so an agent and a customer both know what was agreed.
Accountability
A way to stand behind the commitment after the fact, so settlement, refunds, and disputes resolve against terms the merchant actually authored.
Read it carefully
Four misreads the ladder helps you avoid
The upstream and downstream stages are genuinely advancing, and that is exactly where a confident narrative can lead a local operator to skip stage three. These are the misreads the ladder is built to prevent.
Discovery is not commitment. Being found and carted by an agent does not mean the local service has authored whether the specific booking can be made, held, or cancelled.
Validation is not commitment. An agent readiness score or a merchant directory entry measures whether an agent can act, not whether the merchant can keep the promise.
Execution is not commitment. A clean payment rail moves money for a commitment that already exists; it does not author the deposit, the hold, or the cancellation window.
Settlement is not commitment. Clean reconciliation depends on a real stage-three commitment; a rail cannot retroactively make a promise the merchant never declared.
The honest version keeps the stages apart: an agent can discover, validate, execute, and settle, and still be acting on a commitment the local service never actually authored.
Who owns which stage
Three ownership lanes across the five stages
Group the five stages by who can author them. Discovery sits with the surfaces. Validation, execution, and settlement sit with the networks, payment service providers, and protocols. Commitment sits with the merchant, by design.
Keeping the lanes separate is what turns an impressive upstream and downstream stack into a decision a local operator can act on: the commitment lane is theirs to own.
Discovery
Feed, catalog, cart, and category expansion across AI assistants, shopping protocols, and merchant channels, so an agent can find and assemble an order.
Platform and AI surfaces
Validation, execution, settlement
Agent and merchant verification, readiness scoring, payment authentication and tokenization, merchant-of-record preservation, and reconciliation.
Networks, PSPs, and protocols
Commitment
The deposit, reservation hold, cancellation window, fulfillment guarantee, and acceptance terms a local service must author, keep current, and stand behind.
The merchant, by design
Surfaces distribute discovery, rails secure validation, execution, and settlement. The commitment lane is the merchant's to author and keep.
Operator rail
What to do now, monitor, and not assume
A practical split for a senior operator at a local or multi-location service reading the ladder against their own readiness.
Do now
- Author your stage-three commitment terms explicitly: deposit rules, reservation and booking holds, cancellation windows, no-show and late policy, and service guarantees.
- Make the merchant facts agents read, including hours, availability, services, and location scope, accurate and policy-complete enough to be safe to fulfill, not only easy to discover.
- Decide where you stay merchant of record and how your acceptance-side policy is expressed, since every upstream surface preserves merchant-of-record but does not write your policy for you.
Monitor
- Whether open commerce protocols add explicit hold, deposit, or cancellation primitives for local-service categories such as booking and reservations.
- Whether agent readiness scores or agent and merchant directories ever attach to acceptance-side commitment rather than only site readiness and verification.
- Whether payment and orchestration suites move beyond enterprise retail and add merchant policy controls for booking, cancellation, deposits, and post-purchase accountability.
Do not assume
- Do not assume discovery, validation, execution, and settlement add up to committability for a local service.
- Do not assume an agent readiness score or a merchant directory represents merchant-authored commitment truth.
- Do not assume category expansion into local-service verticals is a complete committability layer for those verticals.
Obenan's point of view
Why the commitment lane is the operator layer, and its boundary
Every public move in agentic commerce confirms the same instinct: the stack is being built around the merchant, and merchant facts are becoming the input the whole ladder depends on. The committability ladder names where that input has to become a commitment, and it puts that stage where it belongs, with the merchant. Obenan's role is to help a merchant author and keep that stage true, not to own discovery, payment execution, or settlement.
The boundary with the AI Visibility and Discoverability lane is deliberate. That lane is about platform-owned graph truth: whether AI surfaces find, cite, and represent a business accurately. The committability ladder is about merchant-domain commitment truth: whether a specific promise can be made and kept. They are complementary and must not be collapsed. Being discoverable is a stage-one and stage-two question; being committable is a stage-three question, and only the merchant can answer it.
Evidence discipline
Observed, inferred, and watching
We separate what the public sources state from what Obenan infers and what we are still watching.
Observed
Public sources show identity, intent, cart, readiness, and payment-execution capability advancing across open protocols and networks: OpenAI moved its agentic commerce protocol into product discovery, Stripe broadened agentic payment methods and protocols, the Universal Commerce Protocol shipped versioned catalog, eligibility, and checkout contracts, and Visa expanded agent readiness and verification tooling. Each preserves the merchant as merchant of record.
Inferred
Obenan reads these moves as building the discovery, validation, execution, and settlement stages while leaving the commitment stage, the merchant-authored acceptance contract for local-service execution, unaddressed. This is Obenan's interpretation of an absence in announced scope, not a quoted vendor admission.
Watching
Whether any surface or protocol formalizes merchant-side commitment primitives such as deposits, reservation holds, cancellation windows, and fulfillment guarantees, and whether readiness or verification ever attach to acceptance-side commitment rather than site readiness.
What we do not claim
The claim boundaries on this framework
The named companies here are public source subjects only. This framework makes none of the following claims.
- 01
Obenan does not process payments, tokenize cards, route checkout, act as a payment service provider or acquirer, settle transactions, or own any payment protocol.
- 02
Obenan has no partnership, endorsement, certification, pilot approval, integration, or special access with Mastercard, Visa, Adyen, Google, OpenAI, Stripe, American Express, PayPal, or any company named here.
- 03
Merchant committability does not guarantee agent ranking, recommendation, transaction completion, revenue, or platform inclusion.
- 04
No current Obenan merchant is certified or committable by any public network or protocol; those specifications are still being released and do not yet cover service inventory or reservation semantics.
- 05
Open protocols and network surfaces named here do not, by themselves, author or validate merchant-side commitment for local-service execution in the way this framework describes; the commitment gap is Obenan's reading of publicly announced scope.
- 06
No vendor named here controls how an AI agent ranks, recommends, or transacts with a business, and this framework cites no protected, NDA, partner-sensitive, or non-public detail.
Keep reading
Related reading
See what an agent can actually commit to at your locations
The surfaces above and the rails below keep advancing. The commitment stage is yours to author. Start with what AI answers say about your locations today, then make the commitment a local service must keep current and yours.
OpenAI, Stripe, Visa, Google, and any other company named here are public source subjects. Obenan has no partnership, integration, or endorsement with any of them.
Sources
This framework generalizes evidence already published and sourced in Obenan's Merchant Participation briefings. The anchoring public sources below were each checked live on June 28, 2026. The named companies are public source subjects, not partners.
Primary public sources
- 1.Universal Commerce Protocol release v2026-04-08github.com · Released April 9, 2026 · Checked June 28, 2026
Standards-body versioned contracts for cart state, catalog, eligibility, signing, and totals, anchoring the validation and commitment-contract reading of the ladder.
- 2.Universal Commerce Protocol checkout specificationucp.dev · Living specification · Checked June 28, 2026
Open, developer-verifiable checkout contracts anchoring the execution stage, independent of any vendor press.
- 3.Visa: New AI, stablecoin, and token innovations at Visa Payments Forumusa.visa.com · Published June 10, 2026 · Checked June 28, 2026
Agent Score, an Agentic Directory, and a large transaction model, anchoring the validation stage of the ladder.
- 4.Stripe: agentic commerce and checkout expansionstripe.com · Published March 24, 2026 · Checked June 28, 2026
Payment-method portability and protocol breadth, anchoring the execution stage of the ladder.
- 5.Google: Shopping updates from Google Marketing Liveblog.google · Published May 20, 2026 · Checked June 28, 2026
Universal Cart and category expansion into hotel booking and local food delivery, anchoring the discovery stage of the ladder.
Several anchoring sources carry no single fixed publication date as living specifications or product pages, so the dated anchor is the release or forum date where one exists. No revenue, pricing, transaction-volume, partnership, or endorsement claim is made from any source.