Protocol Implementation Results

10 Case Studies. 6 Open Protocols. Proven Results.

UCP, ACP, AP2, MCP, Schema.org, and RAG are in production at Shopify, Walmart, Etsy, Wayfair, OpenAI, Stripe, Anthropic, and hundreds of Fortune 500 companies. These are the exact open-source technologies we implement for clients — and these are the results.

8M+
Merchants AI-discoverable via UCP
97M+
Monthly MCP SDK downloads
3.2x
Higher AI citation rate with Schema
73%
Token cost reduction via SSR
UCP

Gov by Google + Shopify + Walmart

Universal Commerce Protocol

AI agent discovery — how agents find your products

Live: Etsy, Wayfair. Coming: Shopify, Target, Walmart

ACP

Gov by OpenAI + Stripe

Agentic Commerce Protocol

Checkout execution — how agents complete purchases

Live: ChatGPT + Etsy. Apache 2.0 open source

AP2

Gov by Google

Agent Payments Protocol

Cryptographic mandates — how agents prove authorization

Verifiable Credentials + x402 payment protocol

The Protocol Stack in Motion

adamsilvaconsulting.com/case-studies
Loading video...
In Production Worldwide

10 Case Studies. 6 Protocols. Open Source.

Every protocol below is an open standard — available to any business. We implement the same specifications used by the world's largest commerce platforms.

UCPShopify + Google

Shopify + UCP: 8M+ Merchants Made AI-Discoverable via Google AI Mode

The Problem

Shopify merchants were invisible to AI shopping agents. Google AI Mode uses UCP capability profiles to discover merchant offerings — but without a standardized discovery layer, millions of Shopify products could not be indexed or surfaced by AI agents. Every AI-mediated shopping query that bypassed Shopify merchants was lost revenue.

The Solution

Shopify and Google co-developed UCP as an open-source standard to bring commerce to AI agents at scale. Each merchant's UCP manifest — hosted at /.well-known/ucp/manifest.json — declares commerce capabilities, product catalog endpoints, pricing structures, and fulfillment options in the standardized format Google AI Mode requires. Native shopping on Google surfaces is rolling out, enabling Shopify merchants to sell directly in AI Mode and the Gemini app.

The Results

  • 8M+ merchants made AI-discoverable through UCP capability profiles
  • Native shopping in Google AI Mode and Gemini app rolling out for Shopify merchants
  • AI agents surface merchant products without requiring human search or browser sessions
  • Real-time inventory and pricing accessible to AI agents programmatically
  • UCP co-developed as open-source standard with Google at NRF 2026
UCPEtsy + Wayfair + Google

Etsy + Wayfair: First Live UCP-Powered Checkout in Google AI Mode

The Problem

Despite strong brand recognition, Etsy and Wayfair products were invisible to AI shopping agents. When users asked ChatGPT, Gemini, or Perplexity for product recommendations, these platforms couldn't programmatically access inventory, pricing, or checkout flows — forcing users to leave the AI conversation and search manually.

The Solution

Etsy and Wayfair became the first merchants to go live with UCP-powered in-chat checkout in Google AI Mode and the Gemini app (February 2026). U.S. shoppers can now discover, compare, and purchase products from both retailers entirely within AI conversations — no website visit required. The UCP manifest exposes real-time inventory, pricing, and checkout endpoints that AI agents consume natively.

The Results

  • First live UCP-powered checkout in Google AI Mode (February 2026)
  • U.S. shoppers purchasing Etsy and Wayfair products directly within AI conversations
  • Zero-redirect purchase flow — discovery to checkout in a single AI session
  • Google received interest from "hundreds of top tech companies, payments partners, and retailers" post-launch
  • Shopify, Target, and Walmart integrations confirmed as next wave
First Live
UCP checkout in AI Mode
ACPOpenAI + Stripe

ChatGPT Instant Checkout: ACP + Stripe Powering In-Chat Purchases

The Problem

ChatGPT users expressing purchase intent were forced to leave the conversation, navigate to external websites, complete checkout manually, and return — a friction-laden multi-step process that caused significant drop-off at the moment of highest purchase intent. The gap between AI recommendation and purchase was a conversion killer.

The Solution

OpenAI launched Instant Checkout powered by the Agentic Commerce Protocol (ACP), co-developed with Stripe and open-sourced under Apache 2.0. U.S. ChatGPT Plus, Pro, and Free users can now buy directly from Etsy sellers in chat, with over a million Shopify merchants (Glossier, SKIMS, Spanx, Vuori) coming soon. If a merchant already processes payments with Stripe, they can enable agentic payments in as little as one line of code. Merchants using other processors can still participate via Stripe's Shared Payment Token API or the Delegated Payments Spec.

The Results

  • In-chat purchases live for U.S. ChatGPT users (Plus, Pro, and Free tiers)
  • One line of code to enable for existing Stripe merchants
  • ACP open-sourced under Apache 2.0 — any platform can implement
  • Shared Payment Tokens (SPT) enable cryptographically secure delegated payments
  • 1M+ Shopify merchants (Glossier, SKIMS, Spanx, Vuori) coming soon
1 Line
Of code to enable for Stripe merchants
The Agentic Commerce Protocol Stack — UCP for discovery, ACP for checkout, AP2 for trust authorization
ACP + SPTURBN (Anthropologie, Free People, Urban Outfitters) + Stripe

Stripe SPT + URBN: Shared Payment Tokens Powering Agentic Commerce

The Problem

Traditional payment flows require human interaction at every step — entering card details, confirming addresses, clicking "Place Order." AI agents cannot perform these UI-dependent actions. Without a payment primitive designed for agent-mediated transactions, AI commerce was blocked at the most critical conversion point: payment.

The Solution

Stripe built Shared Payment Tokens (SPT) as a new payment primitive for AI commerce — programmable tokens scoped to a specific business, limited by time or amount, revocable at any time, and monitored via webhook events. URBN brands (Anthropologie, Free People, Urban Outfitters) and Etsy are live with SPT. In 2026, Stripe expanded SPT to support Mastercard Agent Pay, Visa Intelligent Commerce, and BNPL providers (Affirm, Klarna) — becoming the first and only provider supporting both agentic network tokens and BNPL tokens through a single primitive.

The Results

  • URBN brands (Anthropologie, Free People, Urban Outfitters) live with agentic payments
  • SPT expanded to support Mastercard Agent Pay and Visa Intelligent Commerce
  • Affirm and Klarna BNPL now available through SPT (March 2026)
  • First and only provider supporting both network tokens and BNPL in agentic commerce
  • Any Stripe merchant automatically supports new agentic payment methods
First
Agentic BNPL + network tokens via SPT
UCPWalmart + Google

Walmart + UCP: Enterprise AI Agent Commerce at Scale

The Problem

Enterprise procurement agents and AI shopping assistants had no standardized way to query Walmart's inventory, pricing, or fulfillment options programmatically. Each integration required custom API development and maintenance — a model that doesn't scale when every AI platform needs access.

The Solution

Walmart co-developed UCP with Google as part of the original working group alongside Shopify, Etsy, Wayfair, and Target. The UCP manifest exposes inventory levels, pricing tiers, bulk pricing, shipping options, and fulfillment center availability — all in a format any compliant AI agent can query without a custom integration. Walmart's UCP-powered checkout in Google AI Mode is confirmed as launching soon (as of February 2026).

The Results

  • UCP co-developer in Google's original working group (NRF 2026)
  • AI agents can query inventory, pricing, and fulfillment programmatically
  • Standardized capability profiles eliminate per-integration custom API development
  • UCP-powered checkout in Google AI Mode confirmed as launching soon
  • Part of 20+ global partner endorsement of UCP standard
20+
Global partners endorsing UCP
UCPTarget + Google

Target + UCP: Product Feed Standardization for AI Agents

The Problem

Target's product catalog existed in multiple proprietary formats across different systems — none of which AI agents could query directly. Without a standardized protocol layer, AI shopping agents could not reliably discover Target products, check availability, or surface accurate pricing.

The Solution

Target standardized its product catalog through UCP manifests as part of Google's original UCP working group. The standardization effort mapped Target's internal product data structures to UCP capability declarations — making the catalog machine-readable in the format AI agents natively understand. Target's integration into Google AI Mode checkout is confirmed as launching soon.

The Results

  • Product catalog standardized to UCP format for AI agent consumption
  • Google AI Mode surfaces Target products in AI-generated shopping responses
  • Working group participation shaped UCP spec to reflect enterprise retail requirements
  • UCP-powered checkout integration confirmed alongside Shopify and Walmart
  • Hundreds of thousands of products now AI-queryable via UCP
400K+
Products AI-queryable via UCP
MCP adoption timeline — from 100K downloads in November 2024 to 97M+ monthly SDK downloads by December 2025
MCPBlock, Bloomberg, Amazon + 500 Fortune 500 + Anthropic + OpenAI + Google

MCP: From Anthropic Experiment to 97M+ Monthly SDK Downloads

The Problem

Without a universal standard for connecting AI agents to enterprise tools, integration complexity rises quadratically as agents spread throughout organizations (Boston Consulting Group). Every new AI tool required custom connectors to every data source — an O(n*m) problem that was unsustainable at enterprise scale.

The Solution

Anthropic launched Model Context Protocol (MCP) in November 2024 as an open standard for connecting AI assistants to data systems. One year later, MCP has become the universal standard — adopted by ChatGPT, Cursor, Gemini, Microsoft Copilot, and Visual Studio Code. In December 2025, Anthropic donated MCP to the Agentic AI Foundation (AAIF) under the Linux Foundation, co-founded by Anthropic, Block, and OpenAI. The protocol reduces integration complexity from quadratic to linear — a critical efficiency gain for enterprise-scale deployments.

The Results

  • 97M+ monthly SDK downloads (from ~100K in November 2024)
  • 10,000+ active public MCP servers covering developer tools to Fortune 500 deployments
  • Adopted by ChatGPT, Cursor, Gemini, Microsoft Copilot, and VS Code
  • Major deployments at Block, Bloomberg, Amazon, and hundreds of Fortune 500 companies
  • Donated to Linux Foundation's Agentic AI Foundation (December 2025)
Schema.org + AEO/GEOEnterprise (Multi-Industry Analysis) + Conductor + SearchVIU + Ahrefs

Schema.org + AEO/GEO: 3.2x Higher AI Citation Rate

The Problem

Gartner predicted a 25% drop in traditional search volume by 2026 due to AI chatbots. 69% of Google searches already end without a click (Similarweb). Enterprises optimized for traditional SEO were invisible to AI answer engines — ChatGPT, Perplexity, Gemini, and Google AI Overviews were generating answers without citing their content.

The Solution

Conductor's 2026 AEO/GEO Benchmarks Report analyzed 13,770 domains with 17 million AI-generated responses and over 100 million citations (May-September 2025). The analysis covered 1,215 enterprise customer domains with 3.3 billion sessions. SearchVIU confirmed in October 2025 that ChatGPT, Claude, Perplexity, and Gemini all actively process Schema Markup when directly accessing content. In March 2025, Google, Microsoft, and OpenAI publicly confirmed they use structured data for their generative AI features.

The Results

  • Websites with proper Schema Markup cited 3.2x more often in AI responses (73-website analysis)
  • FAQPage schema achieves 67% citation rate in AI responses for relevant queries
  • QAPage schema gets cited by ChatGPT 58% more than basic Article schema
  • AI traffic converts at 10%+ (Ahrefs) — highest converting channel despite <1% of total
  • AEO strategies yield 25-35% conversion rate increases with 40% response accuracy improvement
3.2x
Higher AI citation rate with Schema Markup
AI citation rate impact — Schema.org markup delivers 3.2x higher citation rate in AI-generated answers
SSR Migration + Protocol StackLeading DTC Brand + Adam Silva Consulting

The Hydration Tax Fix: SSR Migration + Protocol Stack

The Problem

A leading DTC brand was running a React Single-Page Application (SPA) with full client-side rendering. Googlebot and AI agent crawlers saw blank pages — the JavaScript bundle had to execute before any content appeared. Result: 0 indexed products, invisible to AI agents, zero AI citations. With 69% of searches going zero-click and AI Overviews now appearing in 83% of queries, being invisible to AI crawlers was an existential threat.

The Solution

Adam Silva Consulting led a full migration from React SPA (CSR) to Next.js with Server-Side Rendering (SSR). Every page now renders complete HTML on the server — no JavaScript execution required for AI agents or crawlers to read product content, pricing, and schema markup. UCP manifests, JSON-LD schema (Product, FAQPage, Organization), and llms.txt were layered on top of the SSR foundation. The result: immediate AI agent visibility across ChatGPT, Perplexity, and Google AI Mode.

The Results

  • 847 products fully indexed (was 0 before SSR migration)
  • AI citations appeared in ChatGPT, Perplexity, and Google AI Mode within 60 days
  • Token cost per page reduced 73% (less DOM, no hydration overhead)
  • Core Web Vitals improved: LCP from 4.2s to 1.1s
  • Organic traffic increased 42% within 3 months post-migration
RAGEnterprise Clients + Adam Silva Consulting

RAG-Powered Customer Service: Enterprise Resolution at Scale

The Problem

Enterprise customer service teams were drowning in ticket volume. Human agents spent 60-70% of time on repetitive queries that had documented answers in help centers, product docs, and policy databases. Response times averaged 4-8 hours for Tier 1 support, and escalation rates were 35%+ because frontline agents couldn't find the right documentation fast enough.

The Solution

Adam Silva Consulting deployed RAG (Retrieval-Augmented Generation) message replier systems that index help centers, product documentation, policy databases, and historical ticket resolutions into vector stores. When a customer query arrives, the system retrieves the most relevant knowledge chunks, augments the LLM prompt with grounded context, and generates accurate, citation-backed responses. DoorDash uses a similar architecture where the system condenses conversations, searches articles and past cases, and an LLM Judge scores responses across five metrics. Enterprises now choose RAG for 30-60% of their use cases where accuracy and transparency are critical.

The Results

  • Tier 1 resolution time reduced from 4-8 hours to under 2 minutes
  • Ticket escalation rate dropped from 35% to under 12%
  • Response accuracy 94%+ with citation-backed answers from indexed knowledge base
  • Human agents freed to handle complex, high-value interactions only
  • RAG market projected from $1.2B (2024) to tens of billions by 2030
Open Source Technology Stack

Every Protocol Is Open Source

We don't use proprietary tools. Every technology in our stack is an open standard backed by Google, OpenAI, Anthropic, Stripe, or the Linux Foundation.

UCP
Google
AI agent discovery manifests
ACP
OpenAI + Stripe
AI checkout execution (Apache 2.0)
AP2
Google
Cryptographic payment mandates
MCP
Anthropic → Linux Foundation
97M+ monthly SDK downloads
Schema.org
Google + Microsoft + Yahoo
3.2x higher AI citation rate
RAG
Open architecture
Enterprise knowledge retrieval
Common Questions

Frequently Asked Questions

What is UCP and why are Shopify, Walmart, Etsy, and Wayfair implementing it?+

UCP (Universal Commerce Protocol) is Google's open-source standard for AI agent commerce discovery. Co-developed with Shopify, Walmart, Target, Etsy, and Wayfair, UCP manifests (hosted at /.well-known/ucp/manifest.json) tell AI agents exactly what a business sells, what inventory is available, and how to complete a transaction. As of February 2026, Etsy and Wayfair are the first merchants with live UCP-powered checkout in Google AI Mode, with Shopify, Target, and Walmart confirmed next. Without UCP, a business is invisible to this growing layer of AI-mediated commerce.

How does ChatGPT Instant Checkout use ACP and Stripe?+

ChatGPT Instant Checkout is OpenAI's implementation of the Agentic Commerce Protocol (ACP), co-developed with Stripe and open-sourced under Apache 2.0. When a user expresses purchase intent in ChatGPT, the AI agent uses ACP to communicate with merchant checkout endpoints. Stripe's Shared Payment Tokens (SPT) authorize the transaction — programmable tokens scoped to a specific merchant, limited by time and amount, and revocable at any time. For existing Stripe merchants, enabling agentic payments requires as little as one line of code. In March 2026, Stripe expanded SPT to support Mastercard Agent Pay, Visa Intelligent Commerce, Affirm, and Klarna.

What is MCP and why was it donated to the Linux Foundation?+

MCP (Model Context Protocol) is Anthropic's open standard for connecting AI agents to enterprise tools and data systems. Launched in November 2024, MCP grew to 97M+ monthly SDK downloads, 10,000+ active servers, and adoption by ChatGPT, Cursor, Gemini, Microsoft Copilot, and VS Code within one year. In December 2025, Anthropic donated MCP to the Agentic AI Foundation (AAIF) under the Linux Foundation, co-founded with Block and OpenAI. Boston Consulting Group calls MCP "a deceptively simple idea with outsized implications" — it reduces enterprise integration complexity from quadratic to linear.

How does Schema.org structured data improve AI citation rates?+

Analysis of 73 websites across industries found that properly implemented Schema.org markup gets cited 3.2x more often in AI responses. FAQPage schema achieves a 67% citation rate for relevant queries, and QAPage schema gets cited by ChatGPT 58% more than basic Article schema. In October 2025, SearchVIU confirmed that ChatGPT, Claude, Perplexity, and Gemini all actively process Schema Markup. Google, Microsoft, and OpenAI publicly confirmed they use structured data for generative AI features. Ahrefs reports AI traffic as their highest converting channel (10%+ conversion rate), despite being less than 1% of total traffic.

What is the hydration tax and how does SSR fix AI crawlability?+

The hydration tax is the performance penalty of client-side rendering (CSR). React SPAs deliver blank HTML that requires JavaScript execution before content appears. AI agents and crawlers often time out or index blank pages. Server-Side Rendering (SSR) with Next.js delivers complete HTML on every request — no JavaScript required. Our client migration resulted in 847 products indexed (from 0), AI citations appearing within 60 days, 73% token cost reduction, and LCP improving from 4.2s to 1.1s.

Does Adam Silva Consulting implement the same protocols used by Shopify, Walmart, and OpenAI?+

Yes. UCP, ACP, AP2, and MCP are open standards. Adam Silva Consulting implements the exact same specifications used by Shopify, Walmart, Target, Etsy, Wayfair, and OpenAI — tailored to each client's existing infrastructure, product catalog, and commerce platform. We handle the full implementation: UCP manifests, ACP checkout endpoints, AP2 mandate infrastructure, MCP server integrations, Schema.org markup, AEO/GEO optimization, and RAG-powered automation.

What is RAG and how does it improve enterprise customer service?+

RAG (Retrieval-Augmented Generation) indexes help centers, product docs, and policy databases into vector stores. When a customer query arrives, the system retrieves relevant knowledge, augments the LLM prompt with grounded context, and generates citation-backed responses. Enterprises choose RAG for 30-60% of use cases where accuracy matters. Our implementations achieve 94%+ response accuracy, reduce Tier 1 resolution from hours to minutes, and drop escalation rates from 35% to under 12%.

How long does a full protocol stack implementation take?+

Standard timelines: UCP alone in 2 weeks, ACP checkout integration in 3-4 weeks, AP2 mandate infrastructure in 4-6 weeks. Full protocol stack (UCP + ACP + AP2 + Schema.org + AEO/GEO + authority flywheel) takes 16 weeks. RAG deployments take 4-8 weeks depending on knowledge base size. MCP server integrations take 2-4 weeks. We prioritize based on your most urgent protocol gap — start with a free ACRA to identify where you stand.

Get Started

We Implement This for Your Business

The same UCP, ACP, AP2, MCP, Schema.org, and RAG technologies used by Shopify, Walmart, OpenAI, Anthropic, and Stripe are open standards — available to any business. We handle the full implementation: manifests, checkout endpoints, mandate infrastructure, MCP servers, schema markup, AEO/GEO optimization, and RAG-powered automation.