Skip to main content
Agentic commerce covers two distinct protocols — Universal Commerce Protocol (UCP) for AI-driven product discovery and checkout, and Agentic Commerce Protocol (ACP) for in-conversation purchases inside AI assistants like ChatGPT. The playbooks below map each implementation scenario to the right starting point, tools, and time estimate so you can skip the guesswork and ship faster.

​
Choose a playbook

UCP via Google Merchant Center

Feed requirements, Business Profile setup, and endpoint development for Google AI Mode.

UCP on Shopify

Enable UCP using Shopify’s native support, Checkout Kit, or a custom manifest.

ACP Instant Checkout

Implement the required ACP endpoints and webhooks for ChatGPT checkout.

Feed specs and formatting

Prepare and submit product feeds for UCP/ACP submission.

​
Playbook 1: UCP via Google Merchant Center

Who this is for: Brands and merchants that already have — or can create — a Google Merchant Center account. What you’ll build:
  • A clean product feed meeting Google AI Mode requirements
  • A /.well-known/ucp manifest
  • Catalog and checkout endpoints
Time estimate: 1–3 days for a developer; 2–4 hours with Shopify native UCP. Start here: Google Merchant Center Guide

​
Playbook 2: UCP on Shopify

Who this is for: Shopify merchants on any plan who want to enable UCP without custom engineering. What you’ll build:
  • A UCP manifest via Shopify Admin or a custom app
  • Checkout Kit-backed checkout endpoints, or native UCP settings
Time estimate: 1–4 hours for native UCP; 1–2 days for a custom implementation. Start here: Shopify UCP Guide

​
Playbook 3: ACP Instant Checkout

Who this is for: Brands that want ChatGPT Instant Checkout — purchases completed inside the ChatGPT conversation. What you’ll build:
  • Three ACP tool definitions
  • /acp/checkout/create, /update, and /complete endpoints
  • Webhook handlers for order fulfillment
  • An idempotency layer
Time estimate: 2–5 days for a full ACP implementation. Start here: ACP Getting Started

​
Playbook 4: UCP + ACP together

Who this is for: Brands that want maximum coverage — Google AI Mode plus ChatGPT. Implement in this order to avoid rework:
1

Start with UCP

UCP via Google Merchant Center is the fastest path to meaningful traffic because Google AI Mode has the largest reach. It also builds the catalog and checkout infrastructure that ACP will reuse.
2

Add the Asva attribution snippet

Once UCP traffic starts flowing from Google, install the attribution snippet so you can measure it in GA4.
3

Extend to ACP

Extend your checkout endpoints to support the ACP state machine. Since your UCP catalog and checkout are already live, this mostly means adding the tool definitions and idempotency layer.
Total time estimate: 3–7 days.

​
Playbook 5: Agency multi-client rollout

Who this is for: Agencies implementing agentic commerce across multiple client domains simultaneously.

​
Prioritization framework

Run Readiness Scores for all clients first. Clients with higher baseline scores cost less to push to launch — prioritize them to get early wins, then backfill the lower-scoring clients. 
# Batch readiness audit for multiple domains
for domain in client1.com client2.com client3.com; do
  curl -X POST https://asva-ai.com/api/audit \
    -H "Authorization: Bearer $ASVA_API_KEY" \
    -H "Content-Type: application/json" \
    -d "{\"domain\": \"$domain\"}" \
    | jq "{domain: .domain, score: .score, critical_gaps: [.gaps[] | select(.severity == \"critical\")]}"
done

​
Multi-domain tooling

  • Template the manifest — generate a base manifest using the Manifest API and customize per domain. Store manifest templates in version control so any domain can be re-deployed in seconds.
  • Standardize the stack — pick one PSP and one implementation pattern, then document it as your internal playbook. Consistency across clients dramatically reduces debugging time.
  • Measure with the Attribution API — use GET /api/report with each client domain to pull multi-client dashboards without logging into separate accounts.

​
Client onboarding checklist

Use this checklist for each new client before implementation begins:
  • Google Merchant Center account confirmed or created
  • Readiness Score run and gaps documented
  • Product feed format identified (JSON, XML, or Shopify native)
  • Checkout platform confirmed (Shopify, headless, custom)
  • PSP confirmed and Razorpay eligibility assessed (India clients)
  • GA4 property access granted for attribution setup

​
Deploy manifest templates at scale

After generating manifests for each client, deploy them in a single scripted pass: 
# Deploy manifests across client domains
for domain in client1.com client2.com client3.com; do
  curl -s -X POST https://asva-ai.com/api/manifest \
    -H "Authorization: Bearer $ASVA_API_KEY" \
    -H "Content-Type: application/json" \
    -d "{
      \"domain\": \"$domain\",
      \"catalog_endpoint\": \"https://$domain/api/ucp/products\",
      \"checkout_endpoint\": \"https://$domain/api/ucp/checkout\"
    }" \
    | jq 'del(.generated_at)' \
    > "manifests/$domain.json"
  echo "Generated manifest for $domain"
done

​
Reporting templates

After go-live, pull weekly attribution summaries across all clients: 
# Pull last 7 days of attribution data for a client
curl "https://asva-ai.com/api/report?domain=client1.com&start=$(date -d '7 days ago' +%Y-%m-%d)&end=$(date +%Y-%m-%d)&granularity=day" \
  -H "Authorization: Bearer $ASVA_API_KEY"
Fire and forget attribution logging — never await the Asva API call inside your checkout handler. Attribution failures must never block a customer purchase.

​
After completing any playbook

Once your implementation is live:
  1. Run the Readiness Score for a final validation: asva-ai.com/tools/readiness
  2. Install the attribution snippet to start measuring AI commerce traffic
  3. Set up GA4 custom dimensions to track AI sessions in your existing reports
See: GA4 Integration