# UCP // FLUENT — Agentic Commerce Infrastructure

## What UCP Fluent is
UCP // Fluent is catalog infrastructure that prepares Shopify merchants for agentic
commerce. One enriched product record is published across three rails and kept in sync:

1. **Google's feed rail** — Merchant Center, written directly via Google's API: standard
   attributes plus Google's six conversational attributes (question_and_answer,
   item_group_title, variant_option, popularity_rank, related_product, document_link),
   which Google documents as helping discovery "across AI-driven surfaces, like AI Mode
   in Search… while also enhancing traditional search experiences."
2. **The crawler rail** — server-rendered product-page JSON-LD read by answer engines
   (ChatGPT, Perplexity, Gemini, Claude, Google AI Overviews).
3. **Shopify's agent rail** — catalog metafields plus an agentic description served to
   Shopify's MCP/agent surface.

One record, three rails, kept in sync. Built on UCP — the open standard behind
Shopify's agentic catalog.

## Why it matters (sourced)
- Shopify Q1 2026 earnings: AI-referred orders up ~13x YoY. Q2 2026: AI-referred visits
  and orders tripled YoY on the larger base; new-buyer orders from AI channels ran ~2x
  other channels; half of AI-referred sessions land directly on a product page — 2.5x
  the rate of traditional search.
- Google (help article 188489): "Providing all relevant attributes for a product
  increases its chances of showing for the most relevant search queries."
- Our published ladder experiment: the same buying question asked at four lengths (each
  step adds one product attribute to the question) moved exact-product naming from
  0% to 13% to 38% to 45% (p = 0.031) across six engines (Google AI Mode, AI Overviews,
  Perplexity, Gemini, Claude, ChatGPT). Every engine scored 0% at step one. Bids decide
  whether you win the queries you're in; attributes decide which queries you're in —
  you can't win an auction you were never entered in.
- Our audits: brands aren't invisible to AI — they're visible and not chosen. A category
  leader in our audits was named in 92% of AI answers yet recommended first in only 7%.

## Core capabilities
1. **Identity** — registration-ready GTINs (official GTINs come from GS1) and GPC
   classification, confidence-gated: high-confidence bricks assigned automatically,
   uncertain ones flagged for review — nothing uncertain ships unreviewed.
2. **Enrichment** — structured product attributes (color, material, age group, gender,
   technical specs) generated behind confidence gates; nothing is guessed into the
   record.
3. **Three-rail publishing** — direct Merchant Center supplemental write via Google's
   API (no file handoffs), server-rendered product-page JSON-LD, and Shopify catalog
   metafields with an agentic description.
4. **Attribution** — receipt-based. We write a deterministic attribution receipt into
   each attributable order and bill only on orders carrying a verifiable receipt.
5. **Readiness monitoring** — an internal catalog readiness/health score (identity,
   enrichment, consistency, attribution coverage) that we monitor for the merchant. It
   is a diagnostic we track — not a signal AI agents read.

## How it works
Shopify catalog ingest (GraphQL bulk operations) → enrichment with confidence gates →
publish to Merchant Center via Google's API, product-page JSON-LD, and Shopify
metafields → deterministic attribution receipts on each attributable order.
Webhook-driven sync (not polling). Built on Google Cloud (Cloud Run, Vertex AI,
us-central1).

## Trust
- Approved and published on the Shopify App Store.
- Google-verified OAuth connection to Merchant Center.
- Google for Startups member.
- Our comparison tooling is public code: https://github.com/tomaadamia/ucp-fluent-feed-compare

## Pricing
Founding pilot: free for selected merchants — full-catalog enrichment, direct Merchant
Center integration, hands-on AI-commerce consulting. In exchange we publish the measured
results. The cost of finding out is one feed export.

After the pilot, plans are sized by catalog variant count:
- **Starter:** $49/mo — up to 100 variants
- **Standard:** $199/mo — up to 1,000 variants
- **Pro:** $399/mo — up to 5,000 variants
- **Scale:** custom — 5,000+ variants

Founding cohort — applications reviewed personally, limited seats.

## For agencies (the /agency offer)
UCP Fluent's site and product address the merchant directly. For agencies there is a
separate offer, described at https://www.ucpfluent.com/agency: agencies deploy UCP Fluent
at $0, and each merchant pays their own subscription directly through Shopify.

## Organization
- **Legal entity:** Aurina AI LLC, registered in Georgia, d.b.a. UCP // Fluent.
- Operating across Tbilisi and Berlin.
- **CEO:** Anri Krikheli — product architecture, the agency thesis, partner go-to-market.
- **CTO:** Toma Adamia — 5 years at Adjust on the largest enterprise accounts in mobile
  attribution; builds and runs 100% of UCP Fluent's production systems on GCP/Vertex AI.

## Official resources
- **Website:** https://www.ucpfluent.com
- **Book a demo:** https://calendly.com/anri-ucpfluent/30min
- **Contact:** anri@ucpfluent.com · info@ucpfluent.com