Free audit
€0
A full audit of the shop and its top twenty products, with no catalog writes.
UCP Optimizer
Answer Engine Optimization
Win inside the AI agents your customers now shop with.
ChatGPT, Copilot and Perplexity don't read keywords — they read structured attributes. UCP Optimizer scores every product, tracks where you rank inside the agents, and writes the fixes back to your catalog.
Live on the Shopify App Store
The problem
Discovery is moving inside AI shopping agents: ChatGPT, Copilot, Perplexity, Shop, Meta AI, Gemini. These agents don't rank on keyword density — they match on structured attributes, answer-ready copy and correct taxonomy. Legacy SEO was optimised for Google's crawler; agents are a different reader, and for them completeness and structure beat keyword density every time.
How it works
- 01
Agentic Readiness Score
A 0 to 100 score for every product, computed without any language model, across five weighted dimensions: attributes, copy, media, normalization and category.
- 02
Visibility Monitor
Queries Shopify's real Catalog MCP with shopper questions and records whether your store shows up, at what agentic rank, and which competitors cluster on the same universal product.
- 03
Enrichment Engine
An AI job proposes missing attributes and answer-ready copy, but only where it can be inferred from real product data — the rest is flagged for the merchant.
- 04
Reversible writes
Every catalog write passes a confidence gate, stores the previous value, and rolls back in one click.
What it does
- Agentic Readiness Score for every product, across five data dimensions
- Grounded attribute and answer-ready copy suggestions, applied in one click
- Visibility Monitor: track your rank against competitors inside AI agents
- Reversible catalog writes, confidence-gated, with nothing invented
- A free agentic audit of your shop and top products, with no writes required
In numbers
- 0–100
- the agentic readiness score of each product, deterministic
- 5
- weighted data dimensions: attributes, copy, media, normalization, category
- 6
- AI agents monitored: ChatGPT, Perplexity, Copilot, Shop, Meta AI, Gemini
- 0.7
- the confidence gate below which nothing is written to the catalog
The product
The engineering call
The app only writes where agents actually read. The code deliberately refuses to write to private app metafields, because those are merchant-readable only and therefore invisible to agents: it writes exclusively to native fields, standard taxonomy category attributes and standard metafields.
The scoring is hybrid by design. The score is fully deterministic, costs nothing in model calls and is covered by tests: the language model is used for the fix, never for the verdict. And the prompt explicitly forbids inventing specs or generating regulated claims — medical, safety or nutritional.
Compliance lives in the schema: Catalog MCP results and images are never cached, and the monitor persists only derived metrics — rank, presence, seller domain. Zero customer personal data, so no Protected Customer Data request at all.
- Remix
- NestJS
- Polaris + App Bridge v4
- Shopify Catalog MCP
- OpenAI
- BullMQ + Redis
- Prisma + PostgreSQL
- Tailwind v4
Pricing
Starter
€29/mo
Five hundred products, two thousand enrichments a month, apply and rollback.
Growth
€99/mo
Five thousand products, daily monitoring on ChatGPT and Perplexity, competitor gap.
Pro
€299/mo
Unlimited catalog, every channel monitored, sponsored products and multi-store.
Win inside the AI agents your customers now shop with.
A Barikreativa product: design, architecture and full-stack development in house. Published on the Shopify App Store.