# One score for how well AI can read your catalog.

Sphen.ai checks every product for the signals AI assistants and search engines rely on, then sums it up as a score out of 100 with the fixes that move it most.

Canonical: https://sphen.ai/features/readiness-score/

## What is the AI readiness score in Sphen.ai?

The AI readiness score is a single number out of 100 that shows how well search engines and AI assistants can read your Shopify catalog. Sphen.ai checks every product for missing descriptions, attributes and structured data, then shows the score, the count of products that are not AI-ready yet and your top fixes.

## What do you see in the app?

- **The score and the gap count**: When the first audit finishes, Get started reveals Your store, as AI sees it. You see the readiness score, how many products are not AI-ready yet and your top fixes. The audit is the last step of the setup pipeline and runs in the background, so you can close the page and come back when the results are in.
- **Fixes ranked by impact**: Review your first fixes opens Recommendations. The Open tab starts with Top recommendations, your highest-impact fixes as cards, each with an impact score and a Review button. Below it, the All recommendations table lists every open action. You can search by title, sort by priority and narrow by object type, so the long tail of the catalog stays workable.
- **Fields you have to fill yourself**: Some gaps cannot be drafted. The No AI fields badge marks items where every open action is a field no AI can invent, such as vendor, category or GTIN. Those need a manual edit in Shopify. Everything else gets a draft you can edit, refine, apply or dismiss. Nothing reaches your storefront until you press Apply.

## The problem

You cannot fix what nobody measured.

- **No baseline**: Without a number for today, there is no way to tell whether last month's catalog work made the store easier to read.
- **Gaps hide in the long tail**: The hero products look fine. The hundreds behind them are where descriptions, brand fields and GTINs go missing.
- **Too many findings**: A flat list of issues does not say which fix to do first.

## How it works

1. **Every product gets checked**: The last step of the first run is Checking every product for AI-readiness. It runs in the background, so you can leave the page.
2. **Read the score**: Your store, as AI sees it shows the readiness score, how many products are not AI-ready yet, and your top fixes.
3. **Work the fixes**: Review your first fixes opens Recommendations, sorted so the highest-impact changes come first.

## You stay in control

- Scoring only reads your catalog. Nothing is written to Shopify.
- Every fix behind the score is a draft you approve, edit or dismiss.
- Fields no AI can invent, such as vendor, category or GTIN, are flagged for a manual edit.

## Questions

### What goes into the score?

How complete and machine-readable each product is for search engines and AI assistants. The audit looks at descriptions, attributes, alt text, structured data and related signals. Missing brand or GTIN fields and structured data that cannot be read are typical reasons a product counts as not AI-ready yet.

### Does a higher score guarantee AI mentions?

No. The score measures whether your catalog can be understood, not whether an assistant recommends you. Whether AI engines cite your store is measured separately by AI visibility scans, which ask real buyer questions and record whether your store appears in the answer.

### How do I raise it?

Apply recommendations. Each applied fix closes a gap that the next audit picks up. Start with Top recommendations, because they are sorted by impact. For fields marked for manual editing, such as vendor or GTIN, update the value in Shopify yourself.

### Does scoring change anything in my store?

No. Scoring only reads your catalog. Every fix behind the score is a draft that you approve, edit or dismiss, and it is written to Shopify only when you press Apply. Applied changes can be reverted later from the History tab in Recommendations.

### How long does the first score take?

First audit results usually appear during setup, within minutes. Large catalogs take longer because the audit scales with catalog size. It runs in the background, so you can leave the page. Your progress, profile and questions are saved while it works.

### How many products can I audit each month?

That depends on your plan. Each plan includes a monthly allowance of product audits, and the allowance resets at the start of the next month. You can check how much you have used under Settings → Plans. Existing data and applied changes are not affected by a limit.