AI Readiness Score: What It Measures

> The FeedBridge AI Readiness Score is a 0–100 composite score that measures how ready a product is for discovery, evaluation, and purchase by AI shopping agents — calculated from 50 scoring vectors organised across four dimensions: Protocol Compliance (30%), Content Quality (25%), AI Enrichment (30%), and Commerce Signals (15%).

---

What Is the AI Readiness Score?

The AI Readiness Score is FeedBridge's primary measurement of a product's fitness for the AI commerce channel. It is a numeric score from 0 to 100 that reflects how completely and correctly a product's data layer addresses the requirements of AI shopping assistants — not just whether the product exists in a feed, but whether it has the structured fields, enrichment data, protocol signals, and commerce attributes that AI agents need to discover, evaluate, and transact on it confidently.

The score is calculated from 50 individual scoring vectors — discrete data checks across the product record — aggregated into four weighted dimensions. Each dimension captures a different layer of AI readiness, and the combined score reflects the product's overall fitness across all four layers simultaneously. The score is calculated at the per-product level, meaning every product in a merchant's catalog receives its own score, and an overall brand score is derived from the aggregate of the individual product scores.

The AI Readiness Score is not a subjective assessment or a benchmark comparison — it is a structured evaluation of whether specific, documented field requirements are met at the product level. Each vector has a binary or graded outcome; the score is the weighted aggregate of those outcomes.

---

The Four Dimensions

Protocol Compliance — 30%

Protocol compliance is the highest-weighted dimension. It evaluates whether a product's data meets the field requirements of ACP and UCP — the two primary AI commerce protocols. Protocol compliance vectors include:

Presence of ACP trust signal fields: `accepts_returns` (boolean), `return_deadline_days` (integer), `is_digital` (boolean)
Presence and structure of `variant_dict` for products with variants
ACP feed format correctness (JSON-LD)
UCP `native_commerce` attribute presence
GTIN and/or MPN presence (required for UCP Catalog Lookup)
Taxonomy path structure and normalisation

Protocol compliance is weighted at 30% because its gaps are the most consequential — a product that fails protocol compliance cannot complete an AI-assisted checkout regardless of how good its content is. The ACP trust signals in particular are a hard requirement for Instant Checkout eligibility; without them, the checkout agent cannot answer the buyer's basic pre-purchase questions about returns.

Content Quality — 25%

Content quality evaluates the completeness and clarity of the core product content fields. Vectors in this dimension include:

Title: length, specificity, absence of keyword stuffing
Description: word count, factual richness, absence of templated or placeholder text
Images: presence of at least one image URL; URL accessibility
Category: category field populated, category path specificity
Attributes: number of product attributes populated relative to the vertical's expected attribute set

Content quality is weighted at 25% because content gaps are the second most common cause of poor AI discovery performance. An agent that cannot read a clear, specific title and a detailed description has limited material for matching the product to a buyer's stated intent.

AI Enrichment — 30%

AI enrichment evaluates the semantic and contextual fields that make a product interpretable beyond its basic data. This is joint-highest with protocol compliance at 30%, reflecting FeedBridge's assessment that enrichment is as important as compliance for AI commerce performance. Vectors include:

Intent tags: present, minimum count, semantic relevance
`who_should_buy` persona arrays: present and populated
Use case descriptions: present and substantive
`q_and_a` JSON array: present, minimum Q&A pair count, factual accuracy of answers
Voice snippet: present and voice-optimised
Taxonomy normalisation path: present and resolved to a valid leaf node
Eligibility mappings: vertical-specific attributes (diet type, compatibility, fit type) where applicable

AI enrichment gaps are the primary reason products with otherwise complete catalogs underperform in AI-assisted recommendation. A product that has no intent tags, no persona array, and no Q&A pairs gives the agent nothing to work with beyond its title and description when a buyer asks a nuanced question.

Commerce Signals — 15%

Commerce signals evaluate the product attributes that support transactional confidence. Vectors include:

GTIN presence (Global Trade Item Number)
MPN presence (Manufacturer Part Number)
Brand field populated
Availability status declared and current (`in_stock` or `out_of_stock`)
Pricing completeness: price, currency, and valid price format
Sale pricing: `sale_price` with `sale_price_start_date` and `sale_price_end_date` where applicable
Review data: structured reviews as JSON array present

Commerce signals are weighted at 15% because their absence creates friction rather than complete failure. A product without a GTIN is still discoverable and purchasable — but the agent has lower product identity confidence. A product without structured reviews cannot surface social proof to a buyer who asks "what do other buyers say about this?"

---

Score Output Format

The AI Readiness Score is presented in two formats in FeedBridge:

Per-product score: Every product receives a numeric score (0–100) and a color-coded tier label based on where that score falls:

Excellent — top score range; product meets requirements across all four dimensions
Good — strong score; one or more dimensions have minor gaps
Needs Work — moderate score; meaningful gaps in one or more dimensions
Poor — low score; significant gaps across multiple dimensions

Overall brand score: The aggregate AI readiness score across all products in the brand's catalog, reflecting the average readiness of the full portfolio.

Alongside the score, FeedBridge generates actionable fix suggestions per product — a per-product checklist of the specific vectors where the product is scoring below threshold, with one-click navigation to the problem field in the product detail view. This closes the loop between measurement and remediation: the score identifies the gap, the fix suggestion labels it, and one-click navigation takes the merchant to the exact field that needs attention.

---

How the Score Is Used in Practice

The AI Readiness Score is used by merchants in three practical ways:

1. Catalog audit. Running the score across a full product catalog gives merchants a prioritised view of which products have the most critical readiness gaps. Products in the Poor tier with protocol compliance failures should be addressed first; they are the most likely to cause failures in live AI checkout flows.

2. Enrichment prioritisation. For merchants with large catalogs who cannot enrich all products simultaneously, the score provides a priority queue — products with the most improvement potential from AI enrichment (i.e., Good on compliance but Poor on enrichment) are the highest-leverage enrichment targets.

3. Pre-certification check. Before submitting a product feed to ChatGPT Shopping (ACP) or Google Merchant Center (UCP), merchants can use the score as a pre-certification signal. Products scoring Excellent on Protocol Compliance are less likely to fail schema validation during ACP or UCP integration approval. Products with known compliance failures are flagged for correction before submission.

---

Why It Matters for Merchants

The AI Readiness Score gives merchants something they have not had before: a per-product, structured view of why their catalog may not perform in AI commerce, expressed in terms they can act on. Without this measurement, a merchant whose products are not appearing in AI-assisted searches or not completing AI-initiated checkouts has no way to know which specific data gaps are responsible.

The score translates the abstract requirement of "AI readiness" into a concrete, actionable checklist. For a merchant preparing to onboard to ACP or UCP, the Protocol Compliance dimension of the score is a direct pre-certification readiness signal. For a merchant trying to improve product recommendation performance in AI shopping surfaces, the AI Enrichment dimension points to the exact fields that would unlock better agent evaluation.

---

FeedBridge Relevance

The AI Readiness Score is a live FeedBridge platform feature. It measures 50 vectors across four dimensions, produces color-coded tier labels (Excellent, Good, Needs Work, Poor), generates per-product actionable fix suggestions with one-click navigation, and produces an overall brand score. The Public Readiness Checker at feedbridge.ai/score is a free lead-capture tool that gives any merchant a sample AI readiness assessment before they create a FeedBridge account, making it the entry point for the scoring funnel.

The score is computed from the enriched product record in FeedBridge's catalog — meaning that applying AI enrichment through the platform directly improves the AI Enrichment dimension score, and adding trust signals or protocol fields through the ACP/UCP feed pipeline directly improves the Protocol Compliance dimension score. The score is not a static label; it updates as the product record improves.

---

Frequently Asked Questions

Q: Is a score of 100 achievable for all products? A: In principle, yes — a product that meets all 50 vector requirements across all four dimensions would score at the maximum. In practice, some vectors apply only to specific product types (e.g., `is_digital: true` is only meaningful for digital goods; variant data is only relevant for products with variants), so the effective scoring model is adjusted for the product's vertical and type.

Q: How is the overall brand score calculated? A: The overall brand score is derived from the aggregate of individual product scores across the merchant's catalog. The specific aggregation method (mean, weighted by product importance, etc.) is a FeedBridge platform detail. Merchants can improve their brand score by raising the scores of their lowest-performing products.

Q: Can I see which specific vectors a product is failing? A: Yes. The actionable fix suggestions feature in FeedBridge provides a per-product checklist of the specific vectors where the product is below threshold, with one-click navigation to the problem field in the product detail modal. This is available for every product in the catalog.

Q: Does the score change automatically when I update a product? A: The score reflects the current state of the product record. After applying enrichment through FeedBridge — for example, adding intent tags or populating trust signals — the score updates to reflect the improved data. The specific refresh cadence is a platform implementation detail.

Q: Is the AI Readiness Score the same as a protocol conformance certificate? A: No. The AI Readiness Score measures the data layer quality at the product level within FeedBridge. Protocol conformance certification (for ACP or UCP) is a separate process run by the AI platform (OpenAI for ACP, Google for UCP) that validates the merchant's endpoint behaviour, not just the product data. A high Protocol Compliance score in FeedBridge indicates strong product data readiness; it is not a substitute for the platform-level certification process.

---

Source Documentation

| Claim | Source | Source Class | Reference | |---|---|---|---| | 0–100 Score, 50 scoring vectors, 4 dimensions with weights: Protocol Compliance 30%, Content Quality 25%, AI Enrichment 30%, Commerce Signals 15% | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Color-coded tiers: Excellent, Good, Needs Work, Poor | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Actionable fix suggestions: per-product checklist with one-click navigation | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Public Readiness Checker at feedbridge.ai/score — free lead-gen tool | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | ACP protocol fields: trust signals, variant_dict, q_and_a, reviews, sale price | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md |

AI Readiness Score: What It Measures

AI Readiness Score: What It Measures

What Is the AI Readiness Score?

The Four Dimensions

Protocol Compliance — 30%

Content Quality — 25%

AI Enrichment — 30%

Commerce Signals — 15%

Score Output Format

How the Score Is Used in Practice

Why It Matters for Merchants

FeedBridge Relevance

Frequently Asked Questions

Related Topics

Source Documentation

Related Topics