Commerce Signals That Improve AI Shopping Readiness
> Commerce signals are the product data attributes that give AI shopping agents confidence in the transactional accuracy of a product — including GTIN, MPN, brand, availability status, pricing completeness, sale pricing with valid date ranges, and structured review data — and are scored as the Commerce Signals dimension, contributing 15% to FeedBridge's AI Readiness Score.
---
What Are Commerce Signals?
Commerce signals are the product-level data points that communicate transactional trustworthiness to AI shopping agents. They sit at the intersection of product identity, pricing accuracy, inventory status, and social proof — the specific attributes an AI agent checks when determining whether a product can be confidently recommended and purchased on a buyer's behalf.
The term "signal" is appropriate because these attributes are not content — they do not describe what the product is or how to use it. They are structured data points that signal to the agent: this product has a verified identity (GTIN/MPN), this product is currently available (availability status), this product's price is correct and current (pricing fields), this product has a sale with valid dates (sale pricing), and this product has social proof from real buyers (structured reviews). Each signal independently contributes to the agent's confidence that recommending and transacting on this product is appropriate.
In FeedBridge's AI Readiness Score model, Commerce Signals is the fourth dimension, contributing 15% to the overall 0–100 score. It is the lowest-weighted dimension — not because commerce signals are unimportant, but because their individual impact is more focused than the broader dimensions of protocol compliance, content quality, and AI enrichment. Commerce signal gaps create friction; they rarely create complete failures. But at the moment of purchase decision, a product without commerce signals loses credibility in ways that matter.
---
The Commerce Signals Evaluated
GTIN — Global Trade Item Number
GTIN is the barcode identifier that uniquely identifies a physical product at the item level — the number printed as a barcode on the product packaging. For AI commerce, GTIN is a critical product identity signal for two reasons.
First, it allows agents to cross-reference a product across multiple sources. If a buyer asks an AI assistant to "find me the Sony WH-1000XM5 headphone," the agent can look up that specific product by GTIN across any UCP-enabled merchant's Catalog Lookup API — without relying on the merchant's title or description to confirm the product identity. Products with GTIN are unambiguously identifiable; those without it depend on title-matching, which is inherently less precise.
Second, GTIN is required for the UCP Catalog Lookup API's barcode-based retrieval path. The specification supports lookup by `item_id` or by barcodes (GTIN/MPN). Merchants whose products lack GTIN cannot be found via barcode lookup, narrowing the range of agent workflows that can surface them.
For private-label products without assigned GTINs, GTIN absence is a known limitation rather than a remediable gap. FeedBridge's commerce signals scoring accounts for product type in this context — the absence of GTIN on a clearly private-label item is weighted differently than its absence on a name-brand product that should have an assigned barcode.
MPN — Manufacturer Part Number
MPN is the manufacturer's identifier for a specific product model. Like GTIN, it provides a cross-reference point for product identity — particularly for electronics, automotive parts, and industrial products where MPN is the primary identifier used in procurement and comparison.
MPN presence is a commerce signal for the same reason as GTIN: it gives agents a precise, machine-readable identifier that is more reliable than title-matching for confirming product identity. For product categories where MPN is standard (electronics, hardware, components), its absence is a signal quality gap. For categories where MPN is not typically used (fresh food, hand-crafted goods), its absence is expected and not penalised in the commerce signals score.
Brand
The brand field declares which brand manufactures or sells the product. It is both a content quality attribute and a commerce signal. As a commerce signal, brand serves two functions:
Branded query matching. When a buyer asks for "a Samsung TV under ₹30,000," the agent uses the brand field to filter for Samsung products. A product without a brand field will not appear in branded queries even if it is made by Samsung and the word "Samsung" appears in the title — structured brand field data is more reliable for agent filtering than title extraction.
Merchant credibility. For agents evaluating products in a category, brand data adds a layer of trustworthiness — particularly for products from well-known manufacturers. A product page that has a declared brand can be cross-referenced more confidently than one without.
Availability Status
Availability is the single most operationally critical commerce signal. It declares whether the product is currently `in_stock` or `out_of_stock`. For AI commerce, stale or missing availability data causes two distinct failure modes:
False positives: A product shown as `in_stock` in the feed but actually unavailable will cause checkout sessions to fail with `out_of_stock` errors when the agent attempts to create a session. This creates a poor buyer experience and may cause the AI platform to deprioritise the merchant's products.
False negatives: A product that is back in stock but still showing `out_of_stock` in the feed will not be recommended or purchased, representing missed revenue.
Availability scoring in FeedBridge evaluates whether the availability field is explicitly declared (not absent) and whether the feed scheduling is configured to refresh frequently enough that availability changes are reflected in a timely manner. The known gap in FeedBridge's roadmap is webhook-based real-time inventory sync, which is a high-priority item not yet live — meaning that for merchants where inventory changes minute-to-minute, frequent scheduled refresh is the current mechanism for keeping availability current.
Pricing Completeness
Pricing completeness covers three fields: `price` (numeric), `currency` (ISO 4217 code, e.g., USD, INR, GBP), and the validity of the price format. Commerce signals scoring evaluates:
- Price field present and numeric (not a formatted string like "₹2,999")
- Currency field present and using a valid ISO 4217 code
- Price value within a plausible range for the product category
Sale Pricing with Valid Date Ranges
When a product is on promotion, sale pricing must be declared with three fields:
- `sale_price` — the promotional price (numeric)
- `sale_price_start_date` — the start of the promotional period (ISO 8601 date-time)
- `sale_price_end_date` — the end of the promotional period (ISO 8601 date-time)
Structured Reviews
Structured reviews are a commerce signal for social proof. FeedBridge's AI Readiness Score evaluates whether the product has a `reviews` JSON array containing structured review objects (reviewer name, rating, review text). Aggregate star ratings alone do not satisfy this vector — structured review objects provide narrative evidence of product quality that agents can surface to buyers who ask "what do other buyers say about this?"
The signal scored here is the presence and structure of review data, not the sentiment or volume of the reviews themselves. The presence of structured reviews signals that the product has transaction history and that the merchant has committed to surfacing that history in the data layer.
---
How Commerce Signals Interact with Other Dimensions
Commerce signals do not operate in isolation — they reinforce the other three readiness dimensions:
- With protocol compliance: GTIN and MPN are required for UCP Catalog Lookup by barcode, making them protocol signals as well as commerce signals. Brand presence is evaluated in the UCP compliance context as well as the commerce signals dimension.
- With content quality: Brand is also a content quality attribute — its absence affects both the content quality sub-score and the commerce signals sub-score.
- With AI enrichment: Review data complements Q&A pairs in the AI enrichment dimension. A product with structured reviews and Q&A pairs gives agents a richer foundation for answering buyer questions than one with either alone.
---
Why It Matters for Merchants
Commerce signals are the finishing layer of a well-constructed product data record. They are not the most complex fields to add — GTIN is typically available from the manufacturer, brand is almost always known, and availability status is a direct system value. But they are frequently absent from merchant catalogs because they are not surfaced prominently in standard product management workflows.
For AI commerce specifically, these signals have an outsized impact at the moment of purchase decision. A buyer who has been recommended a product by an AI agent and is ready to purchase may ask "is this in stock?" or "does this have a sale price today?" or "what do other buyers say?" The agent's ability to answer these questions confidently depends on the commerce signals being present, accurate, and current.
The return on commerce signal investment is particularly strong because the effort is low. Adding GTIN, populating the brand field, and confirming availability status are typically single-field corrections — unlike AI enrichment, which requires generating new semantic content. For merchants prioritising quick score improvements, commerce signals are the highest-leverage, lowest-effort readiness work in the catalog.
---
FeedBridge Relevance
FeedBridge's AI Readiness Score includes Commerce Signals as its fourth dimension (15% weight), evaluating GTIN, MPN, brand presence, availability status, pricing completeness, sale pricing with valid date ranges, and structured reviews as a JSON array. The actionable fix suggestions in the score identify specific commerce signal gaps per product with one-click navigation to the relevant field in the product detail modal.
Feed health monitoring and feed scheduling in FeedBridge directly support the availability accuracy signal by keeping the feed current with the merchant's actual inventory state. ACP structured reviews — formatted as JSON arrays with reviewer name, rating, and review text — are generated and validated within FeedBridge's ACP feed pipeline. Sale pricing fields (`sale_price`, `sale_price_start_date`, `sale_price_end_date`) are supported as live features in the ACP feed output.
The Public AI Readiness Checker at feedbridge.ai/score surfaces commerce signal gaps as part of the initial readiness assessment, giving merchants visibility into their signal completeness before beginning a full catalog enrichment project.
---
Frequently Asked Questions
Q: Does my product need a GTIN if it is a private-label product I manufacture myself? A: Private-label products without assigned GTINs are a known category where GTIN absence is expected. If you manufacture your own products, you can apply for GS1-issued GTINs, which would improve both commerce signals scoring and cross-platform product identity. If your products do not have assigned GTINs, their absence is accounted for in the commerce signals scoring model and will not disproportionately penalise private-label merchants.
Q: How do I fix a stale availability status? A: The most reliable fix is implementing scheduled feed refresh in FeedBridge, so that the feed reflects inventory changes within your normal stock update cycle. FeedBridge's feed scheduling feature supports configurable auto-refresh intervals. For real-time inventory accuracy, the known roadmap gap is a webhook-based real-time inventory sync (currently high-priority, not yet live), which will allow push-based availability updates when it becomes available.
Q: Why is Commerce Signals only 15% of the score when availability accuracy seems so critical? A: Commerce signal gaps create friction and loss at the margin — a product with incorrect availability fails individual transactions, but does not block all agent interactions. Protocol compliance gaps (30%) are gate conditions that prevent entire workflows from functioning. AI enrichment gaps (30%) reduce recommendation quality across all queries. Content quality gaps (25%) limit discoverability broadly. Commerce signal fixes are often faster to implement than the higher-weighted dimensions, meaning merchants can close a 15% gap with relatively low effort and then focus investment on the higher-weighted dimensions.
Q: Is availability the most important commerce signal? A: For operational reasons, availability is the most immediately impactful — a stale availability status can cause real-time checkout failures. For discoverability, GTIN and brand are often more important because they determine whether the product can be found in branded and barcode-based agent queries. For social proof at the purchase decision stage, structured reviews matter most. The relative importance depends on where in the agent journey the merchant wants to optimise.
Q: Can I add structured reviews to my product data if my store uses a review platform like Yotpo or Judge.me? A: Yes — the review data needs to be in the correct JSON array format in the FeedBridge product record to contribute to the commerce signals score. Reviews from third-party platforms can be mapped into the `reviews` field as structured objects as part of the catalog import or enrichment workflow. The specific integration path depends on the review platform and import method used.
---
Related Topics
Parent hub: AI Commerce Readiness Scoring
Related concepts:
- AI Readiness Score: What It Measures
- Content Quality Scoring for Product Discoverability
- AI Enrichment Scoring Explained
- Missing GTIN, MPN, and Brand Data
- FeedBridge AI Readiness Score Features
Breadcrumb:
---
Source Documentation
| Claim | Source | Source Class | Reference | |---|---|---|---| | Commerce Signals 15% of AI Readiness Score; vectors: GTIN/MPN, availability, brand presence | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | ACP structured reviews as JSON array, sale price with start/end dates | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | UCP Catalog Lookup by item_id or barcodes (GTIN/MPN) | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Feed scheduling, feed health monitoring, stale feed alerts | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Real-time inventory sync: high-priority roadmap gap, not yet live | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Actionable fix suggestions, per-product checklist, one-click navigation | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md | | Public AI Readiness Checker at feedbridge.ai/score | FeedBridge Platform Capabilities April 2026 v2.0 | T1 – FeedBridge Internal | FeedBridge-Platform-Capabilities-April2026.md |