Catalog Intelligence

Catalog Intelligence helps you understand the quality of your product catalog and how Verada AI can improve it.

A product catalog is the foundation for merchandising, product recommendations, search, filters, feeds, personalization, and AI-powered shopping experiences. Catalog Intelligence analyzes your uploaded catalog, identifies missing or weak product data, shows what Verada AI has enriched, and reports whether the catalog is ready for downstream use.

Use Catalog Intelligence to:

• Diagnose source catalog data quality
• Understand where product data is missing or incomplete
• Compare uploaded catalog quality with Verada AI enriched data
• Review category, attribute, search, and product intelligence coverage
• Inspect individual products and the enriched data generated for them
• Identify future opportunities for review, text enrichment, and activation

Prerequisites

Catalog Intelligence requires a product catalog to be available in Fanplayr Merchandising.

For best results, your catalog should include:

• Product ID or SKU
• Product title
• Product URL
• Brand
• Category path
• Price and currency
• Availability
• Product images
• Product descriptions
• Variant or parent-product relationships, where applicable
• Attribute fields such as color, size, material, gender, age group, pattern, or other category-specific facts

Catalog Intelligence can still analyze sparse catalogs, but missing source data may lower the uploaded catalog score.

How Catalog Intelligence Works

Catalog Intelligence separates analysis into two views:

1. Uploaded catalog diagnosis
   Measures only the data supplied in your catalog. This provides a strict view of source data quality before enrichment.

2. Verada AI enriched catalog intelligence Measures the catalog after Verada AI has assigned standardized categories, identified product facts, generated search language, captured audience and variant signals, and prepared product intelligence for downstream systems.

This separation helps you see both:

• What was available in the source catalog
• What Verada AI added or improved

Accessing Catalog Intelligence

Open the Fanplayr Portal and go to:

Merchandising > Catalog Intelligence

If your account has multiple catalogs or projects, use the project selector at the top of the page.

If multiple languages are enabled, use the language selector to view the report in the supported language.

Catalog Intelligence Overview

The Overview tab summarizes the current catalog quality and the impact of Verada AI enrichment.

Catalog Diagnosis

The Catalog Diagnosis card shows the uploaded catalog score.

This score is based only on the uploaded catalog data. It does not include Verada AI enrichment.

The uploaded catalog score evaluates areas such as:

• Core coverage
• Media coverage
• Classification coverage
• Identifier coverage
• Category attribute coverage
• Variant coverage
• Audience and search coverage
• Syndication readiness
• Content quality

A low uploaded catalog score does not mean the catalog cannot be improved. It means the uploaded source data is missing information that may be needed for better search, discovery, feeds, recommendations, or AI experiences.

Enrichment Status

The Enrichment Status card shows the enriched score after Verada AI processing.

This score reflects the catalog after Verada AI has generated or normalized product intelligence, including:

• Standardized category assignments
• Product-specific attributes
• Attribute values
• Search terms and tags
• Audience signals
• Variant and option signals
• Embeddings, where available

The score uplift shows the difference between the uploaded catalog score and the enriched score.

Verada AI Summary

The Verada AI Summary highlights the overall improvement and the largest remaining opportunities.

Use this section to quickly understand:

• How much the catalog improved after enrichment.
• Which data areas still limit catalog readiness.
• Whether remaining gaps require better source data, more enrichment, or review.

Raw Score Breakdown

The Raw Score Breakdown explains each source catalog score component.

Each row includes:

• The data area being scored
• The score
• A short explanation of what is being checked
• Why the data matters

Common data areas include:

Data Area	What It Measures
Core coverage	Basic product identity, such as title, URL, brand, price, and currency.
Media coverage	Image presence and image depth.
Classification coverage	Category and product-type specificity.
Identifier coverage	GTIN, barcode, MPN, vendor SKU, or manufacturer part number.
Category attribute coverage	Expected product facts for the assigned product category.
Variant coverage	Parent-product and SKU grouping, option values, and variant signals.
Audience/search coverage	Gender, age group, search terms, tags, occasions, or use-case signals.
Syndication readiness	Structured data useful for search engines, marketplaces, feeds, and AI crawlers.

What Verada AI Added

This section shows the enriched data now available after Verada AI processing.

Examples include:

• Product intelligence score
• Core product facts
• Search-ready products
• Evidence-backed attributes
• Filter attributes
• Profile depth
• Product knowledge graph readiness
• Discovery language
• AI retrieval readiness
• Choice and option understanding

Use this section to explain how enrichment turns source catalog data into structured product intelligence.

Highest Impact Issue Drivers

This table ranks the most important remaining catalog gaps.

Each issue includes:

• Impact rank
• Driver
• Score
• Affected products
• Resolution path
• Reason
• Status

Use this table to prioritize cleanup or improvement work.

Uploaded Catalog Tab

The Uploaded Catalog tab shows the quality of the source catalog before Verada AI enrichment.

Use this tab when you want to understand what product data was originally supplied.

Top Metrics

The metric cards summarize source-data coverage across key catalog areas.

These scores are strict. They are based on uploaded catalog data only.

Examples:

• Core coverage
• Media coverage
• Classification coverage
• Identifiers
• Attributes
• Variant structure
• Audience/search signals
• Syndication readiness

Raw Data Coverage Breakdown

This table explains how each source-data area was evaluated.

Use it to understand why a score is high or low.

For example:

• Identifier coverage may be low if GTIN, MPN, barcode, or manufacturer part number is missing.
• Media coverage may be partial if products have images but only one image per product.
• Category attribute coverage may be low if the source catalog does not provide expected product facts for the product category.

Category and Attribute Gaps

This section shows missing facts by category.

Category profiles determine which attributes are expected for a category. For example, a clothing item may need size, material, color, fit, or care instructions, while jewelry may need material, color, design, or length type.

Missing category attributes can affect:

• Filters
• Product comparison
• Search recall
• Recommendations
• Feed quality
• AI answer quality

Content Quality Diagnostics

Content Quality Diagnostics is a future area for deeper analysis of titles, descriptions, bullets, SEO metadata, tone, and product claims.

Verada AI Tab

The Verada AI tab shows the enrichment output and current product intelligence.

Use this tab to understand what Verada AI added to the catalog.

Top Metrics

The top metrics summarize enrichment impact.

Common metrics include:

Metric	Meaning
Score uplift	Improvement between uploaded catalog score and enriched score.
Product Intelligence	Overall enriched catalog intelligence score.
Core facts	Products with required product facts completed.
Search-ready	Products ready for search language and discovery.
Filter facts	Products with useful attributes for filters and comparison.

Experience Readiness

This table translates enrichment into business capabilities.

Examples:

• Search and discovery
• Product knowledge graph
• AI answers and comparison
• Recommendations and personalization
• Feeds, crawlers, and external discovery

Each row shows the readiness score, supporting evidence, why it matters, and status.

Product Intelligence Distribution

This section groups products by readiness level.

Example bands may include:

• Excellent
• Ready
• Nearly ready
• Needs review

Use this to understand whether enriched intelligence is broadly available across the catalog or limited to a smaller set of products.

Product Intelligence Scorecard

The scorecard explains how the enriched score is calculated.

Common score components include:

Component	Meaning
Product Intelligence Score	Overall product-level readiness after enrichment.
Core attributes	Required category facts that are present.
Search-ready products	Products with enough generated search language for discovery.
Evidence-backed attributes	Attributes grounded in product evidence.
Filter attributes	Attributes useful for faceting, comparison, and feeds.
Profile depth	Optional facts that deepen product understanding.

Enrichment Quality Signals

This section shows whether enrichment is usable at scale.

Examples:

• Search readiness
• Generated search language
• Attribute coverage
• Option signals

Use this section to verify that enrichment is not just present in a few products, but available broadly across the catalog.

Enrichment Coverage by Area

This table shows how much enriched data exists by area.

Examples:

• Core attributes
• Filter attributes
• Profile depth
• Evidence quality
• Generated search language

Commerce Capabilities Enabled

This section explains where enriched data can be used.

Examples:

• Onsite search and retrieval
• Filters and comparison
• Feed and taxonomy portability
• Personalization and recommendations
• Variant and option understanding

Sample Identified Categories

This table shows standardized category paths identified by Verada AI.

These categories support:

• Reporting
• Feed mapping
• Search
• Recommendations
• Enrichment profiles

Sample Identified Attribute Families

This table shows product fact families identified across the catalog.

Examples:

• Product Type
• Gender
• Color
• Fabric
• Size
• Pattern
• Age Group
• Care Instructions
• Jewelry Material
• Sleeve Length

These attribute families can support filters, comparison, search, recommendations, and AI answers.

Readiness Tab

The Readiness tab shows what experiences the current catalog intelligence can support.

This tab evaluates the current catalog state after enrichment, where enrichment is available.

Top Metrics

Readiness metrics may include:

• Search language
• Facet readiness
• Vector embeddings
• Channel data
• Behavioral data and revenue connectors

These metrics help you understand whether the catalog is ready for search, recommendations, feeds, comparison, and AI-powered experiences.

Experience Readiness

This table lists supported experiences and their readiness.

Examples:

Experience	What It Means
Search and discovery	Products have search language, facts, and structure for discovery.
Product knowledge graph	Products have structured facts useful for comparison and AI understanding.
AI answers and comparison	Product facts can support answer and comparison experiences.
Recommendations and personalization	Product traits and audience context can improve matching.
Feeds and external discovery	Product data can be used outside the storefront.

Search Intent Coverage

This section shows how catalog intelligence can support shopper intent.

When behavioral data is connected, prioritization can shift from data completeness to impact areas such as:

• Revenue
• Zero-result searches
• High-demand product intents
• Search and browse conversion

Categories Tab

The Categories tab groups product intelligence by assigned ontology category.

Use this tab to understand catalog readiness by product family.

Top Metrics

The top metrics summarize category-level coverage.

Examples:

• Assigned categories
• Average intelligence
• Core readiness
• Search readiness
• Evidence quality
• Enriched data points

Category Intelligence Table

Each row represents a category or product family.

Common columns include:

Column	Meaning
Category	The assigned ontology category.
Products	Number of products in the category.
Intelligence Score	Average product intelligence score for the category.
Core Readiness	Products with required category facts completed.
Search Readiness	Products ready for search and discovery.
Avg. Enriched Data Points	Average enriched facts, tags, terms, categories, and other signals per product.
Evidence Quality	Whether enriched facts are supported by product evidence.
Status	Overall readiness status.

Use this view to identify categories that are strong, ready, or need more attention.

Products Tab

The Products tab shows representative product-level examples.

Use this tab to inspect how individual products are scored and summarized.

Each product row may include:

• Product title
• Ontology category
• Product intelligence score
• Core readiness
• Search readiness
• Enriched data points
• Evidence quality
• Status

This tab is useful for moving from catalog-level reporting to product-level investigation.

Lookup Tab

The Lookup tab lets you inspect a specific product or variant.

Enter a product ID or variant ID and select Lookup.

The lookup returns matching product intelligence data for the selected catalog.

Summary

The top of the result shows:

• Product title
• Product ID
• Variant ID
• Ontology category
• Enrichment status

Assigned Category

This section shows how the product was categorized.

Fields may include:

• Ontology Category
• Attribute Profile
• Assignment
• Confidence

The assigned category determines which product facts are expected and how enrichment is evaluated.

Enriched Product Facts

This section shows product facts identified by Verada AI.

Facts are separated into:

• Core Facts — required or primary facts for the category
• Additional Facts — supporting facts that improve filtering, comparison, search, recommendations, and AI answers

Each fact includes:

• Attribute
• Value
• Normalized value
• Confidence

Generated Search Language

This section shows search language generated from product evidence.

It may include:

• Primary terms
• Long-tail terms
• Alternate terms
• Localized terms
• Negative terms
• Script variants
• Search tags
• Creative tags

For Japanese catalogs, script variants may include original script, hiragana, katakana, and romaji.

Source JSON

The Source JSON section is primarily used for inspection and troubleshooting.

It may include:

• Uploaded Catalog data
• Product Export data

The Product Export view shows how enriched product data can be prepared for downstream systems such as search, recommendations, feeds, AI answers, and product export workflows.

Text Tab

The Text tab is reserved for future text enrichment capabilities.

Text enrichment focuses on improving product language while preserving source facts and brand tone.

Planned areas may include:

• Titles
• Descriptions
• Bullets
• SEO metadata
• Claims
• Tone and clarity

Review Tab

The Review tab is reserved for future review and approval workflows.

Review workflows may include:

• Low-confidence suggestions
• Sensitive claims
• Manual overrides
• Approval queues
• Change history
• Write-back controls

This will help teams decide which changes can be applied automatically and which should be reviewed by a person first.

Common Terms

Term	Meaning
Uploaded catalog score	Score based only on source catalog data.
Enriched score	Score after Verada AI enrichment.
Score uplift	Difference between uploaded catalog score and enriched score.
Ontology category	Standardized category assigned by Verada AI.
Attribute profile	Category-specific set of expected product facts.
Core facts	Important facts expected for a product category.
Filter facts	Facts useful for filters, comparison, feeds, and recommendations.
Search-ready	Product has enough language and evidence to support discovery.
Evidence quality	Whether enriched facts are grounded in product evidence.
Product intelligence	Structured enriched product understanding available for downstream use.
Variant signals	Data that helps identify product options such as size, color, shade, scent, or other supported choices.

Best Practices

To improve Catalog Intelligence results:

• Provide complete category paths, not only broad category names.
• Include product descriptions when available.
• Include high-quality product images.
• Provide product identifiers such as GTIN, barcode, MPN, vendor SKU, or manufacturer part number when available.
• Provide variant relationships and option values where products have meaningful options.
• Include category-specific attributes such as material, color, size, pattern, fit, care, compatibility, ingredients, or dimensions where relevant.
• Keep catalog imports current so pricing, availability, and product details stay accurate.