Product Data Reference

Use this page when you need to query product-level DataHawk data or join product tables together. For conceptual explanations, start with DataHawk Data Concepts.

Identifier fields

Product detail feature flags

PRODUCT.PRODUCT_DETAILED stores product-page attributes collected from tracked products. The FEATURES column lists activated page features detected for a product.

For Walmart products, FEATURES can include:

This value appears only when DataHawk detects the feature on the page. If the section is not active or not detected, about_the_brand is absent from the FEATURES list.

Star-rating breakdown

PRODUCT.PRODUCT_DETAILED includes each product's star-rating breakdown: the share of ratings at each star level, from 1 to 5 stars. If a product has ratings but all five percentages are 0, DataHawk shows them as NULL, because that breakdown wasn't captured.

A single 0 is still a real value. For what NULL means, the effect on rating alerts, and the current known gap, see Ratings and Reviews.

Parent and child ASIN joins

When joining product tables, confirm whether the table is parent-level, child-level, or mixed.

• PRODUCT.PRODUCT_LISTING contains one row per ASIN, which may be a parent or child.
• To get all children of a parent, filter where PARENT_ASIN = '<your_parent_asin>' and ASIN != PARENT_ASIN.
• Some visibility tables are parent-oriented, while sales and inventory often need child or SKU granularity.
• When joining public product data to private Seller data, use the ASIN/SKU mapping available in private account tables or curated report tables.

Listing quality tables

Listing Quality Analysis data lives in the PRODUCT schema and refreshes daily.

Null and empty values

• Null value: DataHawk could not extract this data, so the state is unknown.
• Empty value: DataHawk confirmed the characteristic is absent. For example, an empty "Emojis in title" field means no emojis were found.

Score formula

Section score = sum(rule_compliances * rule_weights) / max_section_points
Global LQS = (Title Score * 0.4) + (Media Score * 0.3) + (Bullets Score * 0.2) + (Description Score * 0.1)

Listing issues (compliance & catalog problems)

Amazon flags catalog and compliance problems on your listings through the Listings Items API: search-suppressed active listings, inactive listings missing an offer or price, items pending approval, and more. DataHawk captures these in SELLING_PARTNER.LISTING_ITEMS.

Where the data comes from: LISTING_ITEM_DETAILS:issues is a semi-structured array on each listing item, one entry per open issue, with a code, severity, message, categories, and attributeNames. A single SKU can have multiple issues open at once.

Recipe: current listing issues per SKU

WITH ranked AS (
    SELECT
        WORKSPACE_ID,
        ACCOUNT_KEY,
        SELLER_NAME,
        MARKETPLACE_KEY,
        SKU,
        ASIN,
        ITEM_NAME,
        EXTRACT_TIME,
        issue.value:code::STRING AS ISSUE_CODE,
        issue.value:severity::STRING AS ISSUE_SEVERITY,
        issue.value:message::STRING AS ISSUE_MESSAGE,
        issue.value:categories::ARRAY AS ISSUE_CATEGORIES,
        issue.value:attributeNames::ARRAY AS ISSUE_ATTRIBUTE_NAMES,
        ROW_NUMBER() OVER (
            PARTITION BY ACCOUNT_KEY, MARKETPLACE_KEY, SKU, issue.value:code::STRING
            ORDER BY EXTRACT_TIME DESC
        ) AS rn
    FROM SELLING_PARTNER.LISTING_ITEMS,
        LATERAL FLATTEN(input => LISTING_ITEM_DETAILS:issues) AS issue
    WHERE ARRAY_SIZE(LISTING_ITEM_DETAILS:issues) > 0
      AND EXTRACT_TIME >= DATEADD('day', -3, CURRENT_TIMESTAMP())
)
SELECT
    WORKSPACE_ID, ACCOUNT_KEY, SELLER_NAME, MARKETPLACE_KEY, SKU, ASIN, ITEM_NAME, EXTRACT_TIME,
    ISSUE_CODE, ISSUE_SEVERITY, ISSUE_MESSAGE, ISSUE_CATEGORIES, ISSUE_ATTRIBUTE_NAMES
FROM ranked
WHERE rn = 1

This dedupes to the latest snapshot of each open issue per SKU, so you don't see the same issue repeated across days. The -3 days window only controls which snapshot the dedup uses as "current". Widen it if a SKU's most recent snapshot is older than that.

Competitor detection table

DataHawk scores candidate competitor pairs using signals such as semantic title similarity, price comparison, shared category depth, keyword intersection, and co-occurrence on listing and search pages. Pairs must reach a 90% confidence threshold to be included.

Pricing and offers tables

AMZN_PRODUCT_COMPETITIVE_PRICES is sourced from SP-API Product Pricing v2022-05-01 (getCompetitiveSummary). It refreshes daily for tracked products and seller listings, and quarterly for other products connected to DataHawk.

AMZN_PRODUCT_FEATURED_OFFER columns