Why Ecommerce Stores Need a Knowledge Graph Audit
A Knowledge Graph audit evaluates how well Google and other AI search engines have formed structured entity records for your brand, products, categories, and people. When those records are incomplete or contradictory, your store loses eligibility for rich results, brand panels, and AI-generated citations that drive zero-click awareness.
This checklist covers 12 discrete checks across four layers: brand entity signals, product entity markup, structured data implementation, and off-site entity consistency. Each item has a binary pass/fail criterion so you can prioritize fixes by impact rather than guesswork.
Brand Entity Checks (Items 1–4)
Check 1 — Google Knowledge Panel Exists. Search your exact brand name in Google. PASS: a Knowledge Panel appears on the right side with your logo, description, and at least one social profile. FAIL: no panel appears, or the panel belongs to a different entity with a similar name.
Check 2 — Wikipedia or Wikidata Entry Is Accurate. If your brand has a Wikipedia article or Wikidata item, verify every factual claim matches your official records (founding year, headquarters, product category). PASS: all facts match. FAIL: any discrepancy exists or no Wikidata item exists for a brand that qualifies.
Check 3 — Google Business Profile Is Claimed and Complete. For brands with a physical presence or a defined service area, the GBP must be claimed. PASS: profile is verified, category is specific (e.g., "Sporting Goods Store" not "Store"), and NAP (name, address, phone) matches your website footer exactly. FAIL: unclaimed, suspended, or NAP mismatch.
Check 4 — Brand Name Appears Consistently Across All Authoritative Sources. Check Crunchbase, LinkedIn company page, Better Business Bureau, and your top-5 referring domains. PASS: brand name, URL, and description are identical across all sources. FAIL: variations like "Co.", "LLC", abbreviations, or old domain names appear in any source.
Product Entity Checks (Items 5–8)
Check 5 — Product Schema Markup Is Present on All PDPs. Crawl a sample of 50 product detail pages using a structured data validator. PASS: every PDP contains valid Product schema with at minimum name, image, description, sku, and offers properties populated. FAIL: any PDP is missing Product schema or has critical properties absent.
Check 6 — GTIN or MPN Is Populated in Product Schema. Google uses GTINs (UPC, EAN, ISBN) and MPNs to disambiguate products in its entity graph. PASS: every product that has a manufacturer-assigned identifier includes it in the schema. FAIL: identifier fields are empty even though the physical product carries a barcode.
Check 7 — AggregateRating Schema Reflects Live Review Data. If your PDPs display star ratings, the AggregateRating schema must match what users see. PASS: ratingValue, reviewCount, and bestRating in schema match the displayed values within a 24-hour sync window. FAIL: schema values are hardcoded and stale, or schema is present but reviews are not displayed on-page.
Check 8 — Product Images Pass Google's Image Requirements for Rich Results. Check Search Console's Rich Results report. PASS: product images are crawlable, at least 160×90 px, not blocked by robots.txt, and image URLs are stable (no session tokens in the URL). FAIL: any image returns a 4xx, is blocked, or uses dynamic query strings that change per session.
Structured Data Implementation Checks (Items 9–10)
Check 9 — Organization Schema Exists on the Homepage With sameAs Properties. The homepage should carry Organization (or Brand) schema that lists official social profiles and authoritative directory URLs in the sameAs array. PASS: sameAs includes at minimum the brand's LinkedIn, Facebook, Twitter/X, and one authoritative directory (Wikidata, Crunchbase, or GBP URL). FAIL: Organization schema is absent, or sameAs is an empty array.
Check 10 — BreadcrumbList Schema Matches the Visible Breadcrumb on Every Category and PDP. PASS: the itemListElement order and names in schema exactly match the on-page breadcrumb trail. FAIL: schema breadcrumb paths differ from visible paths, or breadcrumb schema is present on pages where no visible breadcrumb is rendered.
Off-Site Entity Consistency Checks (Items 11–12)
Check 11 — Social Profile Descriptions Match the Homepage Meta Description Within One Sentence. AI systems reconcile entity descriptions across sources. PASS: the brand description on LinkedIn, Facebook About, and the homepage meta description convey the same core value proposition with no conflicting claims. FAIL: one source describes the brand as B2B while another describes it as B2C, or founding years differ across profiles.
Check 12 — Press Coverage and Backlink Anchor Text Use the Canonical Brand Name. Pull your top-50 referring domains. PASS: at least 80% of brand-mention anchors use the exact canonical name, not a truncated or informal version. FAIL: widely cited articles consistently use a variation that differs from the name in your Organization schema, creating an entity disambiguation problem for crawlers.
Prioritizing Fixes After the Audit
Address checks in order of entity foundational weight. Fix Organization schema and sameAs properties first — these are the signals search engines use to confirm your entity exists before assigning any rich results. Product schema GTINs and AggregateRating accuracy come second because they directly affect Shopping tab eligibility.
Off-site consistency issues (Checks 3, 4, 11, 12) take longer to resolve because they involve third-party platforms, but they carry compounding authority over time. Set a quarterly calendar reminder to rerun all 12 checks. A single rebranding event or platform migration can cause multiple simultaneous failures, so the checklist is most valuable run as a regression test, not a one-time exercise.