AI-ready PDP checklist for Shopify brands.

Crawl access (items 1-2)

1. robots.txt allows AI crawlers. GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot, ClaudeBot, Google-Extended. Verify by visiting yourstore.com/robots.txt. The default Shopify file is usually correct; custom rules often are not.
2. llms.txt at the root, current. Markdown, served as text/plain, refreshed when the catalog changes meaningfully. See how to verify llms.txt.

Structured data (items 3-5)

3. Complete Product JSON-LD. name, brand, sku, gtin where available, image, description, offers (with price, availability, priceCurrency), aggregateRating where 3+ reviews exist. See how to add JSON-LD to a Shopify product.
4. FAQPage JSON-LD. Five real Q&A pairs in the schema, matching visible H3 question headings verbatim. Mismatch is a negative signal.
5. BreadcrumbList JSON-LD. Reflects the actual navigation path to the PDP. Validate in the Rich Results Test.

Copy and headings (items 6-10)

6. H1 names product and function. "The Northwest Parka — heavyweight down jacket for Pacific Northwest winters" beats "The Northwest Parka."
7. First 300 characters answer the buyer's top question. Three quotable facts. No brand-voice opener. See write PDP copy for AI search.
8. Three to five H3 headings as buyer questions. Verbatim where possible. "Is this waterproof?" beats "Materials & Build."
9. Each H3 followed by a 40-to-80-word answer. Self-contained passages with at least one quotable fact each.
10. Facts not claims. Every adjective replaced with a number, material, certification, or duration. "1000-denier nylon" beats "durable."

Grounding and specs (items 11-14)

11. Visible review count above the fold. Star rating plus a number. "4.8 out of 5 across 412 reviews." Hidden review widgets fail this check.
12. Third-party grounding source. At least one link or mention from an independent review site or publication. Outside Magazine, Wirecutter, category-specific media.
13. Scannable specs block. Dimensions, weight, materials, capacity, warranty. A table or list, not a paragraph.
14. Use-case examples. At least two specific use-case sentences. "For two-night trips" and "carry-on legal on Delta and American" are use cases; "great for travel" is not.

Technical and measurement (items 15-20)

15. Variant clarity. If the product has variants, each variant's distinguishing fact is named on the PDP. Color is a variant; for an apparel SKU it should explicitly say "available in five colors."
16. Three to five internal contextual links. To related collections, comparison pages, or guides. Improves entity disambiguation.
17. Descriptive image alt text on every product image. "Northwest Parka in olive green, front view, on a model" beats "IMG_4392.jpg."
18. No duplicate boilerplate. Shipping, returns, and warranty language linked from a sitewide policy page. Pasting on every PDP dilutes the page's unique signal.
19. Mobile readability above the fold. On a 375px-wide screen, the price, title, primary fact, and CTA are visible without scrolling.
20. Enrolled in a weekly measurement loop. Category-prompt runs against ChatGPT and Perplexity, citation status tracked over time. The SKU Visibility Grader handles the free version.

How to use this checklist

Three sensible patterns:

• Solo audit. Open a PDP, walk the 20 items, score 0 or 1 per item. Below 14, this PDP is at risk and should be prioritized for rewrite. Between 14 and 17, ship the gaps and re-baseline. 18 and above, this PDP is competitive.
• Catalog rollout. Apply theme-level fixes (items 1, 2, 3, 4, 5, 17, 18, 19) once; they amortize across every PDP. Per-SKU work (items 6 to 14, 15, 16) takes one to three hours per product.
• Continuous. Run the checklist quarterly. Enroll high-revenue SKUs in the eCommerce Insights weekly measurement loop so regressions surface before they impact revenue.

If you can only fix five items first

Not every team has the cycles to ship all 20 items at once. eCommerce Insights's prioritization, based on audit data through Q1 2026:

1. Item 6, first 300 characters. The single highest-leverage rewrite. Engines weight the opening passage disproportionately, and the rewrite is achievable in under an hour per PDP.
2. Item 3, complete Product JSON-LD. The schema field engines use first for entity disambiguation. Theme-level fix; ships across every PDP at once.
3. Item 4, FAQPage JSON-LD with matching visible content. Five real Q&A on every PDP, with the JSON-LD and visible H3s sharing source.
4. Item 11, visible review count above the fold. Star rating and number. Engines weight grounded claims, and a visible review count is the most legible grounding signal.
5. Item 20, the measurement loop. Without weekly measurement, the other fixes are invisible to the team and untrustworthy to leadership. Ship the loop on day one.

Ship those five and the typical PDP moves from a 35 to 50 score range into the 65 to 75 range on the AEO Grader's rubric. The remaining 15 items take a PDP into the top quartile.

What happens after week one

Shipping the checklist is one event. Maintaining it is the work. Three patterns eCommerce Insights sees in catalogs that sustain citation lift past the initial fix:

• Schema drift detection. A theme update or app install silently strips a required field. Without monitoring, the page goes from compliant to non-compliant overnight.
• New-product launches. The team launches three new SKUs and skips items 6 through 14 because there is no time. The new SKUs underperform until someone notices.
• Competitive shifts. A competitor publishes a better-grounded PDP and overtakes your citation slot on five prompts. The signal shows up only if you are measuring.

The eCommerce Insights paid product handles all three by running the audit weekly and flagging changes. The free AEO Grader handles point-in-time checks well; sustaining the score over months is a different problem.

Twenty items, none individually hard. The work is doing all twenty on every PDP, every quarter. That work is what tools are for.

Frequently asked questions

Related jobs

Related tools

• AEO Grader — score the checklist score in one click.
• Shopify SKU Visibility Grader — confirm citation lift after the rewrite.
• Product schema generator — fill the schema items.