We Re-Ran Our Shopify AI Visibility Audit a Month Later. Here's What the Checklist Missed.

November 12, 2026

By Steve Merrill, Founder of WRKNG Digital, November 12, 2026

What did a Shopify AI visibility audit miss the first time around?

A Shopify AI visibility audit that only checks crawler access, Product schema, and reviews will miss the slow failures: stale structured data, a brand name that changes from page to page, and pages with no question-style answers. Those three gaps showed up when we re-ran our checklist a month later. None of them were on the original list.

On October 26 we published our 10-point Shopify AI visibility audit checklist. Robots.txt. Product schema. Price accuracy. Reviews. Third-party mentions. Live prompt tests. All still valid. Run it first if you haven't.

Then we did something I should have done before publishing. We ran a full audit on our own site.

What happened when we ran the audit on ourselves?

We got a C-minus.

Our site-wide score came back at 71 out of 100. That's the agency that writes about this stuff for a living. I'll be honest, that one stung. It took a re-audit to really learn that knowing the checklist and passing the checklist are two different skills.

The breakdown is where it got useful. Across 200 pages we audited:

  • FAQ Content averaged an F and affected 100% of pages.
  • Entity Consistency averaged an F and affected 99% of pages.
  • E-E-A-T Signals averaged a D and affected 97% of pages.
  • Content Freshness averaged a D and affected roughly half the pages.

Look at that list again. The original ten points would have passed most of our pages. Crawlers allowed. Schema valid. Pages fast. Meanwhile the things AI models use to decide whether to trust and quote a page were failing almost everywhere.

Your Shopify store probably has the same blind spot. Here are the four checks to add.

Is your structured data fresh, or just valid?

Valid schema can still be stale schema. Google's Rich Results Test will give you a green check on a Product block whose offer expired in March. It passes. AI tools still read it as old.

Freshness is how a model decides whether your price, stock, and specs are current enough to repeat. Google says it plainly in its Article structured data documentation: include datePublished and dateModified so systems can understand when content changed.

What to check on a Shopify store:

  • dateModified in your JSON-LD. Most themes skip it on Product and Collection pages. View source on a product page and search for it.
  • priceValidUntil on Offer markup. Some themes hardcode a date a year out from when the theme was installed. If that date has passed, your offer reads as expired.
  • lastmod in sitemap.xml. Shopify generates this for you, but it only updates when the resource itself changes. Google's sitemap guidelines say they use lastmod only when it's consistently accurate. Fake bumps hurt you.
  • Visible dates on guides and blog posts. A buying guide with no date looks abandoned to a model choosing between sources.

The fix usually lives in one Liquid snippet. Add a dateModified that pulls from the product's updated_at value, and clear out any hardcoded offer dates.

Does every page call your brand the same thing?

Probably not. This was our worst finding.

Entity consistency means AI models see one brand, with one name and one set of details, everywhere they look. A model builds its picture of you by stitching together your Organization schema, your title tags, og:site_name, Product brand fields, your Merchant Center feed, and outside profiles. When those disagree, the model can treat them as separate entities. Each one is weaker than the whole.

Here's what I mean. Picture a store that calls itself "Northpine Outfitters" in the footer, "NorthPine" in the Product brand field, "Northpine Outfitters LLC" in Organization schema, and "Northpine" on Instagram. A human reads that as one company. A model might not.

Our audit flagged two specific gaps on 99% of pages: missing email and telephone fields in our business schema, and brand naming that didn't match between HTML metadata and JSON-LD. Small stuff. Everywhere.

What to check:

  • Pick one exact brand string. Use it in Organization schema, the Product brand field, og:site_name, and the end of every title tag.
  • Add contact fields (email, telephone, address if you have one) to your Organization schema. Google's Organization structured data guide lists the supported properties.
  • Add sameAs links pointing to your real social and marketplace profiles so the model can connect them.
  • Match your Google Merchant Center business name to the same string. Shopping feeds are a major source for AI product answers.

Product apps cause a lot of this. A reviews app, a feed app, and your theme can each output their own brand field. Three apps, three spellings.

Are your pages written as answers to real questions?

Our FAQ Content score was an F on every single page we checked. Every one.

The original checklist told you to audit product descriptions for buyer questions. That was the right instinct, but it didn't go far enough. AI assistants match conversational prompts. A heading like "Materials" gives a model nothing to match. A heading like "Is this jacket waterproof or just water-resistant?" matches the exact thing a shopper types into ChatGPT.

What to check on your top 20 pages:

  • Are the H2s and H3s phrased as questions a buyer would ask?
  • Do the first one or two sentences under each heading answer it directly, in a way someone could quote?
  • Is there an FAQ block with real questions from support tickets or reviews?
  • Does that block have FAQPage markup that matches the visible text?

Pull questions from your customer service inbox. Those are the prompts people are already typing into AI tools. You don't have to guess.

Can an AI tell a real person stands behind your store?

Trust signals failed on 97% of our pages. That's the E-E-A-T bucket: experience, expertise, authoritativeness, trust.

The original list covered reviews and third-party mentions. It skipped the basics on your own site that tell a model a real business with real people runs this store.

  • A named founder or author on guides and blog posts, with a short bio.
  • An About page that says who you are, where you are, and how long you've been at it.
  • Footer links to your privacy policy, terms, shipping, and returns pages on every page.
  • Review or AggregateRating schema, but only if you have real reviews behind it.

When I ran my clothing company, our About page was an afterthought. Two lines and a photo. That worked fine when Facebook ads drove every sale. It doesn't hold up when a model is deciding which store to name.

How should you run the full audit now?

Start with the ten. Then add these four. That makes 14 points total, split into two passes.

  1. Access pass (original points 1, 2, 7, 9). Crawlers, schema validity, speed, llms.txt. Run once and re-check after theme changes.
  2. Trust pass (original points 3 to 6, 8, 10, plus the four new checks). Freshness, entity naming, question structure, trust signals, data accuracy, mentions, prompt tests. Run monthly.

Pass one answers whether AI can read your store at all. Whether it will quote you comes down to the second pass. Most of the drift between audits happens in that second pass, which is why a monthly cadence matters.

Frequently asked questions

What did the original 10-point Shopify AI visibility checklist miss?

Four things: schema freshness (dateModified, sitemap lastmod, expired offer dates), entity consistency (one brand name used everywhere), question-style FAQ structure, and trust signals like named authors and policy pages.

How often should I re-run an AI visibility audit on my Shopify store?

Monthly for the freshness and entity checks, quarterly for the full audit, and right away after any theme change, app install, or rebrand. Freshness problems build up quietly between audits.

What is entity consistency in AI search?

Entity consistency means your brand shows up under one exact name, with the same details, across your schema, metadata, product feeds, and outside profiles. AI models connect those signals into one entity. Mismatched names split you into several weaker ones.

Does Shopify add dateModified to structured data automatically?

It depends on your theme and apps. Many Shopify themes output Product schema without any modified date, and some SEO apps add it only to blog articles. Check your live page source to be sure.

Want us to run this audit on your store?

We ran it on ourselves first. We got a C-minus, and we're fixing it in public. We'll run the full 14-point audit on your Shopify store and tell you exactly what's breaking, page by page.

See how WRKNG Digital audits Shopify stores for AI visibility

We'll post the re-audit of our own site once the fixes are live. More on that soon.

Back to Blog