Structured data is the difference between a business that gets cited by AI and one that gets skipped. Five steps, one afternoon, zero technical background required.
Most Brooklyn independent businesses have a website. Almost none of them have structured data that AI engines can actually read. That gap is why ChatGPT names your competitor when someone asks for "the best optometrist in Crown Heights" and not you. This audit closes that gap.
Step 1: Check What Schema You Have Right Now
Before fixing anything, know your baseline.
Go to validator.schema.org. Paste your homepage URL. Hit "Run Test."
If the result is empty or shows only generic WebSite markup, you're starting from zero. That's fine. At least you know. Most Brooklyn sites we audit fall into one of two buckets: zero schema, or schema that was auto-generated by a page builder and contains errors. Both are fixable in an afternoon.
Log what you see. Write down every type that appears. Common ones include LocalBusiness, Organization, Product, and BreadcrumbList. If you don't see LocalBusiness or a more specific subtype like MedicalBusiness, FoodEstablishment, or SportsActivityLocation, that's your first problem.
Step 2: Audit Your NAP Data for Consistency
NAP stands for Name, Address, Phone. It sounds basic. It is basic. AI engines still get confused by inconsistencies across sources.
Pull up your schema output from Step 1. Then open four tabs: your Google Business Profile, your Yelp page, your Facebook page, and your website footer. Compare the business name, address, and phone number across all five sources, character by character.
Common failures we find in Brooklyn clients:
- "St" vs "Street" in the address
- A suite number listed on the website but missing from GBP
- A phone number with parentheses in one place and hyphens in another
- A business name that includes "LLC" on Yelp but not anywhere else
Any mismatch is a citation confidence problem. AI systems cross-reference sources. When the data doesn't match, they either don't cite you or cite you with lower confidence. Fix every mismatch before moving on.
Step 3: Verify Your Business Type Is Specific Enough
LocalBusiness is a valid schema type. It's also nearly useless.
Schema.org has hundreds of subtypes. A jiu-jitsu gym should be marked as SportsActivityLocation. An optometry practice should be MedicalBusiness with a medicalSpecialty property. A restaurant should be FoodEstablishment with cuisine type declared.
This matters because AI engines use schema type to route queries. When someone asks ChatGPT for "BJJ private lessons in Brooklyn," it's looking for entities typed as sports or fitness businesses. Brooklyn BJJ Lessons was typed correctly from day one. That specificity was a direct factor in getting cited first for that prompt within 41 days of launch.
Action: Look up your business category on schema.org. Find the most specific subtype that accurately describes what you do. If you're a nail salon, that's BeautySalon. If you're a chiropractor, that's Physician under MedicalBusiness. More specific always beats more generic.
Step 4: Check Your Hours, Service Area, and Price Range
These three properties are the ones AI engines pull when answering "is this place open" and "is this worth my time" queries.
In your schema output from Step 1, look for:
openingHoursSpecificationoropeningHoursareaServedpriceRange
If any of these are missing, flag them. If openingHours is present but hasn't been updated since you changed your Saturday hours eight months ago, it's wrong.
The areaServed property is especially underused. A Crown Heights optometry practice that serves patients from Prospect Heights, Bed-Stuy, and Flatbush should declare all of those neighborhoods. Nostrand Optical has areaServed populated with every relevant neighborhood. That's part of why it targets "optometrist Crown Heights" in AI Overviews with four active rich results.
priceRange takes a simple string: "$," "$$," "$$$." It takes 30 seconds to add. Most sites don't have it.
Step 5: Run Google's Rich Results Test
Go to search.google.com/test/rich-results. Paste your URL. Run the test.
This tool tells you whether your structured data is valid enough to trigger rich results in Google Search. Rich results include star ratings, review counts, hours, and address information displayed directly in the search results page. They also correlate strongly with being cited in AI Overviews.
The test will show you errors and warnings separately. Errors block rich results entirely. Warnings reduce their quality. Prioritize fixing errors first.
Document everything you find:
- How many schema types are detected
- How many errors and warnings exist
- Which rich result types you're eligible for
If you launched a new site and ran this test on day one, you'd want to see zero errors and at least one eligible rich result type. That's the standard we held for Nostrand Optical on launch day. Four rich results appeared in Google within 24 hours of the site going live.
If your audit turns up a long list of errors, we run a free 15-minute audit that covers all five of these areas and identifies exactly what's blocking your visibility. Book one at signalai.agency/#audit.
What This Means for Brooklyn Independent Businesses
Schema is not an optional technical nicety. It's the layer that makes your business legible to AI. Without it, ChatGPT doesn't know what you do, where you are, or whether you're the right answer to the question being asked. With it, you become a named entity that AI systems trust and repeat.
This audit takes two to three hours for a single-location business. The outputs are a clear list of missing schema types, NAP inconsistencies, and validation errors. Fix those three categories and your AI search visibility improves in weeks, not months.
Start with Step 1 today. Run the validator. Write down what you see. The rest follows.