How a Dental Practice Went from Position 17.6 to 4.8 in 30 Days

The Starting Point: Invisible to AI Despite Strong Fundamentals

A dental practice in the Southwest came to us with a problem they did not know they had.

On paper, everything looked solid. They had 180+ Google reviews with a 4.7 star average. Their website ranked on page one for several competitive keywords. They had been in business for over 12 years. By traditional SEO standards, they were doing well.

But when we ran an AI visibility audit, the picture was different. We tested 20 variations of queries a potential patient might ask AI platforms. Things like "best dentist near [city]" and "dentist that takes [insurance provider] in [city]" and "emergency dental care [city]."

Their average position across ChatGPT, Perplexity, and Google AI Overviews was 17.6. That number means the practice was almost never recommended. When it did appear, it was buried below a dozen competitors. In practical terms, position 17.6 is the same as being invisible. AI platforms typically recommend 3 to 5 businesses. Anything past position 5 might as well not exist.

The practice owner had never heard of AI search optimization. He was spending $3,200 per month on Google Ads and assumed that covered his digital presence. It did not.

The Diagnosis: Three Fixable Problems

We ran a full audit across their directory listings, website structured data, and Google Business Profile. Three issues emerged immediately.

None of these problems required a website redesign. None required new content. None required ad spend. They were all data problems, and data problems have data solutions.

Optimization 1: Directory Listing Cleanup (Days 1 to 7)

We started with directories because they feed the most data to AI platforms. Foursquare alone supplies roughly 70% of the local business data ChatGPT uses when generating recommendations. If your Foursquare listing is wrong, ChatGPT is working with wrong data.

Here is exactly what we did in the first week:

• Foursquare:Claimed the listing. Updated category from "Medical Center" to "Dentist" with subcategories for cosmetic dentistry, pediatric dentistry, and emergency dental services. Fixed phone number. Added full service descriptions, hours, photos, and website URL.
• Yelp: Claimed listing. Corrected address format to match Foursquare exactly. Added service descriptions and updated business hours.
• Apple Business Connect: Claimed for the first time. This listing feeds Siri and Apple Intelligence recommendations. It had never been touched.
• Bing Places: Updated to match all other listings. This feeds Microsoft Copilot recommendations.
• Data aggregators: Submitted corrected NAP data to Neustar Localeze and Data Axle, which feed dozens of secondary directories and AI data pipelines.

The entire directory cleanup took about 6 hours of actual work spread across 5 days (some platforms have verification delays). The total cost was $0 in ad spend. Every platform offers free business listings.

By the end of week one, all 14 existing listings were claimed, corrected, and consistent. We also created listings on 28 additional platforms the practice was missing from, bringing their total directory presence to 42 platforms with 100% NAP consistency.

Optimization 2: Schema Markup Implementation (Days 5 to 10)

While directory changes were propagating, we moved to the website. Schema markup is the structured data that tells AI systems what your business does in a language they can parse instantly. Without it, an AI crawler has to read through your entire site and try to extract meaning from unstructured HTML. Most do not bother.

We added four types of schema to the existing website:

• Dentist schema (a subtype of LocalBusiness) on the homepage, with the practice name, address, phone, hours, geo-coordinates, accepted insurance providers, and a list of all services offered.
• Service schema on each service page, describing the specific treatment, the area served, price ranges where applicable, and the provider.
• FAQPage schema on the FAQ section, covering 18 questions patients commonly ask. Insurance acceptance, emergency availability, parking, languages spoken, appointment scheduling process.
• AggregateRating schema linking to their Google review profile with the current star rating and review count.

Total development time: about 4 hours. We validated every schema block using Google's Rich Results Test and schema.org's validator to ensure zero errors. One malformed schema block can cause AI to ignore all your structured data, so validation is not optional.

The website itself did not change visually. Same design, same content, same hosting. The only change was invisible structured data in the HTML that AI systems could now read.

Optimization 3: Google Business Profile Overhaul (Days 7 to 14)

Google Business Profile accounts for roughly 32% of the ranking factors in Google AI Overviews. That alone makes it worth getting right. But GBP also feeds data to Apple Maps, Bing, and other aggregators that AI platforms reference.

The practice's GBP was technically "claimed" but barely filled out. Here is what we changed:

• Categories: Added 4 secondary categories (Cosmetic Dentist, Pediatric Dentist, Emergency Dental Service, Dental Implants Provider) to the existing primary category.
• Services: Listed all 8 services individually with descriptions. Previously, zero services were listed.
• Business description: Rewrote from one generic sentence to a 750-character description mentioning specific services, the city name, neighborhoods served, and insurance accepted.
• Q&A section: Added 12 questions and answers covering insurance, emergency hours, parking, languages spoken, and common procedures.
• Photos: Uploaded 15 new photos of the office, team, and equipment with proper geo-tagged metadata.
• Google Posts: Published 4 posts in the first two weeks covering seasonal topics and service highlights.

This work took about 3 hours total. Most of it was writing the service descriptions and Q&A content. The practice owner provided the raw information. We formatted and optimized it for AI consumption.

The Results: 30 Days Later

We re-ran the same 20-query audit at the 30-day mark. The results:

Position 4.8 means the practice now consistently appears in the top 5 recommendations across AI platforms. That is the zone where AI actually recommends you to users. Anything above position 5 and you are functionally invisible.

The most dramatic improvement was on ChatGPT, where they went from appearing in 1 of 20 test queries to 12 of 20. That is a direct result of the Foursquare cleanup. Once ChatGPT had accurate, complete data from Foursquare, it started recommending the practice consistently.

Google AI Overviews also showed significant improvement, driven primarily by the GBP overhaul and schema markup. The practice now appears in AI Overviews for more than half of their target queries, up from just 2.

What Moved the Needle Most

If we had to rank the three optimizations by impact, here is how they stack up based on the data:

• Directory cleanup (40% of improvement): Fixing Foursquare alone accounted for most of the ChatGPT gains. Claiming Apple Business Connect opened up Siri recommendations that were previously impossible.
• Schema markup (35% of improvement): Adding structured data gave AI platforms a clean, parseable data source on the practice's own website. This improved performance across all AI platforms, not just one.
• GBP overhaul (25% of improvement): The biggest impact here was on Google AI Overviews specifically. Adding services, categories, and Q&A content gave Google's AI much more data to work with.

Total hours of work: approximately 13 hours across 14 days. Total ad spend: $0. Total website redesign cost: $0.

The practice owner put it simply: "I was spending $3,200 a month on Google Ads and did not know that the fastest-growing search channel could not even find me." He has since reduced his ad spend by $1,000 per month while seeing an increase in new patient inquiries through AI referral channels.

Key Takeaways for Any Local Business

This case study is not unique to dentistry. The same pattern applies to HVAC companies, plumbers, chiropractors, pest control businesses, and every other local service provider. If you want to learn more about the fundamentals, read our complete guide to AI search optimization.

The core lessons:

• AI search is a data problem, not a marketing problem. Fix the data and the recommendations follow.
• Foursquare matters more than you think. 70% of ChatGPT's local data comes from there. Claim it first.
• Schema markup is not optional. Without structured data, AI has to guess what you do. It usually does not guess.
• GBP is 32% of Google AI Overview rankings. That alone justifies spending an hour making it complete.
• You do not need a new website, a bigger ad budget, or a marketing agency on retainer. You need accurate, complete, consistent data across the platforms AI actually checks.

Position 17.6 to 4.8 in 30 days. Three optimizations. 13 hours of work. Zero dollars in ad spend. That is what happens when you fix the data pipeline AI actually uses.

Frequently Asked Questions