Research
We didn't guess. We studied it.
Most agencies make recommendations based on intuition or what worked for Google. We ran real studies — pre-registered research across multiple cities, AI platforms, and industries — to find out exactly what makes AI recommend one business over another.
AI responses analyzed
Cities studied
AI platforms tested
Industries studied (and growing)
Why research matters.
The AI search landscape is brand new. There are no established best practices yet — just a lot of people guessing. We decided to actually measure what works. Here's what makes our approach different:
Pre-registered studies
We design our research methodology before collecting data — not after. This prevents cherry-picking results and ensures our findings are honest.
Multi-city, multi-model
We test across many cities and all major AI platforms. This means our findings aren't just true for one place or one tool — they hold up everywhere.
Practical, not academic
We don't do research for research's sake. Every finding translates directly into action steps for our clients. If it doesn't help your business, we don't focus on it.
What our research found.
Across all our studies, clear patterns emerged. These findings drive everything we build for our clients.
Website structure is the #1 factor
The way your website is organized matters more than anything else. Businesses with clear, well-structured pages are dramatically more likely to be recommended by AI. This had the single largest effect in every study we've run (Cohen's d = 1.14).
Domain authority barely matters
Traditional SEO prizes backlinks and domain authority. Our research found virtually no relationship between domain authority and AI recommendations (r = -0.05 to r = 0.002 across studies). AI doesn't care how many other sites link to you.
Specialty pages are essential
82-89% of businesses that AI recommends have dedicated pages for their specific services. Businesses without these pages are nearly invisible to AI, regardless of how good their homepage is.
Almost no one is doing this
Less than 2% of businesses in our studies had the right structural setup for AI search. The opportunity for early movers is wide open — and it compounds over time.
Independent businesses outperform chains
AI models cite independent businesses at significantly higher rates than corporate chains (75.7% of citations in our PT study). Specialization and depth of information beat brand recognition.
Each AI platform picks different winners
Only about 9% of businesses are recommended across all AI platforms. Each model evaluates content differently, so optimizing for one isn't enough. You need to be visible across all of them.
Studies by industry.
We study each industry individually because what AI looks for varies by market. Here's where we've published research so far.
Published Study
Chiropractic
20 cities, 120 query templates, 3 AI models. The largest published AI citation study in the chiropractic industry.
Read the full study →Published Study
Physical Therapy
21 cities, multiple query types, 3 AI models. Research into how AI recommends physical therapy practices.
Read the full study →Coming Soon
Dentistry
Research in progress. AI citation study for the dental industry across multiple cities and platforms.
Coming Soon
Med Spas
Research in progress. AI citation study for the medical aesthetics industry.
Frequently Asked Questions
Where can I read the full research studies?
Our published studies are available on this site. We currently have full studies for chiropractic and physical therapy, with dentistry and med spa research in progress. Each study page includes methodology, findings, and practical implications.
How is your research different from other marketing studies?
Our studies are pre-registered, meaning we design the methodology before collecting data — not after. This prevents cherry-picking results. We also test across multiple cities and AI platforms, so our findings aren't specific to one location or tool.
Why does website structure matter more than domain authority for AI?
Our research found virtually no relationship between domain authority and AI recommendations (correlation near zero across studies). AI models process and understand content based on how it's organized — clear headings, dedicated service pages, and structured data matter far more than how many backlinks you have.
Do your findings apply to my industry?
The core structural findings — heading hierarchy, dedicated pages, schema markup — apply across all industries we've studied. The specific content and query patterns vary by industry, which is why we run separate studies for each vertical.
How often do you update your research?
We run new studies regularly as AI platforms evolve. Our existing studies are updated when we find significant changes in how AI models evaluate content. We're also expanding into new industries based on client demand.
Want this research applied to your business?
Every recommendation we make is backed by data. Start with our AI Search Foundation to see exactly how AI handles your market.