Case Study
From Website Copy to AI-Readable Authority
An AEO/GEO Website Optimization Case Study
Client context
Finlay Systems’ own consulting website — an internal project, not a paid client engagement. The goal was not to guarantee AI rankings or “hack” AI search — it was to make Finlay Systems easier for AI systems to understand, classify, extract, and recommend when users ask questions about AI consulting, workflow automation, and AI visibility for NYC real estate teams.
The challenge
Traditional SEO helps websites rank in search results. AEO and GEO are different — they focus on whether AI systems can understand a business well enough to mention it, cite it, summarize it, or recommend it inside an AI-generated answer. The site had none of the foundation that requires: no structured schema markup, no explicit entity language, thin FAQ content, unclear E-E-A-T signals, and no stated, repeatable relationship between David Finlay, Finlay Systems, AI consulting, NYC real estate, and AEO/GEO itself.
AI systems cannot recommend what they cannot clearly understand.
The cost of the problem
This wasn’t a time-savings problem, it was a visibility gap, and the baseline made that concrete: zero schema types deployed anywhere on the site, no FAQPage structured content on any priority page, no explicit positioning statement anchoring the website, LinkedIn profile, and service pages to the same language, and no repeatable process for testing whether AI tools could describe the business accurately at all.
The goal
Make Finlay Systems clearly understood as an AI consulting firm for high-volume NYC real estate teams across every AI-facing surface — website, schema, LinkedIn, FAQs — and put a repeatable testing process in place to monitor that understanding over time, without claiming or promising guaranteed AI rankings.
The existing workflow
- 1No schema markup anywhere on the site
- 2Positioning implied, not stated
- 3LinkedIn and website out of alignment
- 4FAQ content thin or absent
- 5No process to test AI visibility at all
- 1JSON-LD schema across every priority page
- 2One explicit positioning statement, repeated everywhere
- 3LinkedIn optimized first, then the website
- 4Buyer-intent FAQ content on every priority page
- 5A manual AI-visibility testing tracker, run on a cadence
The AI system
The core positioning statement, “I specialize in AI consulting for high-volume NYC real estate teams,” became the anchor for the website, LinkedIn profile, schema, service pages, FAQs, and testing process. Audit the site for schema and entity gaps → generate JSON-LD schema with AI assistance → implement it in WordPress → validate it with independent tools → rewrite priority page content around the same positioning statement → add a dedicated buyer-intent page → test AI visibility on a recurring cadence with real buyer-style prompts.
Implementation
Seven layers, in order:
Schema types deployed: Organization, ProfessionalService, Person, WebPage, Service, OfferCatalog, Offer, FAQPage, Article, BlogPosting, TechArticle, CreativeWork, Event, BreadcrumbList, ItemList, CollectionPage, AboutPage, ContactPage.
Human oversight and safeguards
- Every schema block validated with two independent tools (Schema.org Validator, Google Rich Results Test) before being considered live, not just deployed and assumed correct.
- The positioning statement was set deliberately, once, by David, and every downstream page, FAQ, and schema block was checked against it rather than allowed to drift.
- AI-generated schema was reviewed before implementation — AI drafted the JSON-LD, a human confirmed it matched the actual business before it shipped.
- No claim of guaranteed AI rankings anywhere in the output — the project’s own framing is explicit that AEO/GEO improves the foundation, not a promised result.
Results
These are operational, foundation-level metrics — this project is new enough that revenue-attributable results aren’t yet available, which is why they’re reported honestly as operational rather than financial.
0
Structured schema types deployed across priority pages, up from zero
0
Single positioning statement now consistent across website, LinkedIn, and schema
All
Priority pages (Home, Services, About, Contact, buyer-intent page) rewritten with explicit FAQ content
Ongoing
Manual AI-visibility tracker now run on a recurring cadence
Business impact
The biggest outcome wasn’t a single ranking — it was a reusable AEO/GEO framework. Finlay Systems went from having no structured way to know whether AI tools understood the business at all, to having schema deployed across every priority page, FAQ content addressing specific buyer questions, website and LinkedIn brought into alignment, and a testing process to monitor it going forward. A real estate team doesn’t only need to rank on Google anymore — it needs to be understood by AI systems when prospects ask “Who are the best real estate agents in the West Village?”
I couldn't tell you before this whether an AI tool would describe my business accurately. Now I have a process to actually check.— David Finlay, Founder, Finlay Systems
Client quote
An internal Finlay Systems project applied to its own site, not a client engagement — David’s quote above reflects the operator’s own perspective rather than a third-party client.
What happened next
The manual AEO/GEO testing tracker built during this project is now run on an ongoing cadence rather than a one-time check, and the same schema and positioning discipline established here now applies to every new page Finlay Systems ships, including its guides and case studies.
What this means for NYC real estate teams
Most real estate websites are built for human browsing. AEO/GEO requires them to also be built for machine understanding — clear entity signals, deployed schema, FAQ content addressing real buyer questions, and a repeatable way to check whether it’s working, rather than guessing.