Initial Client Overview
Initial Challenges
The engagement began with a sensitive dynamic:
- Referral-based business model. The partners were hesitant to alter a strategy that had sustained the firm for decades.
- Design-first principals. Leadership viewed the website as a visual showcase of their work, not a business development tool, and were resistant to “SEO content” cluttering the site.
- Zero baseline in search visibility. Despite marquee projects, the firm ranked for virtually no high-intent architecture terms (e.g., “hospital architecture firm,” “university architecture firm”).
- Lack of content hierarchy. The website contained a sleek but shallow portfolio section and no keyword-targeted editorial hub.
Initial Strategy
Our SEO program combined strategic planning with respect for the firm’s design sensibility.
Keyword Prioritization: We conducted an in-depth keyword study to identify the highest-value terms in healthcare, civic, and higher education architecture. These informed the structure of all new SEO pages.
Projects Section Buildout: Rather than force generic landing pages, we architected keyword-rich pages under the Projects menu. Each project type (e.g., “Hospital Architecture Firm,” “University Architecture Firm”) lived in this section, aligning seamlessly with the principals’ desire for a design-forward presentation while creating SEO visibility.
Editorial Content: To complement project pages, we launched a blog that explored industry topics such as sustainable hospital design and trends in educational architecture, which targeted secondary keywords. Content was styled to match the firm’s branding, and doubled as a resource for backlinks.
Conversion Pathways: Each SEO page included subtle but clear CTAs to inquire about services, ensuring that visitors could convert without disrupting the visual aesthetic.
Initial Results
By balancing design integrity with strategic SEO, we transformed the firm’s website into a lead generation engine. Over the first two years of our campaign, until the firm was acquired by a leading architecture company, SEO became a core source of business development.
| Metric | Baseline | Year 1 | Year 2 (At Acquisition) |
| Average Monthly Qualified SEO Leads | 0 | 6 | 11 |
| High-Value Keywords in Top 3 | 0 | 18 | 47 |
| Organic Traffic to Project Pages | ~250/month | ~2,100/month | ~4,300/month |
Our client went from zero SEO-driven leads to a reliable pipeline of 10 to 12 MQLs per month. After 2 years, the firm was acquired by a national architecture & design firm. Over the following years, our client’ practice was folded in and came to operate primarily through their acquirer’s website, and the SEO engine it had built carried forward into the next phase of work.
Phase 2: Generative Engine Optimization
GEO Challenges
- Invisible in AI answers. Despite strong rankings, the firm was rarely named when buyers asked AI assistants for firm recommendations. Generalist directories and larger national firms dominated the responses.
- Attribution gap. LLMs frequently described the firm’s expertise without crediting it as a source, meaning the practice earned no discovery value from content it had originated.
- Preserving the brand, again. As in the SEO phase, leadership was protective of the site’s visual elegance and wary of “robotic” content added purely to satisfy machines.
- A moving target. AI answers were volatile. Citations shifted week to week across ChatGPT, Perplexity, Gemini, and Google AI Overviews, with no established playbook for measuring or influencing them.
GEO Strategy
We extended the same design-respectful discipline from the SEO phase into a generative-engine program built specifically to convert AI visibility into qualified leads.
Prompt & Question Research: We mapped the actual prompts buyers type into AI assistants, such as “best hospital architecture firms,” “who designs LEED-certified university buildings,” and “top civic architects in New York.” From these, we built a tracked set of roughly 180 high-intent prompts to target and monitor, shifting our focus from keywords to buyer questions.
Entity Authority Building: We made our client legible to large language models as a distinct, credible entity. This meant expanding structured data, standardizing the firm’s descriptors across the web, strengthening its knowledge-base and reference-grade citations, and reinforcing the associations (healthcare, civic, and higher-education architecture) carried over from the original practice that we wanted the models to learn.
Answer-Ready Content: We restructured the Projects pages and blog on the parent firm’s site into formats LLMs can extract and attribute, including crisp definitional lead sentences, concrete project statistics, and FAQ blocks. This work was done inside the firm’s established design system, so the site stayed visually elegant while becoming machine-quotable.
Trusted-Source Seeding: Because AI engines lean on third-party sources, we earned the firm placement in the references models trust most, including industry rankings, “top firms” roundups, association directories, and press. As a result, the parent firm surfaced in answers even when its own site wasn’t the cited source.
Conversion Continuity: We preserved the subtle inquiry CTAs from the SEO era and added lightweight, AI-friendly contact pathways, capturing the growing stream of AI-referred visitors with the same low-friction, on-brand conversion flow.
AI Visibility Measurement: We stood up monthly tracking of citation share and sentiment across ChatGPT, Perplexity, Gemini, and Google AI Overviews, feeding a prioritization loop that told us which prompts and sources to influence next.
GEO Results
The GEO program built on the SEO foundation the practice had established and, by 2025, made the parent firm a common AI-recommended answer for healthcare and civic architecture in its region. This visibility translated directly into pipeline: AI-driven discovery grew from a negligible share into roughly half of the firm’s qualified inbound.
| GEO Metric | GEO Baseline |
GEO Year 1 | GEO Year 2 |
| Average Monthly AI-Sourced Qualified Leads | 1 | 6 | 15 |
| Average Monthly Qualified Leads (SEO + GEO) | 12 | 21 | 32 |
| AI Answer Citation Share (target prompts) | 3% | 19% | 37% |
| High-Intent Prompts Citing the Firm (of ~180) | 5 | 34 | 68 |
| AI-Referred Website Sessions | ~180/month | ~1,440/month | ~3,690/month |
By the end of the second year of GEO, the healthcare and civic practice, now operating entirely under the parent firm, was generating roughly 32 qualified leads per month across SEO and GEO. That is nearly triple its SEO-only peak, with about half now originating from AI-driven discovery. The firm appeared in more than a third of the buying-intent prompts its clients ask AI assistants, was frequently named alongside far larger national practices, and has done so without compromising the visual elegance of its brand.