From invisible to AI-readable: rebuilding Tissue Alchemy for AI search
Tissue Alchemy is a Boulder bodywork practice whose dated WordPress site wasn't readable by AI engines. We fixed that. We added structured data, built a robust FAQ, and revived a blog that had been dormant since 2016 — all things AI search engines look for, and all good for standard SEO too.
AI readiness, before and after An at-a-glance comparison of how ready each version of the site is for AI search and traditional search. Both scores are out of 100 — higher is better.
WordPress → Astro — static, schema-first. Scored on how well each version can be read, recommended, and cited by AI search.
AEO score AEO = Answer Engine Optimization: how easily AI assistants can read your site and cite it when they answer questions. Driven by structured data, an llms.txt file, and clear content. It matters because more searches now end in an AI answer instead of a list of links.
+65How readable & citable the site is to AI engines (ChatGPT, Perplexity, Google AI Overviews)
SEO score SEO = Search Engine Optimization: how well your site is set up to rank in traditional search like Google. Driven by clean metadata, content depth, crawlability, and page speed. It still drives most discovery for local businesses.
+38Traditional search readiness — metadata, content depth, crawlability, performance
By the numbers The concrete, countable differences between the old site and the rebuild. Every figure is a verifiable on-page fact, not an estimate.
Schema.org structured-data types
Schema markup is labeling in a page's code that tells search and AI engines exactly what they're looking at — the business, its services, its reviews — instead of making them guess. The old site had none.
llms.txt for AI engines
A simple published file that hands AI engines like Claude, ChatGPT, and Perplexity a clean map of the site and what the practice does.
FAQ schema (Question nodes)
AI engines look for discrete question-and-answer pairs they can quote directly. Every FAQ on the new site is marked up so it can be lifted straight into an AI answer.
Pages with unique meta descriptions
The one-line summary that shows under a result in Google. Not an AI-specific thing — just a search basic that most small-business websites are missing.
Fresh education content
Traditional SEO and AI search both favor sites that publish fresh, useful content. A blog that's been silent since 2016 signals the opposite.
Local-SEO service-area cities
Tells search and AI engines which Front Range cities the practice actually serves, so it can show up in searches beyond Boulder proper.
Largest page asset (logo)
This one's pure speed: a 2 MB logo was slowing down every page. Same logo, 99% smaller, faster site — and speed matters to both visitors and search rankings.
Broken internal links
Dead links frustrate visitors and waste a search engine's crawl. Every internal link on the new site was checked and works.
The problem: nobody searches for "Tissue Alchemy" What was holding the original site back from showing up in AI search and traditional search.
People who need this work don't search the practice's name — they search for massage, physical therapy, or bodywork, or they ask something like "what do I do if physical therapy isn't working?" Those are the searches that bring in new clients, so the site had to show up for all of them, in both traditional search and AI search.
The old WordPress site couldn't. It gave search and AI engines almost nothing to work with: no schema markup, no llms.txt, missing meta descriptions, and thin pages with duplicated calls-to-action. The sitemap was padded with demo, test, and duplicate pages, and the blog had been dormant since 2016 — two posts, nearly a decade old. When someone asked an AI assistant about a fascia specialist in Boulder, there was nothing structured for it to cite.
What we built The specific work done in the rebuild to make the site readable, citable, and well-ranked.
Structured data on every page
Site-wide LocalBusiness and Person schema, plus per-page Service, FAQPage, MedicalCondition, Article, Review, BreadcrumbList, and a DefinedTermSet glossary — 11 schema types in all, so AI engines can identify the practice, its services, and its expertise.
An AI-readable content layer
A published llms.txt that hands AI engines a clean map of the site, a canonical 'what is Tissue Alchemy' definition repeated in key places for entity clarity, and a 15-term plain-English glossary with deep-linkable definitions.
Deep, honest topical content
Seven condition pages (sciatica, frozen shoulder, TMJ, plantar fasciitis and more), four 'how this compares' pages (vs Rolfing, PT, classic myofascial release, traditional massage), four sport-specific pages for Front Range athletes, and three deep-dive science articles — each written in the practice's own voice with explicit 'when this isn't the right tool' referral language.
Local visibility & performance
A 20-city Front Range service area in schema, a dedicated /visiting/ logistics page, and a performance pass that cut the logo from 2 MB to 20 KB and deferred animation scripts so pages render fast.
New capabilities the old site never had Capabilities the rebuild added that the original site simply didn't have.
- ✓ Schema.org structured data across every page (LocalBusiness, Service, FAQ, MedicalCondition, Review)
- ✓ llms.txt so ChatGPT, Perplexity, and Google AI Overviews can read and cite the practice
- ✓ Condition-specific landing pages for the problems clients actually search
- ✓ 'How this compares' pages vs Rolfing, physical therapy, and traditional massage
- ✓ Sport-specific pages for Boulder runners, cyclists, climbers, and skiers
- ✓ An 18-question FAQ with FAQPage schema, plus per-page FAQs
- ✓ A plain-English, deep-linkable glossary of fascia and bodywork terms
- ✓ 20-city Front Range local-SEO footprint and a visiting/logistics page
- ✓ A Keystatic CMS so the practice can edit its own content — no developer needed
- ✓ Accessibility and performance: reduced-motion support, alt text everywhere, a 99% smaller logo
How we scored this
These are approximate, rubric-based readiness scores — Averde's assessment of each site's public, on-page markup in June 2026, weighted across structured-data coverage, AI-crawlability (llms.txt and a clean sitemap), content depth, local-search signals, metadata, and page hygiene. They're rounded and carry a few points of judgment either way; they are not certified tool output. They deliberately exclude off-page factors — backlinks, domain authority, current Google rankings, and organic traffic — so the original site's roughly decade of domain history likely carries real search equity this on-page score doesn't credit. Measured performance (e.g. Lighthouse / Core Web Vitals) is also not yet folded in. 'Before' is the previous WordPress site as it was live at tissuealchemy.com in June 2026, assessed from its homepage and sitemap; 'after' is the rebuilt site now live at the same domain.
Want to see where your site stands?
Run the free Website Audit to see how AI search reads your business today — and what a rebuild could change.
Run my free AI audit