Generative Engine Optimization
Be the answerAI gives.
Search is moving from links to answers. Classic SEO gets you ranked; AI SEO gets you cited by ChatGPT, Claude, Perplexity, and Google's AI Overviews. One engagement covers both — implemented on your site, measured before and after.
What is AI SEO?
A growing share of search never reaches a results page. People ask ChatGPT, Claude, and Perplexity — or read Google's AI Overview — and act on the answer. Those systems don't rank ten blue links; they assemble one answer from sources they can parse and verify.
AI SEO — generative engine optimization — is the practice of being one of those sources: clean semantic HTML, structured data, an llms.txt, consistent entity naming, and content shaped to answer the questions your customers actually ask. It's not a replacement for classic SEO. It's the layer on top — and this service does both.
| Classic SEO | AI SEO (GEO) | |
|---|---|---|
| Where you win | A spot on the search results page | Inside the answer itself |
| Unit of success | Rankings and clicks | Citations and mentions in AI answers |
| Optimized for | Crawlers and keywords | Models, entities, and questions |
| Core artifacts | Titles, links, sitemaps | Structured data, llms.txt, entity clarity |
| How it’s measured | Rank tracking, organic traffic | Before/after AI-answer citation tracking |
Not either/or — assistants read the same web Google crawls, so the two columns reinforce each other. One engagement covers both.
Who Is This For?
Anyone whose customers ask questions — which is everyone. It matters most where buying decisions start with a question to an assistant.
Founders
When a prospect asks an assistant "who does X?", the answer is your new first impression — and it's assembled from whatever the machines can read about you. Early-stage sites are usually the least machine-legible of all.
What you get:
- Show up when buyers ask assistants about your space
- A machine-readable site without hiring a specialist
- One fixed engagement, clearly scoped
- Before/after measurement you can share
What You Get
Found by humans.Cited by machines.
A fixed-scope AI SEO engagement, done end-to-end on your actual site. We audit how both search engines and AI assistants read you today, implement the full checklist below, verify every change, and hand you a before/after report — implementation and measurement, not a slide deck of recommendations.
Technical SEO audit + fixes
Crawlability, speed, canonical hygiene, sitemap and robots configuration — audited, then actually fixed.
Structured data (JSON-LD)
Schema.org markup for your organization, offers, and FAQs, so machines can quote you accurately.
llms.txt + AI-crawler readiness
A factual llms.txt for assistants, and crawler access configured so AI systems can actually read your site.
Content & entity optimization
Consistent naming for who you are and what you sell, and key pages reshaped to answer real questions.
AI-answer citation tracking
The category questions your customers ask, run against major assistants — recorded before and after the work.
Before/after report
Every change shipped, every check re-run, and an honest comparison of how assistants answered — then and now. Then the engagement is finished, and whether we stay on to maintain it is your call.
Four Weeks. Audit to Evidence.
A fixed arc: audit → implement → verify → report. What comes after it is a choice made at the report, not a retainer that starts by default.
Audit
A full crawl and review: technical SEO health, structured-data inventory, AI-crawler access, entity consistency, and content gaps — plus a baseline snapshot of how the major assistants answer your category questions right now.
Implement
The fixes go in: technical SEO corrections, JSON-LD structured data, llms.txt, robots and sitemap hygiene, and content and entity edits — applied directly to your site, or handed over as exact, copy-pasteable changes.
Verify
Every change gets validated: markup checked against schema validators, crawls re-run, pages re-tested for how assistants and crawlers actually parse them, and anything that regressed gets fixed.
Report
You get the before/after report: every change shipped, the citation snapshot re-run against the same prompts, what moved, what hasn’t yet, and what to keep watching — walked through live with your team. What follows the report is decided then: a maintenance contract, or your own team and the report.
The Report, In Practice
A condensed sample built around two fictional businesses — Stylish, a B2C fashion retailer, and Millbrook Trade, a B2B furniture maker. The companies and numbers are invented; the structure — findings, severities, fixes, before/after citations, and code-level diffs — is exactly the format you'll receive.
Stylish is a fictional online fashion retailer. Its sample audit shows the classic B2C pattern: rich pages for humans, almost nothing machine-readable behind them.
Findings
Fictional sample — the format you’ll receive| Issue | Severity | Fix |
|---|---|---|
| Product pages have no structured data — price, sizes, and availability live only in styled markup | High | Add schema.org/Product JSON-LD with Offer, price, and availability to every product template |
| Brand entity is ambiguous: “Stylish”, “Stylish Shop”, and “stylish-store” used interchangeably across the site and profiles | High | Standardize one entity name, add Organization JSON-LD with sameAs links to every official profile |
| robots.txt blocks all unknown crawlers, which silently includes the major AI assistants | Medium | Allow reputable AI crawlers explicitly and publish an llms.txt describing the store factually |
| Category pages are image grids with no text answering “best X for Y” questions shoppers ask assistants | Medium | Add short, factual buying-guidance copy per category, structured as questions and answers |
| Size guide exists only as an image, so its contents are invisible to crawlers and assistants | Low | Republish the size guide as a semantic HTML table with proper headers |
AI-answer citations, before and after
Illustrative mock — real reports use live transcriptsPrompt tracked: “What are good online shops for sustainable everyday basics?”
The assistant recommends four competitors with brief descriptions. Stylish is not mentioned — its sustainability certifications exist only in a PDF lookbook the assistant never sees.
The assistant’s answer now includes: “Stylish — an online retailer focused on certified organic everyday basics, with published material and sizing information.” — cited to stylish.example
Structured-data diff — product template
templates/product.html<h1>Organic Cotton Crew Tee</h1>- <div class="price">$48</div>+ <div class="price" itemprop="offers">$48</div>+ <script type="application/ld+json">+ {+ "@context": "https://schema.org",+ "@type": "Product",+ "name": "Organic Cotton Crew Tee",+ "brand": { "@type": "Brand", "name": "Stylish" },+ "offers": {+ "@type": "Offer",+ "price": "48.00",+ "priceCurrency": "USD",+ "availability": "https://schema.org/InStock"+ }+ }+ </script>
Fictional sample: Stylish and Millbrook Trade do not exist, and nothing above is a claim about any real company. Your report is built from your site, your category prompts, and live assistant transcripts.
One Engagement. Everything Implemented.
What You're Buying
Implementation and measurement — not a slide deck of recommendations, and not a ranking guarantee. Nobody can honestly promise what Google or an AI assistant will do next week.
What we guarantee: every item in the checklist shipped on your site, and an honest before/after report of how assistants answered.
Ready to be citable?
One fixed-scope engagement, purchased directly through checkout — the full price is shown on the Stripe page before you pay. We kick off with a scoping call. Maintenance afterwards is a separate, optional contract.
Not sure it fits your site yet? Book a call first and we'll tell you honestly.
What's Included
After the Report
The engagement ends with the report. What happens after it is decided then, by you, not by default. We offer a maintenance contract: recurring, with its own scope and notice period, agreed separately from the engagement. Take it, or run with the report and your own team. Either is a fine answer.
The changes keep working as the platforms change
When Google or an AI assistant changes how it reads or cites a site, or a CMS update breaks the structured data or llms.txt, we fix or re-route the implementation. That is the part nobody can build around in advance.
A scheduled re-run
The same prompts run again on a cadence written into the contract: what is still cited, what has drifted, and what to fix.
First call when something breaks
You contact us before your developer or the platform. We work out whether it is the site, the markup, or the assistant, and fix it or say plainly who can.
The checklist kept current
When a platform or a rule changes, the written checklist and the guidance your team follows change with it.
Everything is already yours
Every change shipped lives in your site and your CMS, in your accounts. There is nothing to transfer and nothing to switch off.
The report is written for this
Every change shipped, every check re-run, and what to keep watching. It assumes nobody from our side is around.
Keeping it working is yours
Once the report is delivered without a contract, the site, the markup underneath it, and what the assistants do with it are your business to run. That is the honest edge, and it is why the choice is made at the report rather than assumed.
More work later is scoped separately
A new set of pages or a change of scope is its own piece of work, agreed and written down before it starts.
The contract covers the changes we shipped and the checklist we wrote. It does not cover the platforms themselves: what Google or an AI assistant chooses to show, their outages, or their rules. Nor does it cover changes made on your side without telling us. Those edges are written into the contract, not discovered afterwards.
Ryan
AI-Native UX Leader & UX Researcher
Your engagement isn't handed off to a junior operator. The audit, the implementation, and the final report come from a senior UX/UI designer, researcher, and prototyper — years of hands-on product experience applied to how machines read your site.
- 20+Years Experience
- 4Continents
- 15+Industries
- B2B & B2CBoth Worlds
- Cloud Infrastructure
- Enterprise Software
- Healthcare
- E-commerce