Found · Flagship service
A fast website that AI crawlers can read without JavaScript, that assistants can quote, and that turns the people they send into enquiries. Built in three to six weeks and handed over into your own accounts.
Typical project €8k–16k. Fixed scope, agreed in writing before work starts.
Proof on file
CLIENT · NAME CHANGED · PRE-LAUNCH15
would-be 404s caught before launch
Orvelle Dental, a premium clinic in the DACH region: every old URL, internal link and archived address followed hop by hop before the switch.
observed 2026-10-03 · hop-by-hop redirect audit (92 sitemap URLs, 132 internal links, 17 archive URLs)
03 Fit
Built for you if
Not for you if
04 The problem
My own previous website looked polished in every browser. The HTML it actually sent to crawlers contained 0 words of page content, and its sitemap listed 2 URLs. GPTBot, ClaudeBot and PerplexityBot don’t run JavaScript, so to the assistants your buyers ask, that site barely existed.
Most business websites have a milder version of the same problem. Pages assembled from a dozen plugins. Three versions of the same service page competing for one search. A company description that changes from page to page. A contact form whose messages land in spam. Each one looks small. Together they decide whether an assistant can name you, and whether the person it sends ever gets in touch.
A rebuild fixes the foundation once: what crawlers receive, how your expertise is structured, and the path from first visit to a booked call.
Signs you need this
06 What you get
Everything below is in the fixed scope. Anything extra is quoted before it starts, never discovered on an invoice.
Astro on Cloudflare’s edge by default. A lean custom WordPress theme with native blocks where your editors genuinely need WordPress. No page builder either way.
Duplicate and competing pages merged into one strong page per topic, decided by your real search data rather than taste.
Every old URL, internal link and archived address requested and followed on the new site before launch, so no link you’ve earned ends in a 404.
robots.txt, Cloudflare bot settings and content signals set on purpose, bot by bot, instead of left to defaults.
Organization, Service and Person markup in one connected graph, an llms.txt map, and a clean Markdown copy of every page for AI agents.
A multi-step enquiry form, booking, WhatsApp and attribution that tells you which page and campaign each enquiry came from.
Speed targets agreed up front and measured on a mid-range phone before handover, not promised in a pitch.
Visits, sources and enquiries counted without cookies, in an account you own.
Everything in your accounts, a written runbook and a recorded walkthrough. Bugs found in the first 30 days are mine to fix.
07 How it runs
Agents do the heavy lifting: crawling, drafting, building, testing. Every decision that changes your business waits for a person. The schedule below is the typical six-week version; smaller sites finish sooner.
Agents at work Human gate: nothing moves on until a person approves
Step 1: Week 1
Search Console, analytics and a full crawl of the current site. The agents draft the page inventory, the merge plan and the new sitemap.
Human gate You approve the sitemap and the merge plan.
Step 2: Week 2
Copy drafted from your existing pages, documents and an interview with you. Every claim is checked against a source. Design is built as real pages, not static mock-ups.
Human gate You approve the key pages on a live preview link.
Step 3: Weeks 3–4
Pages, forms, schema and integrations built to the approved spec. Automated accessibility, speed and link tests run on every change.
Step 4: Week 5
Every old address requested and followed hop by hop against the new site. Anything that would break is fixed before launch, not after.
Human gate I sign off the redirect map. You sign off go-live.
Step 5: Week 6
DNS switch, Search Console and Bing submission, crawler and uptime watch. Handover call, runbook and access review.
Human gate You confirm the handover checklist.
Step 6: 30 days after launch
Errors, AI-crawler visits and form deliveries watched daily. Anything broken is fixed at no charge.
08 Proof
One is a client project awaiting launch sign-off, and labelled that way. The other is my own live product, rebuilt static so crawlers stopped costing compute.
15 would-be 404s caught
Page builder replaced by a lean theme. 43 German pages re-optimised, 33+ English pages retranslated, and 19 treatment pages ending in a booking path.
564 pre-rendered pages
Three languages, served as static files. Crawler compute in a 45-second sample after the rebuild: 103 → 0 Worker invocations.
And this website. It replaced a single-page app that sent crawlers an empty shell. Use See as AI in the header to read this page the way a crawler that doesn’t run JavaScript reads it.
09 Price
Most rebuilds land at €8k–16k, quoted as one fixed price after a 20-minute call. These are the things that move it.
Costs more when
Costs less when
Paid by you, directly to each provider, in your own accounts. I add no markup and no hosting fee.
| What | How it works | Paid to |
|---|---|---|
| Hosting | Cloudflare. A static site usually runs on the free plan; a WordPress theme stays on your current host. | Cloudflare / your host |
| Domain | Stays where it is today, in your name. | Your registrar |
| Analytics | Cloudflare Web Analytics, cookie-free, included in the free plan. | Cloudflare |
| Forms and booking | Forms deliver straight to your inbox. Booking uses the calendar you already have. No new subscriptions required. | Nobody new |
If a running cost appears that we didn’t agree in writing, I’ve made a mistake, and I fix it.
10 Handover
No licence fees to me, no lock-in, no proprietary platform. If I disappeared tomorrow, any competent developer could pick it up from the runbook.
11 Questions
Rather talk it through first?
20 minutes with me, not a sales rep. You leave with a clear next step — whether or not we work together.
Text extracted from the HTML this server sends — no JavaScript executed, the way GPTBot, ClaudeBot and PerplexityBot fetch pages. It is also the Markdown twin I publish for AI agents.
Loading…
<!doctype html> <title>Custom Web Tools & Cookie-Free Websites | AIWebRevolution</title> … <body> <div id="root"></div> <script type="module" src="/assets/index.js"></script> </body> → 0 words of page content. Beautiful to people. Blank to ChatGPT.
Two minutes · personal reply