SEO, SEM & AI search
SEO, SEM and AI search visibility
Nobody can promise you the first position. What can be promised is that nothing on your own site is standing in the way of it.
Built and running at Lawang Estate, a property developer selling villas and land. The figures below are theirs.
01 · What it is
Most of the SEO we do happens in the code, which is why a studio that builds the site is well placed to do it: clean addresses, one page per intention, real language versions that point at each other, structured data that says what the business is, and a site that loads fast on a phone. None of that is a trick; all of it is easy to get wrong.
More detail
Search has a third audience now. Assistants like ChatGPT, Claude and Perplexity answer questions about businesses by reading their sites. We publish what they read — an llms.txt generated from the same catalogue as the pages, crawlers allowed on purpose — so the answer they give about you is yours.
Paid search (SEM) is the opposite discipline: it buys the visit today. We run it where the numbers justify it and connect it to the CRM, so what is reported is enquiries and sales, not clicks. And we will say so plainly when it does not pay.
02 · Seen working
lawangproperties.com, and this site
Sample data in the shape of the real tool. No names, no amounts.
What a search engine or an AI assistant reads from their site, and where each piece is generated from.
| Item | Where | Generated from | Status |
|---|---|---|---|
| sitemap.xml | Both sites | The page catalogue | Live |
| hreflang pairs | EN ↔ ES ↔ ID | The page map | Live |
| llms.txt | Both sites | The page catalogue | Live |
| Structured data | Every page | The page itself | Live |
| Search Console | axisworks.studio | — | Pending |
IN PRODUCTION
The developer’s site publishes in three languages with every page paired to its translations, a sitemap and an llms.txt generated by code rather than kept by hand, and an accessibility statement. axisworks.studio follows the same rules: one sheet per service, English and Spanish pairs, and new addresses submitted through IndexNow when they go live.
MODULES · 9
- Technical audit
- One page per intent
- Language pairs
- Structured data
- Generated sitemap
- llms.txt
- IndexNow
- Keyword research from real data
- SEM tied to the CRM
03 · How it flows
- 01AuditWhat stops the site being read.
- 02StructureOne address per intent and language.
- 03ContentWritten for the person searching.
- 04PublishSearch engines told the same day.
- 05MeasureBefore judging anything.
04 · What’s included
| # | Component | What it means |
|---|---|---|
| 01 | Audit | What stops your site being read today: speed, structure, duplicates, broken signals. |
| 02 | Structure | One address per intention, in each language, with its pair declared. |
| 03 | Structured data | Organisation, services, FAQ and breadcrumbs, generated from the page itself. |
| 04 | Keyword research | From your Search Console, your leads and the real results page — not an invented volume. |
| 05 | AI search | llms.txt, crawler permissions and content written to be quoted. |
| 06 | SEM | Paid search where it pays, tied to the CRM. |
| 07 | Measurement | Search Console and analytics installed before anything is judged. |
05 · Questions
Can you get us to the first position?
How long until we see results?
What is AI search optimisation?
Do you write the content?
Do we need a new website for this?
Start a project
Have something worth building well?
Tell us what you are working on. We reply in English or Spanish, usually within a day.