- Retrieval
- the mechanism that matters
- Chunks
- how content is actually consumed
- Attribution
- cited, and cited correctly
Strategy and measurement live on AI SEO.
LLM Optimization
The technical half of AI visibility. If AI SEO is deciding which answers you want to appear in, this is the work of making your content mechanically easy for a model to retrieve, quote and attribute to you rather than to someone else.
- Mechanism
- Understand this and the tactics follow.
How retrieval actually works
Most assistant answers about current information are not recalled from training. The system searches, retrieves a handful of documents, splits them into passages, and generates an answer grounded in the passages it selected. Your content competes at the passage level, not the page level.
That has direct consequences. A page burying its answer in the eleventh paragraph, after autoplaying video and three interstitials, may never have that passage selected even if it ranks well conventionally. A page stating the answer plainly under a heading that matches the question is far more likely to be chosen.
It also means attribution can fail independently of retrieval. Models routinely use content and credit the wrong source, or no source. Entity clarity is what reduces that.
- 8 areas
- One body of work, not a menu.
What’s included
Content chunking and structure
Organising pages so that self-contained, extractable answers sit under headings that match how questions are actually phrased.
Semantic HTML and heading hierarchy
Correct document structure so a parser can determine what belongs to what. Div soup with styled text is legible to humans and ambiguous to machines.
Schema and structured data
Explicit machine-readable statements about your organisation, services, authors and content, reducing the inference a model has to perform.
Entity disambiguation
Making clear which entity you are — particularly where your name collides with others — through consistent identifiers and corroborating references.
Author and expertise signals
Named, credentialed authorship with verifiable external presence, which affects whether content is treated as authoritative.
Crawler accessibility
Verifying the crawlers these systems use can actually fetch your content, including checking your CDN or WAF is not silently blocking them.
Citation and attribution tracking
Monitoring not just whether you are cited but whether you are cited correctly, and diagnosing misattribution.
Content freshness signals
Clear, honest dating and revision history, which affects selection for queries where recency matters.
- Common questions
- Asked on most first calls.
Questions
Is this just structured data with a new name?
Schema is one component. The larger part is content architecture — whether a self-contained answer exists in an extractable passage — plus entity clarity and corroboration. Adding schema to a badly structured page achieves little.
Does this help traditional SEO too?
Generally yes. Clear structure, semantic markup, unambiguous entities and genuine authorship all help conventional ranking. That overlap is the main reason this work is defensible even amid uncertainty about the engines.
Will you write content specifically for AI?
No, and we would push back on the idea. Content written for machines reads badly, and these systems increasingly detect and discount it. We structure genuinely useful content so machines can parse it.
How do we know if it is working?
Citation tracking across a defined query set, plus monitoring whether attribution names you correctly. We report both, including when the number does not move.
Our CDN might be blocking these crawlers. How would we know?
Checking that is one of the first things we do. Blocking at CDN or WAF level is common, usually unintentional, and completely removes you from consideration. It is also among the easiest problems to fix.
- Related
- Often scoped alongside this.
Related services
AI SEO
Visibility inside AI Overviews and assistant answers.
Technical SEO
Crawlability, indexation, rendering and speed. The foundation everything sits on.
Content Marketing
Topic strategy, editorial calendars and the capacity to publish on schedule.
JavaScript SEO
Rendering diagnosis and remediation on framework-driven sites.
Digital PR
Original research and data stories that earn coverage and citations.
- Next step
- We look at your site before the call.
Start with a free audit
An automated crawl plus a 30-minute walkthrough of what we can see from outside your site. If what you need isn’t something we do well, we’ll tell you.