01
Clarity
One claim per page, stated plainly and early. If the answer to “what do they do” takes three paragraphs to assemble, it will not be assembled.
AI search readiness
Getting quoted when nobody sees a results page.
People ask an assistant now and take the answer. If it does not know who you are, you were not beaten — you were not in the room.
What changed
There is no page two any more. There is no page one either.
A results page gives you ten chances to be noticed. An answer gives you one sentence, and either you are in it or you are not.
When someone asks an assistant which agency to use, or what a thing costs, or who does this in Northamptonshire, they get a paragraph. Two or three names, sometimes with links. Nobody scrolls past it, because there is nothing to scroll past.
Models decide what goes in that paragraph from what they can read and verify. A page that buries its claim in a carousel, or states it only in an image, or describes the company differently on every page, gives them nothing to hold on to.
That is the whole job: make the true things about you easy to find, easy to attribute, and consistent enough that a model will repeat them without hedging.
The framework
CLEAR.
Five things a page needs before an assistant will use it as a source. We built it because the SEO checklists were written for a search engine that shows ten links.
02
Legibility
Readable without running scripts. Content that only appears after JavaScript, or lives inside an image, is content a crawler has to guess at.
03
Evidence
Numbers, names, dates and sources. Models weight specifics: “over 8,000 bookings” survives a summary in a way “trusted by many” does not.
04
Authority
The same entity everywhere. One name, one address, one description, marked up so the company on your contact page and the one in a directory are provably the same.
05
Resilience
It still holds when it is quoted out of context. A sentence that only makes sense next to the heading above it will be repeated wrong or dropped.
What you get
The work itself.
Markup
- Organisation and service schema
- FAQ and article structured data
- Entity consistency across the site
- Canonicals that point at real URLs
Content
- Claims moved up the page
- Specifics in place of adjectives
- Answer-shaped headings
- Rewrites where marking up will not fix it
Machine-readable
- llms.txt written and maintained
- robots and sitemap sanity
- Server-rendered so nothing needs executing
- Before-and-after citation testing
Proof
We did it to ourselves first.
This site carries organisation, service and FAQ schema, one consistent entity description, and an llms.txt that states plainly what OYNK is and what it should not be called. Every page is server-rendered, so nothing has to be executed to be read.
It is also generated rather than hand-maintained, which matters more than it sounds: the description of the company is written once and appears identically everywhere, so there is no page quietly disagreeing with the others.
You can read the file at oynk.co.uk/llms.txt. It is the same one we would write for you.
Questions
The ones we get asked.
What is AI search readiness?
Making your site something an AI assistant can read, trust and quote. That is a different job from ranking: there is no results page to come first on, so what matters is whether a model can find a clear claim, tie it to you, and repeat it without hedging.
What is CLEAR?
Our framework for it. Clarity, Legibility, Evidence, Authority, Resilience — five things a page needs before an assistant will use it as a source. We built it because the existing SEO checklists were written for a search engine that shows ten blue links.
Is this just SEO with a new name?
No, and the overlap is smaller than people expect. Classic SEO optimises for position in a list. This optimises for being quoted in an answer where no list appears. Fast, well-marked-up pages help both, which is why we do not sell this to anyone who still has a four-second homepage — fix that first.
What is llms.txt and do I need one?
A plain-text file that tells an assistant what your organisation is, what to call it, and which pages matter. It is not a standard anybody is obliged to honour yet, and it costs an hour. We write one for every site we build. Ours is at oynk.co.uk/llms.txt if you want to see the shape of it.
How do you measure whether it worked?
We ask the assistants. Before we start we record what the major models say when asked about you and your category, and what they cite. Then we do it again afterwards. It is not a rank tracker and it is not perfectly repeatable, but it is the actual outcome rather than a proxy for it.
How much does it cost?
It is part of a build at no extra cost — the structured data, the entity markup and the llms.txt come as standard. As a piece of work on an existing site it starts at £1,500, usually alongside an audit, because the answer is often that the content needs rewriting rather than marking up.
Get started
Ask an assistant about your own company.
Whatever it says back is your starting position. Send us the answer and we will tell you what it would take to change it.
Or email hello@oynk.co.uk