Skip to main content

SEO & GEO

One framework for how a website is discovered, understood and judged

Search visibility, generative discoverability, content usefulness, technical health, brand consistency and infrastructure integrity are usually assessed separately, by different people, at different times. They are not separate problems.

A website is read by three audiences at once, and they want different things. Search engines need to reach it, process it and work out what each page is for. Generative systems need to identify the organisation, understand what it offers and decide whether its claims can be relied upon. People need to grasp within a few seconds whether they are in the right place.

These audiences are usually served by different disciplines. A developer handles performance, a marketer handles content, an agency handles rankings, and nobody handles the question of whether the answers add up. Website Intelligence exists because the interesting problems sit in the gaps between those responsibilities.

The framework covers eleven areas. Four run on every assessment because they apply universally. The rest are selected where they are relevant, and each one is documented in its own right below.

The useful findings are rarely inside a single discipline. They appear where two disciplines were each satisfied and nobody checked the join.

Principle

Measurement, inference and interpretation are not the same thing

A great deal of confusion in this field comes from presenting three different kinds of statement in the same voice. Whether a canonical declaration exists is a measurement. Whether a page is likely to be understood as a service page is an inference. Whether the organisation’s positioning is coherent is an interpretation.

All three are legitimate and useful. Presenting the third with the certainty of the first is not. Every finding IXSEO produces carries a confidence level, and where the evidence only supports a cautious reading, it says so rather than rounding up to a conclusion.

This is also why the automated and human parts of the process are separated deliberately. Code is better at establishing facts. People are better at deciding which facts matter.

Three kinds of statement

Measured
Established directly from the response. Present or absent, reachable or not, this many milliseconds. Not open to disagreement.
Inferred
Derived from measured evidence with a stated assumption. Sound reasoning, but the assumption could be wrong and is always disclosed.
Interpreted
A judgement about meaning or priority. Where experience earns its cost, and where a confidence level matters most.

Depth

Three levels of assessment

The distinction between them is about authorisation as much as depth. What can responsibly be shown about a website depends on who is asking and what they have proven.

No verification

Public snapshot

Passive analysis of publicly available information, available to anyone for any public domain. Indicative scores, a selection of findings and an executive preview. Sensitive technical detail is withheld.

Verification required

Domain-verified review

After control of the domain is demonstrated, findings can include the specific addresses affected, detailed infrastructure observations and organisational exposure indicators. Still passive.

Signed authorisation

Authorised assessment

Active technical testing, conducted only under a separate written agreement defining scope, systems, timing and contacts. Never triggered by anything on this website.

Domain verification confirms control of an approved domain resource. It does not automatically authorise active testing of company infrastructure.

Frequently asked questions

Is Website Intelligence a product I can subscribe to?
No. It is the framework IXSEO uses to assess websites, delivered through professional engagements. There is an automated snapshot on the homepage that applies part of the framework immediately and without charge, but there is no dashboard to log into and no monthly licence.
Do I have to use every module?
No, and most engagements do not. Four modules run by default because they apply to nearly every website. The remainder are selected where they are relevant — email deliverability matters enormously to some organisations and barely at all to others.
How much of this is automated?
The evidence gathering is entirely automated and deterministic: what exists, what returns which status, what is declared, what is measurable. Interpretation is partly assisted by a language model working within strict constraints and, in paid engagements, reviewed by a person before it reaches you.
What are the scores based on?
Observable evidence, weighted by how strongly each signal tends to affect how a site is understood. They are directional assessments intended to focus attention, not measurements of ranking. A score of 68 does not correspond to any position in any search result.
What does the free snapshot leave out?
Depth rather than breadth. It runs the same checks but shows a selection of findings rather than all of them, withholds detail on anything requiring domain verification, and does not include the human review, roadmap or strategic sequencing that a paid engagement provides.

See the framework applied to your website

The free snapshot runs the four core modules and any optional ones you select. It takes about a minute and requires nothing but a website address.