Search visibility systems for the modern web

Make your brand discoverable

A business can be excellent and still be invisible.

Atlas builds the technical, semantic and authority systems that make brands easier to find, understand and trust across modern search.

Build visibility Explore the system

State of the entity — five states, one directionAtlas works the whole route

State of the entityFive states

What Atlas measures — 100 points, eight weighted categoriesThe scale, not a score

Technical foundation20

Search architecture15

Content & intent15

Ecommerce10

International10

Entity10

Authority10

AI visibility10

Visibility riskbelow 60

Growth opportunity60–74

Strong75–89

Elite90–100

Your reading comes from a diagnostic, not from this page. The scale is published so you can see what Atlas counts before anyone quotes you a number.

Search is no longer a single results page

Seven surfaces · none consults the others

Discovery used to have one address. It now has several, and each one builds its own understanding of a brand from whatever it can read. Where the reading is thin, the brand gets substituted for one the system understands better.

That substitution is the entire commercial cost of poor visibility, and it is invisible in a rankings report.

Surface	What it reads	What happens when it cannot read you
Google, classic	Crawlable HTML, links, structured data	You rank below someone with worse product and better architecture
AI Overviews & AI Mode	Extractable statements it can attribute	You are summarised out of the answer entirely
Answer engines	Corroborated facts, cited sources	A competitor becomes the citation for your category
Ecommerce search	Product titles, attributes, taxonomy	Products are filed under the wrong category and never surface
Marketplaces	Feed data, identifiers, availability	Approved listings that earn nothing, because the fields are wrong
Branded search	Entity consistency across the web	A brand with a similar name absorbs your own demand
International search	Language, market and intent signals	One market's pages compete with another's for the same query

Visibility is a system

Six layers · each bears the ones above

Rankings are an output. This is the thing that produces them.

Search intelligence

If this is wrong: you rank, and still are not cited. Every discovery system is measured, not only the one with ten blue links.

Authority

If this is wrong: nothing outside the site confirms it exists. Not a link count — a body of evidence.

Entities

If this is wrong: a system substitutes a brand it understands better. Usually one with a similar name and clearer signals.

Language & intent

If this is wrong: you answer questions nobody asks. Demand has a different shape in every market, and sometimes it is absent.

Architecture

If this is wrong: every later decision inherits it. The only layer where a mistake compounds instead of staying local.

Technical foundation

If this is wrong: none of the five above is ever read. This layer fails silently — a canonical pointing the wrong way costs nothing visible and everything measurable.

Atlas works from 06 upward. Architecture before scale. Technical integrity before content volume. Entity clarity before authority campaigns. A defect at the base is inherited by everything resting on it, which is why content added to a broken foundation returns nothing.

The ATLAS visibility framework

Five channels · 44 systems

Five letters, and every one of them is a list of things that can be measured. Select a channel.

How the site and its information are structured — the shape a crawler walks and a buyer navigates.

site architectureURL architecturecategory hierarchyinternal linkinginformation architecturecrawl pathscontent relationships

How accessible and machine-understandable the system is. The layer that fails silently.

crawlabilityindexationcanonicalisationrobotsXML sitemapsJavaScript SEOrenderingcore web vitalsstructured datamigrationsinternational infrastructure

How the brand connects to what people actually search, in every market it sells in.

keyword researchsearch intentsemantic relationshipstopic architecturelocalised languagecommercial intentcontent architecture

How the wider web corroborates the brand. A hundred spam domains and two real citations is a weaker position than the two alone.

backlinksmentionsdigital PRcitationsreferencesbrand signalsentity corroborationexternal trust

How the brand performs across every discovery system. Answer engines do not rank pages; they assemble an answer and cite what they trust.

GoogleAI searchanswer enginesgenerative searchAI OverviewsAI Modeecommerce searchmarketplace searchentity discoverycitation visibilitysearch behaviour

Services

Five capabilities

Technical SEO

crawlabilityindexationcanonical architecturerobotssitemapsJavaScript renderingcore web vitalsinternal linkingstructured datamigrationsShopify

The question it answersCan these systems read the site at all — and does the site say the same thing twice?

Ecommerce SEO

product architecturecollection architecturetaxonomyfaceted navigationvariant architectureproduct schemamerchant datacommercial intent

The question it answersDoes the catalogue's structure match how people actually shop it — and does the feed agree with the site?

International SEO

multilingual architecturehreflanglocalised keyword researchmarket intentcountry targetinglocalised entitiesduplicate-language management

The question it answersDoes each market get a page written for its own demand, or a translation of another market's?

AEO — answer engine optimisation

answer-first contententity understandingknowledge architecturestructured informationquestion intentfactual claritycitation potential

The question it answersWhen a system answers instead of linking, is the answer yours?

GEO — generative engine optimisation

entity relevanceexternal corroborationbrand mentionsdigital PRevidencecitation visibilityAI search discovery

The question it answersCan a generative system state a fact about this brand and defend it?

AEO and GEO are not AI ranking hacks. They are what search visibility becomes when the result is assembled rather than listed.

How the work runs

Two-week loops

DiagnoseUnderstand the system as it is, not as it was specified.
MapIdentify the visibility architecture, and where it breaks.
PrioritiseFind the highest-leverage constraint — the smallest change that moves the number.
BuildImplement it. Shipping in two-week loops, not quarterly plans.
MeasureTrack what actually moved, and say so when nothing did.
CompoundThen the next one.

What Atlas does not sell

Left: not on offer

ActivityImproved visibility

SEO as content productionSEO as infrastructure

Technical work separate from business strategyOne argument, connected end to end

AI search as magicAI search as another discovery system

Scaled noiseScaled systems

A dashboard to hide behindAn explanation of what actually matters

The record

What exists, and what does not

Atlas publishes what it finds, with the method attached. Where a figure is not evidenced it is not printed here, and the absence is stated instead of filled. That rule costs something on a page like this one — and it is the reason anything else on it can be trusted.

8 Sep 2026

Answer engines do not rank pages. They assemble answers.

What changes when the unit of retrieval is a sentence rather than a document.

Research 2 Sep 2026

A 396-domain backlink profile with two real citations

Why disavowing it would be the wrong move, and what to do instead.

Technical note 21 Aug 2026

Hreflang

The attribute that tells a search engine which language version belongs to whom, and the three ways it fails silently.

Glossary

Ambar — the audit that cancelled a rebuildWomenswear, four locales, 1,508 products. 158 findings, 44 refuted before the client saw them.Published

Client testimonialsAtlas does not publish them until they are given in writing, uneditedNone yet

Provenance — framework and audit weights defined by Atlas · no client measurements appear on this page

Build visibility

Seven questions · no discovery call

Atlas reads the site before the first conversation. These are the seven things worth knowing first — the same seven a diagnostic starts from.

01 Website02 Business03 Markets04 Current search situation05 Main challenge06 Desired outcome07 Contact

Start the diagnostic See how Atlas works

00 Overview Build visibility