Generative systems repeat what the rest of the web already agrees on.
Ten points on the scale. Authority is not a link count — it is whether an independent source will corroborate a claim about you.
What actually breaks
6 defects · observed, not hypotheticalAsk a generative system about a brand and it assembles an answer from what it can find said about that brand elsewhere. Your own site is the least persuasive source in that set, because every brand says good things about itself.
So the work is corroboration, not promotion: making sure the basic, checkable facts about the business exist somewhere other than your own homepage, stated consistently, by sources that had no obligation to say them.
| Defect | What it looks like | What it costs |
|---|---|---|
| A backlink profile that is mostly spam | Hundreds of referring domains and a handful of real ones. | The headline number looks healthy and describes nothing. |
| No external corroboration of basic facts | Nothing outside your own site states what you do or where you do it. | A system has only your word for it, and will not use it. |
| Mentions unlinked to the entity | You are named, but not connected to any identifier you control. | The credit accrues to a name, not to you. |
| Digital PR measured in placements | Coverage counted by outlet and domain rating, not by what it corroborates. | A stack of links that answers no question a system would ask. |
| Inconsistent organisation data | Name, address, founding and description differ across profiles. | Every inconsistency is a reason to trust the record less. |
| A lapsed asset still being cited | An old domain or profile that outside sources still point at. | Authority delivered to an address you no longer answer. |
What it moves on the scale
Weights, not scoresMoves 30 of 100 points. The muted categories are not this service’s job, and saying so is the point: a proposal that claims to move all eight is describing a retainer, not a diagnosis.
What you get
Same format as every Atlas documentDelivered as a document, not a dashboard login: dark cover, verdict columns, evidence IDs, a provenance line stating where every number came from. Shipped in two-week loops — the smallest change that moves the number, then the next one.
Proof
What exists, and what does notNo fabricated clients, statistics, testimonials or rankings appear on this site. Where a figure is not evidenced it is not printed, and the absence is stated instead.
Common questions
Answer first, then the reasoningWhat is generative engine optimisation?
Building the external corroboration and entity clarity a generative system needs before it will state a fact about a brand. Where AEO is about what your pages say, GEO is about whether anyone else confirms it.
How is GEO different from AEO?
AEO is work on your own pages. GEO is work on the evidence outside them, which makes it the one search discipline you cannot complete on your own site — and the reason a technically flawless site can still be described wrongly.
Can a brand put itself into a knowledge graph?
No. Entry is a consequence of consistent, externally corroborated facts, not something a site can assert about itself. That makes tidying the places which already describe the business more productive than adding markup to your own pages.
Where this sits
The boundary, stated plainlyWhere it bites hardest
By sectorThree ways to start
None of them is a discovery callScan your own site
A visibility scorecard against the published 100-point scale, an hreflang validator, an answer-readiness reader and a SERP preview measured in pixels — all running in your browser. Nothing is uploaded and nothing is stored.
Run a scanThe AEO & GEO field guide
What changes when a result is assembled rather than listed — and the eight things to fix first. Written for engineers and founders, not for a newsletter funnel.
Read the guideBook a working call
Bring the site and the constraint you already suspect. You leave with the diagnosis whether or not there is an engagement in it.
Book on Cal.comAtlas reads the site before the first conversation. Seven questions; what comes back is a reading against the 100-point scale, the bottleneck named, and the smallest change that moves it.