Where the term came from
GEO entered general use in late 2023, popularised initially by a Princeton-led paper proposing a set of optimisation methods for content destined for generative search. The acronym caught on quickly because it filled a vacuum: practitioners and journalists needed a term for the new work, and SEO's analogy was familiar.
Since then, several adjacent terms have circulated. AEO (Answer Engine Optimisation) describes the same work seen from the answer side. LLMO (Large Language Model Optimisation) appears occasionally in technical writing. AI SEO is used loosely. The institution treats GEO as the most established of these terms but does not regard the terminology as settled.
What the discipline actually is
Stripped of its terminology, the discipline does five things.
- Reconciles the subject's identity inside the citation graphs AI systems consult.
- Closes the distance between the subject and the authority sources those systems treat as load-bearing.
- Produces and surfaces content shapes the systems can quote and attribute.
- Builds the credentialing and editorial signals the systems use as confidence inputs.
- Measures the resulting standing across systems, longitudinally, against a defined methodology.
These five functions describe the bulk of the institution's work. Whether the discipline is called GEO, AEO, AI Discovery, or some future term, the work is approximately stable.
The terminology question matters less than the framing question. Teams that approach the work as optimisation tend to underestimate its editorial and institutional dimensions. Teams that approach it as discovery tend to make better decisions on time horizons and on quality thresholds.
Why naming matters
Naming a discipline shapes how it is staffed, scoped, and bought. SEO has historically been bought as a marketing line item, scoped quarterly, and staffed by specialists. GEO, when bought through the SEO frame, inherits the same scoping habits. The institution's experience is that this is the largest single source of disappointed engagements in the field.
The reframing the institution proposes is to call the work AI Discovery. Discovery names the actual phenomenon: the AI's act of finding the subject and treating it as authoritative. Optimisation names a tactic. The first is what is happening; the second is part of how it is occasionally done.
How the institution practises it
The Standard, the Index, and the Body of Knowledge describe how the institution practises the discipline in detail. In summary: the work is conducted as professional services, methodology-led, on twelve-month engagements with quarterly Score reviews, against a published methodology that is itself amendable through a defined process.
A closing observation
Whether the field settles on GEO, AEO, or another term, the underlying discipline is now established. The job of the institution is to practise it well, regardless of what it is called this season.
Naming will continue to evolve. The work, in its essentials, has stabilised. The institution suggests — with the appropriate restraint — that AI Discovery describes the work most accurately, and is willing to be told otherwise in a future edition.