Generative engine optimisation (GEO)
Definition
Generative engine optimisation (GEO) is the practice of editing content so that AI engines, which write an answer from several sources instead of listing links, are more likely to use and cite it. The term comes from a research paper accepted to KDD 2024.
The paper, by Aggarwal and colleagues including researchers at Princeton University, tested content edits on a benchmark of 10,000 queries. Adding citations, quotations and statistics lifted position-adjusted word count, a measure of how much of an answer cites a source and how prominently, by a relative 30 to 40%. Keyword stuffing didn't perform well, and lower-ranked sites gained more than higher-ranked ones.
The headline figure of up to 40% is the best result in a research benchmark from 2023 and 2024, and it measures visibility inside generated answers, not traffic, leads or revenue. A separate test on Perplexity gave smaller gains. Quoted as a forecast for one business, the number claims more than the paper does.
GEO overlaps heavily with AEO and AI SEO. For Google's own AI features, Google says nothing extra is needed beyond normal indexing and snippet eligibility, so the edits that did best in the paper are ordinary good editing: name the source and date of each statistic, and replace adjectives with figures. Measure it by asking the assistants the same set of buyer questions every month, in English and Arabic.
Definitions are free.So is the cost review.
A short call, then a written review inside 72 hours: what your marketing costs, what it produces, and what is worth fixing first.