GEO applies to documentation because answer engines increasingly sit between users and docs. When a developer asks ChatGPT how to configure your product, the answer comes from whatever content the engine could crawl, parse, and quote confidently.
For documentation, practical work includes clear answers, self-contained sections, stable URLs, and checking which crawlers can access the content. Training crawlers and search or retrieval crawlers have different purposes. An llms.txt file can provide a curated index, but publishing it does not guarantee citations.
GEO complements SEO rather than replacing it. The qualities that make a passage citable by an answer engine, such as clarity and one idea per section, also make it a good search snippet and a good retrieval chunk for your own docs chatbot.