Some classic ranking factors carry over but with shifted emphasis. Keyword targeting matters less in absolute terms — semantic matching is good enough that exact-match phrasing is no longer required. But H1 specificity matters more, because the H1 is the single strongest passage-level signal an LLM uses to disambiguate which question you answer.
Backlinks still matter, but anchor diversity matters more than anchor exactness. Mention authority — being cited in news, podcasts, and academic work — now carries comparable weight to a traditional editorial backlink for AI citation.
For most growth-stage businesses we work with, the right initial split is 70% SEO foundation, 30% GEO layer. The SEO foundation pays the rent — it drives the bulk of measured traffic and conversions. The GEO layer is the strategic investment that compounds over the next 24 months.
If you are starting from a healthy SEO baseline (good rankings, clean technical), shift the split closer to 50/50. The marginal dollar in GEO has higher returns right now because the competition is light.
Not within any reasonable planning horizon. The SERP is not going away — Google continues to iterate on it, and a meaningful share of users will always prefer to compare results themselves. GEO is additive.
For most teams, no. The skills overlap heavily with content-led SEO. The right move is to upskill your existing senior SEO into the GEO layer — they already understand the foundations.
Roughly yes. AEO (Answer Engine Optimization) is the older term, mostly used pre-ChatGPT. GEO is the newer term used in academic literature and by most agencies in 2026.
Almost never. The exception: if you over-edit your H1s purely for AI question-matching, you can lose long-tail SEO traffic that came from the old phrasing. Always check Search Console before/after a major H1 rewrite.