The case for yes
Search still converts: the clicks that remain skew high intent, and someone wins them. And the second reason is newer and bigger: the pages good SEO produces, indexable, direct, current, genuinely authoritative, are exactly what AI systems cite when they compose answers. SEO stopped being just a channel and became the substrate of every answer surface at once. Kill it and you disappear from both.
The case for not-like-before
What is no longer worth buying: rankings as the end product. Position reports on informational queries mean less every quarter as AI Overviews absorb those clicks, the mechanics in AI Overviews vs Google Search. Commodity content programs mean even less, because engines and assistants both ignore restated consensus.
The reframe: pay for outcomes on both surfaces. Clicks and pipeline from the queries that still click, plus citations and recommendations in the answers that do not, measured per assistant, the modern program is laid out in how to do SEO in 2026.
The math to run for your own case
Take your organic pipeline value, then test your ten most valuable buying questions across ChatGPT, Gemini, Claude and Perplexity. If you appear in most answers, your SEO is compounding on both surfaces and is absolutely worth continuing. If competitors own those answers while you hold rankings, part of your SEO spend is defending a surface your buyers are leaving, and the highest-return next dollar goes to the answer layer. Run that test free: the AI visibility audit maps it in 2 business days, and is my SEO good gives the ten minute self-assessment.
