"Should I still care about Google?" is now a real question from local business owners. Short answer: yes — but the work has changed shape, and most of the new work pays into both channels at once.
What's actually different
Google SEO is a ranking contest: ten links, and you fight for position with content depth, authority, and technical polish.
AI search (ChatGPT, Perplexity, Claude, AI Overviews) is a trust contest: one synthesized answer, and the model includes you only if it can verify who you are and what you offer. There is no position #4 — you're in the answer or you don't exist.
That changes what wins:
| Factor | Google SEO | AI search |
|---|---|---|
| Content volume | Matters a lot | Matters little |
| Backlink authority | Core signal | Secondary |
| Fact consistency | Helpful | Decisive |
| Structured data | Rich results | Core input |
| Plain-text answers | Featured snippets | The whole game |
The overlap is your budget's best friend
The highest-ROI moves pay into both channels simultaneously:
- Plain-language answers to buyer questions → featured snippets on Google, quotable facts for AI.
- LocalBusiness schema → rich results on Google, verified facts for AI.
- Consistent NAP across the web → local pack rankings on Google, confident answers from AI.
- A well-maintained Google Business Profile → Maps visibility, and a source AI Overviews leans on heavily.
The only AI-specific artifact with no Google payoff is llms.txt — and it takes ten minutes.
Where they genuinely diverge
Classic SEO tactics with weak AI payoff: chasing volume with thin blog posts, exact-match keyword pages, aggressive link building. If your only goal were AI visibility, you'd skip all three and spend the time on fact hygiene and answer-first content.
AI-specific work with no classic-SEO payoff: llms.txt, testing what models actually say about you, and monitoring answer drift as models retrain. This is the genuinely new maintenance burden — a new kind of uptime monitoring.
A sane split for a small business
If you have four hours a month for visibility work: spend two on the overlap (answer-first content, schema, consistency), one on your Google Business Profile and reviews, and one checking what AI actually says about you. That last hour is the one everyone skips — and it's the only way to catch the wrong answer that's quietly costing you calls. A free scan compresses it to 90 seconds; Starter makes it automatic.