SEO in 2026: ranking in Google during the AI era

Executive overview

AI overviews now cost the top-ranking page 58% of its clicks — up from 35% a year ago. Three shifts in search behavior define what works now: multi-platform discovery, brand mentions as the new ranking signal, and action-oriented queries where traditional SEO still dominates.

Build brand mentions on the right pages, then target queries AI can't fully satisfy.

Where AI pulls its answers from

• AI assistants research YouTube, Reddit, forums, and review sites — the same way humans do.
• 86% of top-cited sources are unique to each AI assistant.
• Google AI Overviews lean on YouTube, Reddit, and Quora; ChatGPT prefers publishers; Perplexity favours niche and regional sites.
• Use Ahrefs Brand Radar's cited domains report to see which sites AI cites in your niche.

Brand mentions as the core ranking signal

• AI fans a single query into dozens of subqueries, then stitches answers from the pages that rank for each — this is called query phantom.
• Branded mentions across those subquery pages have the strongest correlation with AI overview visibility — stronger than backlinks or domain rating.
• Target list roundups, comparison posts, and dedicated reviews; reach out to authors and get your product mentioned.
• Use the cited pages report in Brand Radar, filtering for queries containing "best", "top", "versus", "review", or "alternative".

Action-oriented queries where traditional SEO wins

• AI overviews rarely appear for queries where users need to complete an action (e.g. "backlink checker", "mortgage calculator", "snow removal service").
• Informational queries ("what is a mortgage?") are dominated by AI overviews; transactional and tool-based queries are not.
• Before targeting any keyword, ask: can AI fully satisfy the user for this query? If no, use traditional SEO.
• Find these opportunities in bulk by filtering keyword tools for modifiers like "calculator", "checker", or "generator", or by selecting transactional intent.