Featured Snippets vs AI Overviews: Which Drives More Traffic in 2026?

TLDR: Featured snippets are being absorbed into AI Overviews. Ahrefs data from July 2025 shows AI Overviews now appear on 48% of queries while traditional snippet appearances dropped proportionally. Optimization tactics have diverged: snippets reward concise direct answers in a fixed format, while AI Overviews reward broad entity coverage and citation worthiness. If you optimise for only one in 2026, you are leaving traffic on the table.

The state of SERP features in 2026

AI Overviews exploded from a beta feature into a default surface for nearly half of all Google queries. According to Ahrefs SERP feature research, AI Overviews appeared on 48% of tracked queries by mid-2025, while featured snippet share declined in roughly inverse proportion.

A parallel Analyze.ai breakdown confirmed the cannibalisation: when an AI Overview is present, the classic featured snippet is suppressed roughly 80% of the time.

Traffic dynamics: which actually delivers clicks

Featured snippets historically captured 8 to 12% of click share for informational queries. AI Overviews are different: they answer the question inline, and click-through rates on cited sources sit closer to 1 to 4% depending on industry.

• Featured snippets: Higher CTR per appearance, but appearing on fewer queries.
• AI Overviews: Lower CTR per citation, but appearing on 48% of queries and growing.
• Net traffic: AI Overviews now drive more total cited-source traffic in absolute volume despite the lower per-impression CTR.

The shift is from concentrated snippet wins to distributed citation share across many queries. Brands that only chased snippets are losing visibility faster than they realise.

Optimization tactics that diverge

For featured snippets (still relevant)

1. Lead with a 40 to 60 word direct answer immediately under the H2 that matches the query.
2. Use the exact query phrasing in the H2 and the first sentence.
3. Format as a definition paragraph, ordered list, or comparison table depending on intent.
4. Keep the answer self-contained so Google can lift it without surrounding context.

For AI Overviews (the bigger prize)

1. Cover the full entity space around the query, not just the literal question.
2. Cite primary sources with inline anchor links so AI engines can verify your claims.
3. Include statistics, named frameworks, and dated facts that LLMs prefer to attribute.
4. Maintain consistent author attribution and freshness signals at the page level.

How to optimise for both without doubling your work

The good news: a well-structured page can win both. Start every key section with a snippet-ready 50-word answer, then expand into the deep entity coverage that AI Overviews reward. The lead paragraph serves Google traditional, the body serves AI Overviews and ChatGPT.

Track both surfaces side by side using the GEO/AEO Tracker. Most teams discover that pages losing snippet share are gaining AI Overview citations, which means total branded visibility is often flat or up even when snippet dashboards look red.