By Mehdi HassanIs SEO dead because of AI search?
No. SEO isn't dead, but the deal it offered has changed. For most of search's history, ranking meant traffic. In 2026, ranking still means being seen, but more of that visibility happens inside Google's results page and AI answers, and less of it turns into a visit. That's a real loss for some businesses and a smaller change for others.
We think the "SEO is dead" headline gets the cause right and the conclusion wrong. AI answers are built from web pages. The companies that disappear won't be the ones that keep doing SEO; they'll be the ones whose sites can't be found, read or trusted. This piece is part of our answer engine optimization guide.
What does the zero-click data actually show?
The data shows a real and growing share of searches that end without a click, but the headline numbers need careful reading. Two studies get quoted most often.
- Of US Google searches ended with no click, Jan to Apr 2026 (SparkToro)
- 68%
- Lower CTR for the top result when an AI Overview shows (Ahrefs, Dec 2025 data)
- 58%
- The same CTR drop in Ahrefs' 2025 analysis
- 34.5%
SparkToro found 68% of US Google searches in January to April 2026 ended without a click to any website. That's striking, but SparkToro switched its panel data provider that year, so comparing it to earlier years isn't a clean like-for-like trend. Ahrefs' study is tighter: for queries with an AI Overview, the top organic result's click-through rate was 58% lower, up from a 34.5% drop in its 2025 analysis.
Why is SEO still worth doing in 2026?
SEO is still worth doing because AI answers are assembled from pages search systems can find. Google's AI Overviews draw on Google's index, and Google's own documentation says its AI features need nothing beyond normal SEO best practice. ChatGPT search and Perplexity run their own crawlers. A page that isn't crawlable, indexed and clearly written isn't eligible for any of them.
What changes is the target. Ranking for a broad term was the goal; being the quoted source for specific questions is now part of it. Conductor reports 57% of AI Overview citations come from outside the top 10, which means a page that answers a narrow question well can be cited without ranking first. That's an opening for small companies, not a dead end.
What's weaker | What still works | What's newly important |
|---|---|---|
Traffic from broad informational queries | Ranking for specific, high-intent queries | Being cited inside AI answers |
Generic how-to blog posts | Pages that answer buyer questions precisely | Brand mentions across the web |
Measuring success by sessions alone | Conversions from search visitors | Prompt audits and self-reported attribution |
Volume-first content calendars | Technical health and crawlability | Answer-first structure on money pages |
Why does the website matter more, not less?
The website matters more because it's now the place buyers check after an AI tool has shaped their view. Someone who asks ChatGPT for options, reads an AI Overview and then visits your site arrives better informed and closer to a decision. There are fewer of those visits, so each one counts for more.
TripleDart's Q1 2026 study of 28 B2B SaaS companies shows where the value went: homepages earned 81.7% of AI-attributed leads, while general blog posts drew 12,000 sessions and no leads. The page that converts is the page that says clearly what you do, for whom, at what price and why you're credible. If that page is vague, the AI-informed buyer leaves for the competitor whose page isn't.
When the click gets rarer, the page it lands on has to work harder.
For example, suppose two startups lose half their search traffic
For example, suppose two B2B startups each see blog traffic fall by half after AI Overviews expand in their category.
The first measures sessions, sees the drop, and cuts content. The second notices that demo requests held steady, because its pricing and comparison pages still answer what buyers ask and are cited in AI answers. It shifts effort from generic posts to money pages, adds a "How did you hear about us?" field and finds AI tools named often. Same traffic loss; very different conclusions. This is a hypothetical scenario, not data.
What should startups do differently now?
Startups should spend less on volume and more on the pages and signals that survive zero-click search. In practice, that means five shifts.
- Build money pages before blog posts: homepage, pricing, product, comparisons, integrations and security.
- Write answer-first, so each page and section can be quoted on its own.
- Make the site readable by every crawler, with server-rendered content; see building an AI-search-ready website.
- Earn mentions. In Ahrefs' study of 75,000 brands, web mentions correlated 0.664 with AI Overview visibility, against 0.218 for backlinks.
- Measure differently: AI referral traffic, a source field on forms and a monthly prompt audit.
Our AEO checklist turns this into a list of fixes, and zero-click search strategy covers winning buyers who never visit. If you're tempted to replace writers with AI to cut costs, read AI-written content and SEO first: generic content is the kind most exposed to zero-click.
Who should worry, and who shouldn't?
Sites that earned most of their traffic from broad informational queries and monetised it with ads or affiliate links should worry most, because AI answers do exactly that job. B2B companies whose value is a product people evaluate, compare and buy have less to fear, provided their sites answer buyer questions clearly.
If your pipeline depends on search, don't panic and don't stop. Audit which queries actually brought customers, check whether your key pages are readable and quotable, and rebalance from volume to quality. Named expertise helps here too, as we explain in author pages for SEO.
