AI search answers more questions without sending a click. Google's AI Overviews, AI Mode, ChatGPT, Perplexity and other assistants summarize information directly, so some visits that used to reach websites never happen. But the effect isn't uniform. Simple informational queries lose the most; complex, commercial and high-trust queries still send people to sites. And being named in an answer has value of its own. The right response is to change what you create and how you measure it, not to panic.
Zero-click isn't new, but AI accelerates it
Featured snippets, knowledge panels, weather boxes and calculators were already answering queries on the results page. AI answers extend this to far more queries, including ones that used to need a page visit: multi-part questions, comparisons and summaries.
Several independent analyses have reported lower click-through rates for results that show AI Overviews, and publishers in informational niches have reported traffic declines. Google, for its part, says that people who click from AI features tend to be more engaged. Both are consistent with the same picture: fewer, but more intentional, clicks.
Which traffic is most at risk
| Query type | Exposure | Why |
|---|---|---|
| Definitions ("what is a 401k") | High | Fully answerable in a few sentences |
| Simple facts ("how many ounces in a cup") | High | One-line answer |
| Quick how-tos ("how to reset a router") | High to medium | Steps can be summarized |
| General overviews ("benefits of yoga") | Medium | Summaries satisfy casual readers |
| Comparisons ("X vs Y") | Medium | Answers summarize, but buyers often click to verify |
| Commercial research ("best CRM for agencies") | Lower | Decisions need depth, trust and specifics |
| Brand and navigational ("Brand login") | Low | Users want the site itself |
| Tools, calculators, data, templates | Low | Users need to interact or download |
| Experience-based content (reviews, tests, case studies) | Lower | Hard to compress without losing the value |
If most of your traffic comes from the top rows, expect pressure. If it's mostly from the bottom rows, you may see little change, or even gains from being cited.
The value of a mention without a click
For brands, the more important question is often not "did we get the click?" but "were we in the answer?"
When an assistant names your brand in a shortlist or cites your page:
- You're considered. For recommendation questions, being named is the gateway to the buyer's shortlist.
- You're framed. The assistant's description shapes how buyers think about you before they visit.
- You build demand. Buyers who see your name often search for you later, which shows up as branded or direct traffic.
That's why AI visibility metrics such as mention rate and share of voice matter even when referral traffic is small.
How to adapt your content
Lean into what can't be summarized away
- Original data and research. Assistants can cite your findings, but readers who want the full data still click.
- Tools and interactive content. Calculators, templates, configurators and checkers.
- Depth and experience. First-hand testing, detailed case studies, expert analysis.
- Community and opinion. Perspectives and debate that a summary can't replace.
Make informational content work for visibility
Informational pages still matter. They establish topical authority and get cited in AI answers, which puts your brand in front of people. Make them count:
- Answer directly and specifically, so you're the source that gets cited. See how to write content AI assistants cite.
- Add a clear next step for readers who want more: a tool, a deeper guide, a product page.
- Include your brand's perspective, so a citation also communicates who you are.
Strengthen commercial pages
Commercial pages hold up better, but only if they're clear: use-case pages, comparison pages, pricing pages and documentation. These are what assistants retrieve for buying questions and what buyers click to verify.
How to adapt your measurement
If you only watch sessions, AI search will look like a loss even when it's helping. Add:
- AI visibility: mention rate, share of voice and citations across a fixed prompt set.
- AI referral traffic: sessions and conversions from chatgpt.com, perplexity.ai, gemini.google.com, claude.ai and copilot.microsoft.com.
- Brand demand: branded search impressions in Search Console and direct traffic trends.
- Self-reported attribution: an "AI assistant" option in "How did you hear about us?"
- Conversion rate by channel: AI referrals often convert differently from generic organic traffic.
Report these together so stakeholders see the full picture: fewer visits on some queries, more visibility and better-qualified visits overall.
A realistic plan
- Audit your traffic by query type. Estimate how much comes from the high-exposure rows.
- Baseline your AI visibility for the commercial questions that matter most.
- Rebalance content investment toward original, interactive, experience-based and commercial content.
- Upgrade top informational pages to be the best-cited source and link to commercial next steps.
- Change the scorecard to include visibility, brand demand and conversions.
The bottom line
Zero-click answers are real, and some informational traffic won't come back. But the brands that do well in AI search treat the answer itself as a channel: they aim to be named and cited, they create content worth clicking for, and they measure influence as well as visits. Traffic is one outcome of search, not the only one.
For how this plays out on Google specifically, see our Google AI Overviews guide.
For the latest numbers on AI usage and clicks, see our AI search statistics.
