Search behavior has changed for good. Out of every 1,000 Google searches in the United States, only 360 clicks reach external websites. The other 640 searches end without a click, send traffic to Google properties, or lead to another search. [1]
The kind of device makes a big difference. On mobile, 77 percent of Google searches end without a single click to an outside website. On desktop, the zero-click rate is 46.5 percent. When Google shows an AI Overview, the zero-click rate across all devices jumps to 83 percent. In Google AI Mode, it hits 93 percent. [2]
Those numbers point to a structural change in how the web works. For thirty years, the deal was simple. Developers made content, search engines indexed it, and people clicked links to visit websites. That deal is over. Search engines are now the destination. They pull your data, summarize your facts, and answer the question right there on the results page.
That changes web design going forward. A page can’t count on its visual layout alone to hold someone’s attention once they arrive. Developers have to build for two kinds of visitors: people who come for a deeper look, and AI crawlers that read the content in an instant for summary engines. The same goes for civic content, including my post on software vendor vetting for local government boards and my other writing on public-side deal structuring. If people can’t find it, the substance doesn’t help anyone.
If your site can’t talk plainly to AI scrapers, your message disappears.
1. Implement Deep JSON-LD Schema Markup
AI bots don’t guess how your content is structured. They read structured data. You need exact JSON-LD schema across your site. [3], [4]
- Add
TechArticleorArticleschema to every post, with publication dates, author metadata, and publisher information. - Use
FAQPageschema for question-and-answer sections. With plain paragraphs, the AI has to pull facts out blind. Schema hands the facts straight to the crawler. - Add
ItemListmarkup for roundups and lists, so the parser credits each list item to your domain.
2. Place Direct Answers in the Top 100 Words
AI models give more weight to text near the top of the page. Long introductions waste crawl budget and lower your odds of getting cited.
- State the main fact or answer in the first two sentences.
- Use a plain question as your H2, then answer it in the very next paragraph.
- Keep answer sentences under 20 words. Short statements drop cleanly into AI summaries.
3. Use HTML Tables and Unordered Lists for Raw Data
AI summaries pull data from tables well. Plain blocks of text often get misquoted or skipped.
- Put pricing, specs, and comparison data in HTML tables with labeled
thheaders. - Use
ulandollists for step-by-step technical instructions. - Skip custom JavaScript table renders that block static crawlers.
4. Increase Web Server Speed for AI Bots
AI scrapers visit millions of pages per minute. If your server takes longer than 300 milliseconds to respond, crawlers skip your deeper content.
- Serve static HTML through a CDN, and keep dynamic server rendering on content pages to a minimum.
- Remove blocking JavaScript that delays DOM completion.
- Check your
robots.txtfile to make sure you aren’t blocking the AI user agents modern search engines use.
5. Add Verifiable Source Attribution
AI models favor content backed by data. Claims nobody can check get thrown out when the summary is written.
- Include exact statistics and cite primary sources right in your text.
- Link out to reputable primary sources. AI engines check outbound links to judge how trustworthy a page is.
- Update your content timestamps any time you update the numbers.
The zero-click search numbers, device splits, and schema standards in this post come from the following industry studies and documentation:
References
- Fishkin, Rand. (2024). 2024 Zero-Click Search Study: For Every 1,000 Searches, 360 Clicks Leave Google. SparkToro. https://sparktoro.com/blog/2024-zero-click-search-study/
- Bain, Mark. (2025). Measuring the Impact of AI Overviews and Google AI Mode on organic CTR. SEO Clarity. https://www.seoclarity.net/
- Schema.org Community Group. (2026). Schema.org Vocabulary: TechArticle, Article, and FAQPage Guidelines. W3C Schema.org Community Group. https://schema.org/TechArticle
- Google Search Central. (2025). Google-Extended and Googlebot Documentation: Structuring Data for AI Features. Google LLC. https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data






