SEO in 2026 looks different on the results page. AI answers now sit above the links for many searches, and fewer of those searches end in a click. But the work that earns rankings has barely moved: match the search, keep your title’s promise, keep the page fast, and earn links from relevant sites.
I have worked in SEO since 2014, and every few months someone tells me it is dead. Then I open a client’s Search Console and see pages that answer the search well still bringing in traffic. What has changed is where that traffic starts, and how often people need to click at all.
What changed: AI answers now sit above the links
At Google I/O on May 19, 2026, Google CEO Sundar Pichai said AI Overviews has over 2.5 billion monthly active users. He added that AI Mode passed 1 billion monthly active users about a year after launch. Ahrefs, the SEO data company, notes that AI Overviews have run on Google’s Gemini 3 model since January 2026.
What does the click data say?
The click studies don’t all agree, so read them closely. Ahrefs looked at 300,000 keywords using Search Console data from December 2023 and December 2025. Where an AI Overview appeared, the top-ranking page had a 58% lower average desktop click-through rate. That is a pattern across a large sample, not a forecast for your site.
Seer Interactive’s data, reported by Search Engine Land in April 2026, fills in the detail. CTR on queries with an AI Overview dropped to 1.3% in December 2025, then climbed to 2.4% by February 2026. Pages cited inside the overview averaged about 2.1% CTR. Uncited pages on the same results page averaged about 0.9%. Both sat below the roughly 3.3% seen on searches with no overview.
Seer also found AI Overviews on about 36% of informational queries, but only about 5% of transactional ones. Blog posts and guides feel this change the most, while service and product pages see far fewer overviews. And when one does appear, being cited beats being skipped.
Where do AI answers find their sources?
AI Overviews no longer lean mainly on page one. In a March 2026 study, Ahrefs looked at 863,000 keyword results and 4 million cited URLs. It found that 37.9% of AI Overview citations also ranked in the top 10 for the same query. Its July 2025 study put that figure near 76%. Ahrefs suggests the likely cause is query fan-out, and notes that its own citation tracking also improved.
Google’s documentation explains that AI Overviews and AI Mode may run several related searches across subtopics to build one answer. So a page can be picked for a side question even when it does not rank for the main one.
ChatGPT and Perplexity use their own crawlers. OpenAI says OAI-SearchBot surfaces websites in ChatGPT’s search features, and sites that block it will not show up in ChatGPT search answers. Perplexity says PerplexityBot surfaces and links websites in its search results. I check robots.txt and firewall rules early on every new site. A security plugin or an old rule can block these crawlers, and nobody notices.
What did not change: relevant links still count
Google is direct about this. Its ranking systems guide says Google uses systems that look at how pages link to each other to understand what those pages are about. It also says PageRank is still part of its core ranking systems.
The type of link that helps has changed, though. Google’s spam policies treat bought links, excessive link swaps, automated link building, and guest posts or press releases with optimized anchor text as link spam. A good test for any link: would it exist if rankings did not matter?
That is why relevance matters more than volume. Links help Google understand a page’s topic. So a link from a site in your own field sends a clearer signal than one from an unrelated site. A page your buyers actually read can also send real visitors. A directory listing rarely does.
Before I go after any link, I check five things:
- Does the site rank for keywords in my client’s topic?
- Does it get real organic traffic, mostly from the countries my client sells to?
- Is the linking page itself indexed? A site: search for the exact URL answers this in seconds.
- Would a reader of that page have a reason to click through?
- Is the anchor natural? I lean toward branded or descriptive anchors over exact-match keywords.
If a link fails most of these, I skip it, however cheap it is.
On-page SEO basics that still move rankings
When I take on a new client, I review the target page before any link work. If the page does not fit the search, links will not fix it. These are the checks I start with.
Match the search intent
Search the keyword and look at what ranks. If the top results are service pages with prices, a long blog post will struggle, and the reverse is just as true. Get the page type right first, then work on the words.
Make the title and H1 keep their promise
Google’s helpful content guidance asks whether the main heading or page title gives a descriptive, helpful summary of the content. A title that says “pricing” on a page with no prices sends people straight back to the results.
Use internal links on purpose
Google lists making content easy to find through internal links as a best practice for its AI features. Link to the page you want to rank from your strongest related pages, using anchor text that tells readers what they will get.
Keep the page fast
The same documentation recommends a great page experience. Compress large images, remove scripts you no longer use, and test on a real phone over mobile data, not just office Wi-Fi. The Core Web Vitals report in Search Console shows which pages need work.
Answer near the top of each section
Give the direct answer in the first one or two sentences under each heading, then explain. People skimming get what they came for, and the answer is easy to quote.
How to get a page cited by AI search
Google says there are no extra requirements or special optimizations needed to appear in AI Overviews or AI Mode. The page needs to be indexed and eligible to show a snippet. In other words, getting cited is mostly good SEO, done with care. A few habits help:
- Write headings as real questions. A fan-out search often matches one section, not the whole page, so answer each heading in its first sentence.
- Cover the questions around the main one. Most AI Overview citations now come from outside the top 10 for the main query. A page that also answers the related questions has more chances to be picked.
- Add something only you have. Google asks whether content offers original information or analysis, and whether it shows first-hand expertise. That is the experience side of E-E-A-T. Your own process, tested numbers or a real client example give an AI system a reason to cite you. A page that only repeats others does not.
- Use schema where it fits, and keep it honest. Google says no special schema is needed for AI features, and recommends that structured data match the visible text. Article, Product, Organization or LocalBusiness markup describes a page. It does not replace good content.
- Keep key content in text. Google recommends making important content available as text. Prices locked inside an image are hard for any system to read.
- Think about video for key topics. Ahrefs found that 18.2% of AI Overview citations from pages outside the top 100 were YouTube URLs. A short video that answers the same question can be a second route to a citation.
A checklist you can use today
- Search your main keyword and confirm your page type matches what ranks.
- Read your title and H1, then check the page delivers exactly that.
- Put a direct answer in the first two sentences under every heading.
- Add internal links to the page from your three strongest related pages.
- Test the page on a phone over mobile data and fix whatever loads slowly.
- Check robots.txt and firewall rules for Googlebot, OAI-SearchBot and PerplexityBot.
- Confirm any schema matches what a visitor can see on the page.
- Add one piece of first-hand experience or original data.
- Review your backlinks and stop chasing links that fail the five checks above.
The bottom line
AI search has changed how answers look, and on many informational searches, click-through rates are lower when an AI answer appears. It has not changed what makes a page worth showing. Give searchers exactly what they came for, show that you have done the work, and earn links from sites your audience trusts. Then judge the results over months, not days.
3. Sources
- Google I/O 2026: Sundar Pichai’s opening keynote (Google, May 19, 2026) – https://blog.google/innovation-and-ai/sundar-pichai-io-2026/
- AI features and your website (Google Search Central) – https://developers.google.com/search/docs/appearance/ai-features
- A guide to Google Search ranking systems (Google Search Central) – https://developers.google.com/search/docs/appearance/ranking-systems-guide
- Spam policies for Google web search (Google Search Central) – https://developers.google.com/search/docs/essentials/spam-policies
- Creating helpful, reliable, people-first content (Google Search Central) – https://developers.google.com/search/docs/fundamentals/creating-helpful-content
- Update: AI Overviews Reduce Clicks by 58% (Ahrefs) – https://ahrefs.com/blog/ai-overviews-reduce-clicks-update/
- Update: 38% of AI Overview Citations Pull From The Top 10 (Ahrefs, March 2, 2026) – https://ahrefs.com/blog/ai-overview-citations-top-10/
- Google AI Overviews CTR shows early signs of recovery: Study (Search Engine Land, Seer Interactive data) – https://searchengineland.com/google-ai-overviews-ctr-recovery-study-475566
- OpenAI crawlers: OAI-SearchBot (OpenAI documentation) – https://developers.openai.com/api/docs/bots
- Perplexity crawlers: PerplexityBot (Perplexity documentation) – https://docs.perplexity.ai/guides/bots
4. Author bio
Subham Chowdhury is the founder of SEO by Subham, an independent SEO and link-building consultant based in Kolkata, working with SEO since 2014. He focuses on on-page SEO, white-hat link building and visibility in AI search.