What Changed in 2026 SERPs

If you are still optimizing for blue links, you are already behind. In early 2024, AI Overviews were a novelty that showed up for maybe 15 percent of searches. By March 2026, Google confirmed they appear for over 78 percent of informational queries in the US, UK, and Canada, and they also launched AI Mode, which is a full chat style answer page.

I saw the shift firsthand on my camping gear blog. In December 2025, my how-to posts were getting a 2.8 percent CTR from position 2. By February 2026, same position, CTR dropped to 1.1 percent. Why? Because the AI Overview answered the question at the top and pushed my result below the fold on mobile.

The biggest change is not just the answer box. It is the citations. Google now shows three to five source links inside the AI Overview, usually with a small favicon and the site name. Those links get the clicks, not position 1 below. My testing across 112 keywords shows those citation links get an average CTR of 9.2 percent, even without being ranked first organically.

AI Mode is even more aggressive. It gives a full conversation and only cites sources when users click show sources. That means traditional impressions are down 30 to 50 percent for basic questions. You cannot fight this. You have to optimize to be the source it cites.

AEO vs Traditional SEO

I learned AEO is not a replacement for SEO. It is a layer on top. Traditional SEO helps you rank. AEO helps you get chosen. Here is how I now plan content differently.

The 5 Content Formats AI Overviews Love

After analyzing 112 AI Overview citations across my three sites, five formats appeared over and over. Google loves content it can easily extract without guessing.

1. Direct Definitions

I now start every informational post with a 25 to 35 word definition. For example: Answer Engine Optimization is the process of structuring content so AI systems can find, understand, and cite your answer directly. No fluff before it. My post that added this exact box got cited within 11 days, while the old version with a story intro never was.

2. Step by Step

AI Overviews love numbered steps. Google can turn them into a neat list instantly. I changed my how to clean a water filter post from paragraphs to H3 steps with action verbs. Step 1: Remove the cartridge. Step 2: Backflush for 30 seconds. It was picked up two weeks later. Keep each step under 50 words.

3. Pros and Cons Tables

For any best or versus query, add a simple pros and cons table in HTML, not an image. I use three columns: Feature, Pros, Cons. My backpack review had this table and Google pulled it verbatim into the overview. Use real experience, not generic marketing points.

4. Comparison Tables

If you write x vs y, create a comparison table with clear rows: Price, Weight, Battery life, Warranty. Use checkmarks or short text. On my power bank comparison, the AI Overview cited my table for three different queries. It is the easiest win I found in my tests.

5. FAQ Blocks

The last format is a true FAQ section with questions as H3s. I match the exact People Also Ask phrasing. Then I answer in 40 to 60 words right below. Do not hide it only in an accordion for the main version. Use visible text first, then you can add details tags for UX later.

My 7-Step AEO Framework

Here is the exact workflow I now follow for every new post. It takes an extra 45 minutes, but my citation rate tripled from 8 percent to 27 percent.

1. Understand the intent in AI Mode. Before writing, I search the keyword in Google with AI Overviews on. I screenshot what format it shows. Is it a definition, steps, or table? I then design my content to match that format exactly.
2. Answer first. Put the direct answer in the first 40 to 70 words after the H1. No intro story, no affiliate disclosure at the top. I move my personal story to after the answer. This first paragraph is what Google pulls 80 percent of the time.
3. Use structured headings. I use question style H2s and H3s that mirror real searches. Instead of My Thoughts, I write How long does it last. Each heading gets one clear idea. This helps the AI parse sections cleanly.
4. Add first-hand proof. This is the E-E-A-T part. I add original photos with my hand in them, screenshots of my test data, or a short video. Then I write one sentence like I tested this filter for 14 days in Yosemite. Posts with that sentence got cited 3.1x more often.
5. Cite sources. Ironically, to get cited, you need to cite. I link to studies, manufacturer specs, or Reddit threads for context, and I list the source name in text. I add two to three outbound links per 1,000 words.
6. Add schema. I add FAQPage and HowTo schema for every eligible post. Blogger makes this easy in the HTML view. I also keep Article schema with author name and date updated. After adding schema, my average time to first citation dropped from 31 days to 18 days.
7. Update frequently. AI Overviews favor freshness. I review top posts every 60 days. I change the date, add one new paragraph with updated testing, and resubmit in Search Console. My older posts regained citations within a week after updating.

E-E-A-T Signals That Matter in 2026

Google says E-E-A-T is more important for AI, and my data confirms it.

Experience photos: I now include at least three original photos per review, one unboxing, one in use, one flaw close up. I name the files descriptively and add alt text. Posts with original images were cited 67 percent of the time versus 22 percent for stock photos.

Author bio: I created a real author page on each blog with a photo, credentials, and links to LinkedIn. I link my name in every post to that page. This alone increased my citation rate by 14 percent.

Original data: I publish small tests. For example, I timed how long 5 power banks took to charge an iPhone 15. That table was cited five times across related queries. AI prefers unique numbers over opinions.

Transparency: I now add a Tested by line and an Updates log at the bottom. I disclose affiliate links clearly but after the main answer. Google seems to reward honesty over perfect keyword density.

Mini Case Study: Before and After

Let me show you real numbers from my smallest blog, a camping site with 12,000 monthly visits. I took a post ranking position 4 for how to keep tent warm in winter. It had zero AI Overview appearances in January.

In February, I rewrote it using my AEO framework. I added a 32 word definition first, changed steps to H3s, added a pros and cons table for heaters, added my own photos from a January trip, and added HowTo schema.

Impressions dropped, but clicks went up 39 percent because I was getting the citation traffic from the AI Overview itself. This is the new game.

Common Mistakes That Kill AEO

These are the five errors I see most, and I made them all at first.

• Burying the answer. I loved personal intros, but AI skips them. Give the answer first, story second.
• Using images for tables. Google cannot read text in PNGs. Always use real HTML tables.
• Vague headings. If your H2 is Final Verdict, AI ignores it. Use What is the best budget option.
• No updates. Content from 2023 is dead in AI results. I saw citations drop off after 90 days without a refresh.
• Pure AI content without proof. I tested 10 pure AI posts. Zero citations. Same topics with my photos and one paragraph of real testing got four citations. Experience wins.

FAQ

Conclusion

Ranking in Google's AI Overviews is not about tricking the AI. It is about making your content stupidly easy to understand, extract, and trust. Answer first, structure with questions, prove you actually used the product, and keep it fresh.

I went from zero AI citations in December to 183 across three sites by April just by following this framework. Start with your top 10 posts this week.