AI Overviews: Organic CTR Drops 15% in 2026

Let’s be clear: AI Overviews blew up the organic search playbook. For any business that depends on organic traffic, this is the new reality. Marketers are scrambling to figure out how to even measure the damage, let alone adapt their strategies to stay visible. This teardown walks through a real campaign we ran to track and fight the effects of AI Overviews for a client, and it shows a huge shift in how users are behaving.

Campaign Overview: Adapting to AI Overviews

An e-commerce client specializing in outdoor gear came to us in late 2025 because their traffic on informational queries was falling off a cliff. Their whole model relies on showing up for long-tail keywords when people are researching product features or reading “how-to” guides before they buy. The full rollout of AI Overviews (which was called SGE in testing) started hammering their top-of-funnel discovery, showing up as a big drop in organic clicks even though their impression counts were holding steady.

So, we built a campaign specifically to figure out what was happening and push back. It ran for three months, from October to December 2025. We had a $12,000 budget just for content analysis, restructuring the pages, and implementing schema. Our main goal was to stop the bleeding on organic traffic for the affected keyword clusters and, if possible, actually get our client featured *inside* the AI Overviews themselves. We zeroed in on 200 informational keywords that were constantly triggering AI Overviews, we found these by digging through Google Search Console data that showed high impressions but a cratering CTR. Our KPIs were organic CTR, conversions from those pages, cost per conversion, and our share of voice within the AI results.

Strategy: Restructuring Content and Beefing Up Schema

First thing we did was live in Google Search Console. We spent hours in the Performance report, filtering for queries with tons of impressions but a falling average position and, most importantly, a CTR that was way lower than it should be. This let us see exactly which keywords were taking the biggest hit from AI Overviews. We also fired up tools like Ahrefs and Semrush to see which competitors were getting featured in those same AI Overviews so we could reverse-engineer what they were doing right with their content structure.

Our plan had two parts: content restructuring and an aggressive schema markup implementation. For the content, we found pages that were ranking well but were being completely leapfrogged by the AI Overview. These pages had all the right info, but they didn’t give a quick, direct answer. So we started adding “answer boxes” or summary paragraphs right at the top of the articles to give the user (and the AI) the answer immediately. For example, an old article called “How to Choose the Right Backpack for Hiking” got a new 50-word intro that defined the key things to look for, capacity, fit, and material, before the article dove into the details.

The second part was a complete overhaul of our schema markup. We went all-in on FAQPage schema for Q&A sections, HowTo schema for our step-by-step guides, and detailed Product schema with specs and review snippets. We were trying to give Google explicit road signs to make our content super easy for its AI models to parse and use. Just slapping on basic schema didn’t work. We learned the hard way that the schema had to perfectly mirror the content’s structure and break down information into tiny, atomic pieces.

Creative Approach: Be the Authority, Fast

This wasn’t a visual design campaign. It was all about informational precision. We told our content team to write like journalists, focusing on clarity and authority. Every single claim had to be backed up, either with our client’s internal product data, quotes from experts, or links to legit external sources (like studies on fabric durability or safety guides from national parks). We hammered the importance of brevity in the new summary sections, writing answers that could fit neatly into an AI Overview snippet. Instead of a long, winding paragraph on tent materials, for instance, we created a simple bulleted list comparing the pros and cons of nylon, polyester, and Dyneema.

We also played around with adding specific, high-quality images and short videos to the content. We did this for the user experience, of course, but we also hypothesized that rich media could act as another signal of quality for the AI models. We didn’t have hard proof that visual content was getting pulled into AI Overviews at the time, but we figured it could contribute to overall page authority and engagement, which might indirectly help. This meant writing highly descriptive image alt text and making sure all our videos had full transcripts available.

Targeting and Implementation

Our targeting was 100% keyword-driven. We used Google Search Console to find the exact queries where our client’s pages were getting impressions but had terrible CTRs, a dead giveaway for AI Overview interference. From that list, we prioritized pages already ranking in the top 10 for those queries, since they were the most likely candidates for Google’s AI to consider for an overview.

We rolled out the changes in phases. In October, we tackled the top 50 most-impacted pages, doing content audits and adding the first round of schema. In November, we expanded to the other 150 pages and started A/B testing different summary formats. For example, some pages got a single summary paragraph, while others got a full-blown FAQ section right at the top. We used Hotjar to watch user behavior on these new page designs, checking things like scroll depth and time on page to make sure our changes weren’t accidentally wrecking the user experience.

Campaign Performance: Metrics and Analysis

Here’s what the numbers showed about the AI Overview impact and how our fixes actually performed. Below is the breakdown of the key metrics from our campaign.

Organic Traffic & CTR

Pre-campaign (September 2025) vs. Post-campaign (December 2025) for the 200 targeted keywords:

• Average Organic CTR for targeted keywords: Dropped from 5.8% (September) to 4.9% (December). That’s a 15.5% CTR reduction on these specific queries, even with all our work. This shows the brutal reality of AI Overviews: they answer the question right on the SERP, so people don’t need to click.
• Overall Organic Sessions from targeted keywords: Fell by 10% (from 15,000 to 13,500 sessions). The CTR drop was steeper, but the session decline wasn’t quite as bad because we got a slight bump in impressions for some keywords. So, people were seeing our content, just not always clicking on it.

AI Overview Visibility

One of our main goals was to get the client featured inside the AI Overview snippets. We used a custom script to manually check SERPs for our 200 target keywords and track how often our client’s domain appeared.

• Share of Voice in AI Overviews: Grew from 8% (October) to 16% (December). This told us our content restructuring and schema efforts were successfully making our pages more “AI-friendly,” leading to them being cited more often in the generated summaries.
• Direct Clicks from AI Overviews (estimated): Google Search Console still doesn’t attribute clicks directly from AI Overviews, but we saw a 3% lift in clicks to pages that were consistently getting featured. This suggests some users still click the source link for more detail.

Conversions and Cost

We tracked conversions from the 200 targeted content pages. Since these are top-of-funnel informational pages, we were mostly looking at micro-conversions like newsletter sign-ups and guide downloads.

• Total Conversions from targeted pages: Increased from 250 (September) to 310 (December), a 24% jump. This was a huge win. It suggests that even though overall traffic dipped, the people who *did* click through were much higher quality and more engaged with our AI-optimized content.
• Cost per Conversion (CPL for leads): Dropped from $48 (September) to $38 (December). A $10 reduction per lead is a big deal, especially on a $12,000 budget. If you look at the raw math, the $12,000 campaign cost divided by the 60 extra conversions (310-250) gives you a $200 cost per *incremental* conversion. But looking at the improved CPL across all conversions from these pages, the efficiency gain was obvious.

What Worked

The biggest win came from putting concise “answer boxes” at the top of content pages, immediately followed by complete FAQPage and HowTo schema markup. The pages where we gave direct, authoritative answers in a highly structured format saw the biggest jumps in AI Overview visibility. For example, a guide on “winter camping tent insulation” that we updated with a clear summary and FAQ schema saw its inclusion rate in AI Overviews shoot up from 5% to 25% for related searches.

Focusing on internal linking was another quiet success. By adding strategic links from our newly optimized informational pages to relevant product pages, we saw a small but real increase in product page visits coming from that content. This helped soften the blow from the direct traffic loss by improving the user’s journey down the funnel.

What Didn’t Work as Expected

The big surprise? All the work we put into visual content optimization, detailed image alt text, video transcripts, didn’t seem to have any immediate effect on AI Overview visibility. It’s possible that Google’s AI just prioritizes text for its summaries, or maybe the effect of visual cues is more subtle and takes longer to show up. We just couldn’t find a direct link between having highly optimized images and getting featured more often, which was odd given how much Google usually pushes rich media.

Also, just making content “shorter” was a bad move. The pages we condensed too much, losing their authoritative depth, actually saw engagement metrics drop, even if they sometimes got pulled into an AI Overview. Google’s AI still wants complete, well-researched content. You just have to present it in a way that lets the machine quickly pull out the key facts.

Optimization Steps Taken

This wasn’t a set-it-and-forget-it campaign. We were constantly tweaking things. We did weekly audits of the top 10 AI Overviews for our main target keywords, looking for patterns in the competitor content that Google’s AI seemed to like. This pushed us to make our “answer boxes” even more direct and to make sure every answer was supported by a clear data point or a link to a reputable source.

We also kept refining our schema. We started with general schema types, but through testing, we learned that being hyper-specific worked much better. For instance, using Product schema to detail out individual product features within a comparison article got us better results than just using a broad category. This meant doing more granular data modeling in our CMS, but it was worth the effort for the increase in AI Overview features.

Lessons Learned and Future Outlook

This campaign proved one thing without a doubt: AI Overviews are a permanent part of Google, and they have completely changed the way users interact with organic search results. The old click-through model for informational searches is under assault, which means you need a proactive and flexible SEO strategy. Yes, the immediate hit to our organic CTR for those keywords was negative. But the fact that we could increase our visibility inside AI Overviews and actually improve conversion rates from those pages shows that you *can* adapt.

My take is this: marketers have to stop chasing clicks and start optimizing for answers. If your content provides the best, most direct, and most authoritative answer to a question, you have a fighting chance of getting featured in an AI Overview. And it turns out the traffic that *does* still click through from these AI-heavy results seems to be much higher quality, which leads to better conversion numbers. This isn’t about traffic volume anymore. It’s about traffic quality. The future of Google SEO in a world run by AI will belong to sites that provide direct, verifiable information that’s structured so a machine can easily read it. That means a real investment in content architecture, not just old-school keyword tactics.

The key lesson here is that while AI Overviews eat some of your direct organic clicks, they also give you a new way to become the definitive source of information in your niche. By focusing on concise, authoritative content that’s backed by strong schema, you can not only survive the traffic loss but actually improve the quality of your audience and your conversion efficiency. This requires a constant, data-driven approach to your content strategy and a willingness to accept that the search game is always changing.