Google’s AI Overviews, Opt-Outs, and the Post-Click Problem

Google’s AI Overviews, Opt-Outs, and the Post-Click Problem - Portugal Business News

SEO News - Google Search has entered its AI era, and it’s giving us answers on a silver platter right there in the SERP. Great for hurried users; nerve-wracking for publishers watching their traffic dwindle.

In late January 2026, Google flipped the switch on a major AI Search update: Gemini 3 is now the default model generating those AI Overviews atop your search results. And AI Search isn’t a limited experiment anymore, as we saw previously with rollouts in different geographies at different paces. AI Mode is rolling out globally, woven natively into mobile apps and browsers so that anyone can chat with Google’s AI assistant mid-search.

Google is basically saying: “Here’s your answer right on the results page, no further clicks needed.”

Gemini 3 and the Rise of AI Overviews

This Gemini 3 upgrade means Google’s AI answers are more powerful (Google touts “best-in-class” quality), and they’re seamlessly integrated. On mobile, tapping a “Show more” in an AI Overview now glides you into a full AI chat without ever leaving the search page. It’s a smooth, conversational search experience by design. Google is making Search feel less like a list of blue links and more like an interactive Q&A session. They’ve effectively built a slick on-ramp from a quick AI summary to a deeper AI conversation.

For users, this is cool and futuristic. For content publishers and SEOs, it’s terrifying. Because every answer Google’s AI provides is one less reason for a user to click through to a website. It’s the classic “zero-click search” problem on steroids, now turbocharged by generative AI.

Google’s mission has always been to give users information as directly as possible (remember, their mission statement says nothing about sending traffic to websites). With Gemini 3, they’re closer than ever to that mission, and publishers are feeling the squeeze.

The Post-Click Gut Punch: Vanishing CTRs

Let’s talk numbers. When Google first introduced AI Overviews, publishers saw an alarming drop in click-through rates (CTR). Now, with widespread AI answers, the data is downright brutal. One recent analysis found that if an AI Overview appears and your site isn’t cited in it, organic CTR plunges by about 65% on those searches. Even if you are cited as a source, you still suffer roughly a 50% drop in clicks. Paid search ads are hit even harder – seeing up

to a 78% drop in CTR when AI answers are present.

To make it worse, even when there is no AI box on the page, organic CTR has been falling year-over-year (a ~46% drop in organic clicks was observed independent of AI, thanks to other rich features stealing attention). In short, the click is becoming an endangered species on Google Search.

Publishers are sounding the alarm. In 2025, Chartbeat data showed Google Search referral traffic to publishers fell by 33% globally. The Reuters Institute reports news execs fear search traffic will plunge 43% more over the next three years thanks to AI answers and chatbots. And it’s not just news: content sites across the board – from travel blogs to how-to guides – are seeing double-digit drops in traffic when AI results appear.

It’s a post-click problem: Google’s AI might answer the query perfectly, but the publishers who supplied the info get zero visit in return. As one industry insider put it, Google is “able to extract valuable data without reward,” turning the open web into a one-way street. Being cited as a source in an AI Overview is nice for credibility, but if users rarely click those citations (in one study, they clicked a cited source on only ~1% of AI page visits), that citation isn’t paying anyone’s bills. No click means no page views, no ad impressions, no lead generation, nothing. Little wonder publishers are fuming.

Enter the Opt-Out Debate: Fight or Adapt?

Regulators have noticed the power imbalance. In late January, the UK’s Competition and Markets Authority (CMA) proposed new rules forcing Google to give publishers an “opt-out” switch for AI features. Essentially: publishers could say “don’t use my content in your AI Overviews or training data,” and Google must comply without demoting that site’s normal search rankings. Google, facing the heat, publicly conceded that it’s “exploring updates” to let sites opt out of generative AI in Search – though in the same breath it warned that any opt-outs must not “break” the user experience. (Translation: we’ll give you controls, but nothing that makes Google Search look worse for users.)

But the UK isn’t alone — the EU’s antitrust investigation, which I broke down for Search Engine Land, is tackling even bigger questions around AI training, publisher rights, and how SEO visibility gets decided in this new ecosystem.

For publishers, an opt-out sounds like a possible lifeline. After all, if Google’s AI is cannibalizing your traffic, why not pull your content out of the AI’s feeding trough? In theory, opting out means Google can’t use your articles to answer queries, potentially forcing users to click your link instead. It’s a form of protest – reclaiming your content from Google’s answer box. And according to one poll of the SEO community, about one-third of publishers say they will opt out as soon as that mechanism is available. (In that survey of 350+ folks, 33.2% said they plan to block their content from AI Overviews. Meanwhile, 41.9% said they would not opt out, and 24.9% were undecided. The community is clearly divided.)

Opting out may be a lose-lose situation

But here’s the catch: opting out might feel like fighting back, yet it could be a pyrrhic victory. Yes, you’d withhold your content from Google’s AI summaries – but that doesn’t stop Google from showing an AI summary. It just means it will use someone else’s content instead of yours. The AI answer still appears, the user still gets the info without clicking, and now your brand isn’t even visible as a source. As one publisher advocacy group bluntly put it, “Opting out... won’t change [the traffic loss], it’ll just stop publishers from being able to access the meagre crumbs of traffic they might get from being present in AI Overviews.” In other words, you forfeit even the slim chance that a reader sees your name or clicks that 1% citation link. Those “crumbs” go to a competitor who remained opted-in.

This creates a wicked opt-out paradox. Block Google’s AI and you protect your content from being freely mined – but you also render yourself invisible in a portion of search results. Stay in the AI Overviews, and you’re implicitly allowing Google to keep siphoning off direct traffic – but at least your brand appears in the new AI results, which could have indirect benefits (branding, future credibility, etc.). It’s a damned-if-you-do, damned-if-you-don’t scenario. As one analysis noted, it comes down to weighing “immediate traffic concerns against long-term visibility risks.” Many are asking: Which can we live without – some short-term clicks, or our foothold in the future of search?

Bing’s Transparency Play

While Google wrestles with opt-out pressure, Microsoft’s Bing is taking a different tack, at least in terms of optics. Bing has its own AI chat in search (the new Bing with ChatGPT tech), and publishers have similarly worried about “no-click” info theft there. In response, Bing has begun testing an “AI Performance” report in Bing Webmaster Tools that actually shows publishers how often their content is getting cited in AI answers. It even breaks it down by which pages were cited and what types of queries led to those citations.

On one hand, this is a welcome gesture toward transparency. It acknowledges that AI answers are a new search surface that site owners care about. However, before we applaud Microsoft too much – guess what metric is conspicuously absent from the AI Performance report? Clicks. Bing will not tell you how many people actually clicked through from those citations. So if Bing’s chatbot quoted your blog 100 times, you’ll see that, but you won’t have

a clue if it sent even a single actual visitor your way. As Search Engine Land’s coverage put it, “Sadly, it does not show you clicks, so you still have no idea what your click-through rate is from AI search results.”

It’s basically the same veil Google has drawn over its AI product. Neither search giant wants to quantify just how badly AI answers are cannibalizing clicks – likely because the numbers aren’t pretty. “It just feels like all the search engines are deliberately hiding this data from us,” one editor remarked pointedly. The lack of click transparency is almost an admission that, yes, these AI answer boxes are siphoning off a lot of traffic, but the platforms would rather

keep that inconvenient truth blurry.

Still, Bing’s move is noteworthy. By giving at least citation counts, Bing is implicitly encouraging publishers to start valuing “AI visibility” alongside traditional clicks. If you can’t have the traffic, well, at least you got the mention. This is the new reality: we might need to treat an AI citation or a brand mention in a chatbot’s answer as a success metric, the way we used to treat a page visit. That’s a hard pill to swallow (you can’t pay the bills with “brand visibility,” I know), but it’s where things are headed unless or until the economics of all this catch up.

Opt-In to the Future: Why AI Visibility Can’t Be Ignored

It’s easy to see AI answers as an existential threat to the open web. But they’re also not going away. Google isn’t going to put the genie back in the bottle (Sundar Pichai literally said conversational AI is the future of Search). So publishers and SEOs face a tough choice: resist and opt out, or adapt and find a way to win within the new system.

Increasingly, savvy voices in the SEO community are pushing for the latter. As one commentator noted during the zero-click debate, “SEO is shifting from pure traffic thinking to visibility, brand, and attribution across search and AI surfaces. If clicks keep shrinking, measurement and monetization models have to evolve just as fast.” In other words, we might hate the game, but we have to learn the rules if we want to keep playing.

What might that entail? For one, optimizing for AI visibility – call it Answer Engine Optimization. This means structuring your content in a way that the AI is more likely to grab it (much like featured snippet optimization in the old days). It could mean providing succinct Q&A pairs in your text, using schema markup, and ensuring your content is high-authority so that the AI trusts it. If your content must be the one scraped, better it be your site cited in the

answer than your competitor’s. Yes, a citation without a click is far from ideal – but it’s better than being invisible in the AI answer altogether. Moreover, having your brand appear as the authority in an AI-delivered answer can have indirect benefits: brand recognition, perceived expertise, and maybe a user will remember and seek you out later (or ask the AI follow-up questions specifically about your brand).

There’s also the possibility that down the line Google will refine attribution in AI answers to drive more clicks. The UK CMA explicitly is pushing for “prominent and accurate attribution” because it can improve click-through rates. Google could, for instance, experiment with bigger source links, or multiple source links, or other UI tweaks to encourage exploration. If you’ve opted out, you won’t be around to benefit from any improvements there. In short,

opting out might protect you in the short term, but it could mean forfeiting your spot in the long-term future of search. It’s telling that roughly two-thirds of publishers polled aren’t planning to block AI content – they’re not happy about the situation, but they may see more risk in disappearing from AI results entirely.

So, what’s the play for those of us in SEO and content strategy?

• Monitor and Adapt: Keep a close eye on how much traffic you’re losing where AI Overviews appear. Use whatever data you can (even Bing’s half-baked AI report) to measure your “AI visibility.” If you’re consistently seeing your competitors cited while you’re absent, it’s time to investigate why.

• Optimize for AI Answers: This is the new game. Incorporate concise answers, FAQ sections, and schema markup that might make your content a preferred source for AI. In essence, write for the user and the AI – anticipate the questions and answer them clearly in your content. Being the quoted source in an AI Overview is the new

  Page 1 ranking.

• Leverage Branding: If your content is cited, make sure your brand is front and center. For example, having a strong, identifiable domain name (or author name) that’s shown in citations can at least build brand awareness. Some publishers are even experimenting with subtle content watermarks (like mentioning their brand in the text

  that might get quoted) to ensure the AI doesn’t strip out all identifying info.

• Stay Informed on Opt-Out Developments: Google’s opt-out controls, driven by regulators, are likely coming in some form. Make a strategic decision (with data) on whether to use them when the time comes. But remember the opt-out paradox – know exactly what you stand to gain or lose. It might be that partial opt-out (for certain pages) could be a compromise strategy, if that becomes possible.

• Pressure for Better Attribution and Rewards: As an industry, content creators should continue to push Google (and Bing) for better solutions – whether that’s improved citation links, traffic analytics, or even compensation models for content used in AI. We have to advocate collectively; otherwise, the “AI vs. open web” scales will keep

  tipping away from us.

Key Takeaways for Publishers & SEOs

• Don’t Rush to Opt Out Completely. Opting out of Google’s AI answers might feel like sticking it to Big G, but it also removes you from a chunk of search visibility. Your competitors who stay in will grab that AI spotlight. As one group warned, opting out could just forfeit the “meagre crumbs” of traffic you do get from AI Overviews. Consider the long-term trade-off before pulling the plug.

• Optimize for AI Visibility. Treat AI Overviews as the new battleground. Structure your content to answer questions directly and succinctly. Use FAQ sections, how-to lists, and clear definitions. This increases the odds that Google’s AI will pick your content to feature – keeping your brand in the conversation even when clicks drop.

• Focus on Brand and Engagement Metrics. If 60+% of searches might soon be zero-click, success can’t just be measured in clicks. Start tracking and valuing metrics like brand mentions in AI, direct traffic (from people who later navigate to you), and engagement from those who do click. SEO isn’t just about funneling visits

  anymore; it’s about staying visible wherever the consumer is getting answers.

• Keep an Eye on Bing (and Others). Bing’s new AI Performance report is a sign of how search engines might offer visibility metrics beyond clicks. It’s not perfect (no click data yet), but it’s a start. Use these tools to understand how often your content surfaces in AI answers. If Bing is citing you frequently, that’s a good proxy that your

  content is considered authoritative – intel you can use to adjust your Google strategy too.

• Prepare for an Attribution Arms Race. Regulators are pushing for better attribution in AI answers, and Google may be forced to comply. We could see more prominent source links or even traffic-sharing models in the future. Make sure you’re in the game to benefit from these changes. Completely opting out means if Google finds a

  way to make AI answers send traffic or value back to publishers, you’d miss out.

Bottom Line

The search landscape is fundamentally changing. We’re moving from an era of ten blue links and predictable clicks, to an era of AI-curated answers where visibility might matter as much as traffic. It’s a challenging shift (and frankly a bit of a raw deal), but it’s the reality we have to navigate. Rather than rage-quitting the system, the smarter play for most will be to adapt: ensure you’re one of the voices the AI can’t help but cite, push for fair treatment, and evolve

your metrics for success. Google’s AI Overviews may be eating into our clicks, but with the right strategy, we can still make sure they don’t eat our lunch.

Author:  Maryanna Franco - Founder of  SEO agency BrilliantSEO