AI Overviews Are Eating Your Traffic. Here Is What You Lose

AI Overviews are one of the most consequential changes to Google Search in years, and not in a way that benefits most publishers. When Google answers a question directly at the top of the page, a large share of users never scroll further. Clicks that used to reach your site are absorbed before they leave Google’s interface, and the SEO playbook many businesses built their growth on is quietly becoming less reliable.

The negative impacts of AI Overviews on SEO are real and measurable: reduced organic click-through rates, content cannibalisation, loss of attribution, and a structural shift in how value flows through search. Understanding what you are actually losing, and why, is the first step toward making sensible decisions about where to invest your content effort next.

What Is Actually Changing in Organic Search?

For most of the last two decades, organic search operated on a reasonably predictable model. You created content that answered questions well, Google ranked it, users clicked through, and you earned traffic. The model was never perfect, but it was legible. You could measure it, optimise for it, and build a business around it.

AI Overviews change that model in a structural way. Google now synthesises answers from multiple sources and presents them above the organic results. For a significant share of informational queries, the answer is complete before the user has any reason to scroll. The source content still exists, and Google may even cite it, but the click that used to follow the answer no longer reliably happens.

I have spent time looking at this through the lens of the businesses I have worked with across 30 industries. The pattern is consistent: the sites most exposed are those that built their organic presence almost entirely on informational content, the how-to articles, the explainers, the definition pages. That content still ranks. It just drives less traffic than it used to.

If you want to understand how AI is reshaping search and content more broadly, the AI Marketing hub at The Marketing Juice covers the landscape in depth, from content strategy to tools to what the shifts actually mean for marketing teams.

How Do AI Overviews Reduce Click-Through Rates?

The mechanism is straightforward. When a user searches for something and the first thing they see is a complete, well-formatted answer, a meaningful proportion of them stop there. They got what they came for. The organic listings below the AI Overview exist, but they are competing for attention that has already been partially satisfied.

This is not a new phenomenon in principle. Featured snippets and knowledge panels have been suppressing clicks for years. AI Overviews are the same dynamic at greater scale, with more comprehensive answers and a more prominent position on the page.

The queries most affected are what you might call “resolved” queries: questions with a clear, factual, or procedural answer. What is a conversion rate? How do you calculate ROAS? What does schema markup do? These are the queries that power enormous amounts of content marketing. They are also the queries where AI Overviews perform most confidently, because the answer is bounded and the synthesis is reliable.

When I was running agency teams and we were building out content programmes for clients, the informational layer was always the entry point. You earned trust with the how-to content, then converted it downstream. That model depends on the click happening. If the click does not happen, the funnel does not start.

Which Types of Content Are Most at Risk?

Not all content is equally exposed. The risk is concentrated in specific content categories, and understanding that distribution matters for how you respond.

Informational content at the top of the funnel is the most vulnerable. Definition articles, how-to guides, comparison overviews, and FAQ-style content are precisely the formats that AI Overviews are built to synthesise. If your content strategy has a heavy weighting toward these formats, you are carrying more risk than you may realise.

Transactional content is less immediately affected. When someone is ready to buy, to book, to request a quote, they still need to go somewhere to complete that action. AI Overviews do not process transactions. The closer your content sits to a commercial action, the less exposed it is to zero-click behaviour.

Navigational queries are similarly protected. If someone is searching for a specific brand or tool, they are not looking for a synthesised answer. They want to go to a particular place, and Google knows that.

The content in the middle, the research-phase content, the “best of” comparisons, the detailed guides that sit between awareness and decision, is the most complicated case. AI Overviews can partially satisfy these queries but often cannot fully resolve them. There is still a click to be earned, but it is harder to earn than it was.

Resources like the Ahrefs AI SEO webinar series have explored how query intent maps onto AI Overview behaviour in useful detail. Worth reviewing if you are trying to audit your own content exposure.

Does Being Cited in an AI Overview Drive Traffic?

This is one of the questions I hear most often, and the honest answer is: sometimes, but not reliably, and not at the volume you might hope.

When Google cites a source in an AI Overview, it does include a link. Some users click those links, particularly when they want more depth than the overview provides. But the click rate on those citations is substantially lower than the click rate on a traditional first-position organic result. You are getting attribution without the traffic that attribution used to imply.

There is also a visibility problem. AI Overviews do not always display citations prominently. On mobile in particular, the cited sources are often collapsed behind an expand interaction. A user who does not actively look for them may not know they exist.

I have judged the Effie Awards and spent time thinking about how marketing effectiveness gets measured. One of the consistent patterns I see is that organisations conflate visibility with impact. Being cited in an AI Overview is visibility. It is not the same as a visit, a lead, or a conversion. If your reporting treats them as equivalent, you will overestimate the value of your content and underinvest in fixing the problem.

What Happens to Attribution and Measurement?

The measurement problem is, in some ways, more damaging than the traffic problem. When traffic falls, you can see it and respond to it. When attribution breaks down, you lose the signal you need to make good decisions.

AI Overviews contribute to a broader zero-click ecosystem where users get answers, form opinions, and sometimes convert, without ever appearing in your Analytics data. They may visit your site later via direct traffic or branded search, and you will have no way of knowing that the experience started with an AI Overview that cited your content.

This is not a new problem. Dark social, direct traffic, and attribution gaps have been part of the measurement landscape for years. But AI Overviews add another layer of opacity to a picture that was already difficult to read clearly.

The practical consequence is that organic search performance metrics become less reliable as indicators of content quality or SEO effectiveness. A page that is consistently cited in AI Overviews but drives low direct traffic is not necessarily failing. It may be doing more work than your data shows. The challenge is that you cannot optimise what you cannot measure, and you cannot make confident investment decisions on the basis of incomplete data.

Early in my career, when I was building websites by hand because the budget for external support did not exist, I learned that the tools available to you shape how you see the problem. Analytics is a perspective on reality, not reality itself. That has never been more true than it is now.

How Does AI Overview Behaviour Affect Content Investment Decisions?

This is where the commercial impact becomes concrete. Content programmes are expensive. Writing, editing, publishing, maintaining, and promoting content at scale requires real resource. The business case for that investment has historically rested on organic traffic projections. If those projections are structurally optimistic because they do not account for AI Overview suppression, then content budgets are being allocated on faulty assumptions.

I have run agencies and turned around loss-making businesses. The pattern I saw repeatedly was that complexity accumulates gradually and nobody questions it until the numbers stop working. Content programmes built for a pre-AI Overview search landscape are now carrying that kind of legacy complexity. The articles that justified their existence on the basis of informational traffic projections need to be re-evaluated against a different set of assumptions.

That does not mean abandoning content. It means being more deliberate about what content you create and why. Content that supports conversion, builds authority, earns links, or serves a purpose beyond ranking for a resolved query is more defensible than content that exists solely to capture informational search volume.

The Semrush overview of AI optimisation trends covers some of the strategic shifts content teams are making in response to these pressures. The direction is broadly consistent: less volume, more depth, more focus on the content that AI cannot easily synthesise.

Is There a Compounding Effect on Domain Authority?

One of the less discussed risks is the indirect effect on domain authority over time. Organic traffic is not just a business outcome. It is also a signal. Sites that earn consistent traffic, engagement, and links tend to build authority that compounds. That authority makes future content more likely to rank and more likely to be cited.

If AI Overviews suppress traffic to informational content at scale, the engagement signals that support authority building become weaker. Pages that rank but do not get clicked generate less behavioural data. Content that earns citations in AI Overviews but not direct visits does not build the link profile that traditional high-traffic content would have built.

This is a slow-moving problem. It will not show up clearly in a month’s data. But over a year or two, sites that were growing their authority through informational content may find that the compounding effect has slowed or stalled. The pipeline that fed authority has narrowed.

There are no clean solutions here, but the response is the same as it is for most compounding problems: catch it early, diversify the inputs, and do not wait for the numbers to look alarming before you act.

What About Content Quality and Misinformation Risk?

There is a broader concern worth naming, even if it sits slightly outside the immediate SEO conversation. AI Overviews synthesise content from multiple sources and present it as a unified answer. When the source content is high quality and the synthesis is accurate, this works reasonably well. When it is not, the errors are presented with the same authority as the correct information.

For publishers, this creates a reputational risk that is difficult to manage. If your content is cited in an AI Overview that contains an error, the association exists whether or not the error originated with you. Users who see the AI Overview may not distinguish between what Google synthesised and what your original content said.

This is one of the reasons that original research, proprietary data, and first-person expertise are becoming more valuable in content strategy. AI systems struggle to synthesise what does not already exist in training data or crawlable content. If your content contains something genuinely original, it is harder to cannibalise and more likely to earn a meaningful citation rather than a buried one.

The Moz analysis of AI content creation touches on the quality and differentiation question in useful terms. The argument for investing in genuine expertise over generic coverage has never been stronger.

How Should SEO Strategy Adapt?

The honest answer is that there is no single clean pivot. AI Overviews represent a structural shift, not a tactical problem you can solve with a plugin or a prompt. The adjustments that make sense are strategic and take time to show results.

First, audit your content by intent. Separate the content that serves resolved queries from the content that serves research-phase or commercial queries. The resolved query content is most exposed. Decide whether it is worth maintaining, whether it can be repositioned to add more depth that AI cannot synthesise, or whether the resource is better deployed elsewhere.

Second, invest in content that is harder to cannibalise. Original data, proprietary frameworks, case studies, and expert opinion grounded in direct experience are all formats that AI systems cannot easily replicate from existing content. They also tend to earn links and citations in ways that generic informational content does not.

Third, diversify your traffic sources. This sounds obvious, but many businesses that grew on organic search have allowed their channel mix to atrophy. Email, owned communities, direct relationships, and social distribution are all channels that do not depend on Google’s interface decisions. The businesses that will be least disrupted by AI Overviews are the ones that were already not entirely dependent on them.

Fourth, measure more carefully. The gap between what AI Overviews show and what Analytics reports is going to widen. Investing in better measurement, including brand search tracking, direct traffic analysis, and content attribution modelling, will give you a clearer picture of what your content is actually doing.

For a broader view of how AI is reshaping marketing strategy and tools, the AI Marketing section of The Marketing Juice is worth bookmarking. The pace of change in this space means the picture will keep shifting, and staying current matters.

The Ahrefs AI tools webinar archive is also a practical resource if you are trying to understand how the tooling landscape is evolving in response to these changes.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions