Google’s AI Overviews Are Stealing Your Clicks - Fight Back

If you’re wondering how to optimize for AI Overviews, you need more than a top-10 ranking. You need a page that answers the query directly, covers the related questions Google may explore, gives the system clear passages to cite, and still earns the click when the answer is not enough. That matters because AI Overviews are changing the economics of organic search: fewer searches end in an open-web click, while the pages that remain visible inside the answer can become disproportionately valuable

How AI Overviews Affect Organic Traffic and Clicks

The June 2026 SparkToro/Similarweb analysis put Google’s zero-click rate at 68%, up substantially over the last decade. Similarweb also found AI Overviews appearing on more than 20% of searches in its panel, while searches showing an Overview experienced a sharp reduction in click-through.

The important part is not the headline average. It is where your clicks are disappearing.

A question such as “what is X?” can often be answered completely on the results page. A comparison query may still send a buyer to multiple websites before they decide. A commercial query may be much more dependent on the click.

That makes AI Overview SEO a query-by-query problem, not a site-wide traffic panic.

SEO for AI Search Starts With Queries, Not Algorithms

Ahrefs’ analysis of 1.9 million AI Overview citations found that 76% of cited URLs also ranked in Google’s traditional top 10, with the median cited result around position two. Its broader research also found AI Overviews appearing especially often on question-based, longer and informational queries.

So the first rule of SEO for AI search is surprisingly familiar:

You still need to be good at Google Search.

But ranking is only the starting point.

The more useful question is:

Which of your valuable queries are generating AI Overviews, and what happens to the click when they do?

Pull your top revenue-driving queries from Search Console and classify them by intent.

The averages are useful context. Your own revenue-weighted query set is the strategy.

How to Optimize for AI Overviews: What Actually Works

Google’s official guidance is important here because it cuts through a lot of the mythology around generative search. There are no additional technical requirements specifically for AI Overviews beyond being eligible for Google Search. The same fundamentals still matter: crawlability, indexability, internal linking, useful content, textual access to important information, good page experience and accurate structured data.

That means the question is not: “What special AI markup do I need?”

It is: “Is my existing SEO strong enough, and is the page structured so Google can actually use it in an answer?”

That distinction should shape the entire programme.

Create Content That AI Overviews Can Use and Cite

The simplest content change is also one of the most useful.

Put the direct answer near the top of the page, before the supporting nuance.

For a pricing query, that might look like:

<h1>How Much Does [X] Cost?</h1>

<p><strong>Short answer:</strong> [X] costs $A–$B for

[typical case], driven mainly by [factor 1] and

[factor 2]. Most [segment] teams budget $C

for year one.</p>

The point is not to write for a machine instead of a person. It is to remove the unnecessary distance between the question and the answer.

And there is an important caveat: do not inflate the page just to make it look comprehensive.

Ahrefs found almost no relationship between word count and AI Overview citations. Its research also highlights the risk of intent dilution: adding tangential material can bury the answer that originally worked.

For AI Overview SEO, relevance beats length.

Optimise Content for Questions Behind Every Search

This is where query fan-out changes the work.

Google says AI Overviews may use query fan-out, issuing multiple related searches across subtopics and data sources while constructing a response. That means your page can become useful not only because it answers the original query, but because it answers one of the supporting questions the system explores.

Take: “Which CRM is best for a 50-person SaaS company?”

The useful supporting questions might include:

What does it cost?

How long does implementation take?

Which integrations matter?

What are the best alternatives?

What are the main trade-offs?

Which option works for a smaller sales team?

That is a better content brief than repeating “best CRM for SaaS” 20 times.

Use People Also Ask, Search Console data, customer questions and related searches to build the supporting-question set. Then make each important section capable of answering one of them on its own.

Why Helpful Content Update Still Matters

The helpful content conversation did not become irrelevant because Google added AI Overviews. In fact, the underlying principle matters more: don’t create content simply because a keyword exists.

Google’s people-first guidance asks whether a page provides substantial value, demonstrates knowledge and gives someone enough information to accomplish what they came to the page to do. It specifically warns against producing large amounts of content mainly to attract search traffic or summarising what others have already said without adding meaningful value.

That has a direct AI-search implication.

If your page is another version of information already available everywhere, an AI system has little reason to choose it.

A stronger page adds something: first-hand experience, original data, a useful comparison, a documented process, a named expert view or a trade-off competitors avoid discussing.

That is what makes content more than a summary.

How to Recover Traffic Lost to AI Overviews

The fastest AI Overview win may already be sitting in your content library.

Start with pages that:

rank for valuable queries,

are now exposed to AI Overviews,

have lost clicks or CTR,

and still have business value.

Don’t automatically rewrite all of them.

First ask what changed.

Is the answer now being satisfied directly in the SERP? Is your page failing to provide a concise answer? Are competitors being cited instead? Has the content become outdated? Did a previous update bury the original answer under unnecessary detail?

Ahrefs’ recent research is a useful warning here: adding more content can dilute intent rather than strengthen it.

Sometimes the right AI Overview SEO move is removing content, not adding it.

How to Measure AI Overview SEO Performance

Search Console will not hand you a clean “AI Overview traffic” number. So measure the impact indirectly.

Start by identifying which of your priority queries trigger AI Overviews. Then track three things alongside each other:

HubSpot’s current playbook recommends establishing a keyword-to-prompt baseline, tracking citation presence and segmenting queries by intent and funnel stage.

The reporting change is important.

Don’t report: Organic traffic down 35%.

Report: Informational clicks down 35% on queries with AI Overviews; citation presence up from X% to Y%; commercial traffic stable.

That turns a frightening traffic chart into an optimisation problem.

What Most AI Overview SEO Strategies Still Miss

The leading AI Overview guides already cover the fundamentals: traditional SEO, question-based queries, query fan-out, authority, structured data, direct answers and measurement. Ahrefs is particularly strong on the relationship between traditional rankings and AI Overview citations, while HubSpot goes deeper into content structure, schema and measurement workflows.

Where there is still room to be more useful is the fight-back decision.

Not every declining page needs more content.

Not every AI Overview query deserves the same investment.

And not every lost click is recoverable.

The smarter approach is to identify which query classes still generate valuable clicks, which should be treated as citation plays, and which pages need to be rebuilt rather than expanded.

Differentiating by Industry

Publishers and media: informational clicks are often the most exposed. Compete for citations, build proprietary data and create assets that are harder to compress into a one-paragraph answer.

B2B SaaS: top-funnel informational queries are more exposed, but comparison, pricing and commercial queries can still drive meaningful visits. Shift effort toward the pages closest to revenue.

Ecommerce: AI-generated product and recommendation experiences make product data, reviews and structured information increasingly important. Don’t hide important attributes inside images or inaccessible interfaces.

Local businesses: AI Overviews can pull from business profiles and reviews alongside websites. Your search strategy therefore extends beyond the page itself.

Overview Fight-Back Checklist

Pull your top 100 revenue queries.

Tag each by search intent and AI Overview presence.

Identify where impressions are holding but clicks are falling.

Prioritise the 10 highest-value exposed pages.

Put the direct answer near the top of each page.

Rebuild H2s around real questions and supporting queries.

Remove tangential content that dilutes intent.

Add original evidence, useful comparisons and genuine trade-offs.

Verify crawlability, server-rendered content and structured data.

Track citation presence and conversion alongside rankings.

Bigger Picture

AI Overviews are not simply taking clicks away. They are changing what a valuable search result looks like. For some queries, the SERP is becoming the answer.

For others, it is becoming the filter that decides which sources deserve the next click. That means the fight is no longer just for position one. It is for the answer, the citation and the click that remains after the answer has already been given.

The brands that adapt will not be the ones that publish the most. They will be the ones that understand which questions still deserve a click, which deserve a citation, and what each page needs to earn both.

Frequently Asked Questions

Do AI Overviews appear on every search?

No. Google says AI Overviews appear when its systems determine they add value beyond classic Search, and current third-party studies show their prevalence varies substantially by dataset and query type. Your own query mix matters more than a single market-wide average.

How much traffic can AI Overviews take away?

There is no universal percentage. Current studies report substantial CTR reductions when an Overview appears, but the impact varies by device, query type and dataset. Measure your own revenue-weighted queries rather than applying one average to the whole site.

Can I appear in an AI Overview without ranking first?

You don’t need position one specifically, but strong traditional visibility remains important. Ahrefs found 76% of the AI Overview citations in its 1.9-million-citation analysis also appeared in the traditional top 10.

Should I write longer content to improve AI Overview visibility?

No. There’s almost no relationship between word count and AI Overview citations. The better goal is to match the query, answer it clearly and cover the relevant supporting questions without diluting the page.