Google AI Overviews: The Real Impact on Your Traffic

Search Console impressions holding steady while clicks quietly slide isn’t always a ranking problem. For a growing number of business websites, the effect on Google AI Overviews website traffic is the more likely explanation. Google now answers a large share of informational questions directly on the results page, which means the searcher gets what they need before your listing is even opened.

ProfileTree, a Belfast-based digital marketing agency working with SMEs across Northern Ireland, Ireland, and the UK, has repeatedly observed this pattern across client accounts over the past year: rankings and visibility unchanged, but the click-through rate for certain query types trending downward. The pattern isn’t random. It follows query type closely enough that any business can map its own exposure before deciding what to do about it.

This isn’t a reason to panic, and it isn’t a reason to ignore what’s happening either. It’s a reason to understand exactly which searches are affected, which ones are not, and what the sensible next step looks like for each.

The Pattern: Steady Impressions, Falling Clicks

If a page’s impressions in Google Search Console are flat or growing while clicks and click-through rate are falling, and the average position hasn’t moved, that combination points squarely at AI Overviews rather than a ranking drop. Google still counts the impression when your page appears within or alongside an AI Overview, but a searcher who gets a complete answer summarised on the page has no reason to click through to read it in full.

This shows up unevenly. Some pages lose a noticeable share of clicks. Others barely move. The difference almost always comes down to query type, not content quality. A page targeting “what is local SEO” behaves very differently from a page targeting “SEO services Belfast”, even if both rank in a similar position, because Google treats the two queries as fundamentally different search intents.

It’s worth being precise about what’s actually declining. Rankings measure where a page sits in the results. Visibility measures how often it’s shown. Clicks measure how often someone takes action in response to that visibility. AI Overviews interrupt the third link in that chain without necessarily touching the first two. A business owner checking their professional SEO support results and seeing falling traffic naturally assumes something has gone wrong with the SEO itself. Often, nothing has. The page is doing its job; the results page is just doing more of the answering itself.

None of this means the traffic drop doesn’t matter commercially. Fewer clicks generally means fewer enquiries from that particular page, even if the brand impression (appearing as a cited or referenced source) still carries some value. The right response depends entirely on determining which of a site’s pages are genuinely affected by this shift and which are largely unaffected.

Which Queries Trigger AI Overviews (and Which Don’t)

AI Overviews appear most consistently for queries where Google can provide a complete, self-contained answer without sending the searcher elsewhere. In practice, that means:

Informational and definitional queries are the most exposed. “What is technical SEO?” “How does local search work?” “difference between SEO and PPC” and similar phrasing tend to trigger an Overview because the answer is an explanation rather than a decision. These queries sit at the top of the funnel, and they were never the strongest source of direct enquiries in the first place, but they were often useful for building topical authority and capturing early-stage attention.

How-to and process queries sit close behind. “How to set up Google Business Profile” or “how to improve page speed” often get summarised into a short numbered list on the results page itself, particularly when the steps are widely documented and consistent across sources.

Three categories of query, by contrast, tend to keep sending clicks:

Local intent queries. Searches like “web design agency Belfast” or “SEO company near me” carry a strong commercial and geographic component that Google generally still resolves through the Local Pack, Maps results, and organic listings rather than a generative summary. A searcher looking to hire a business wants to see and compare options, not read a synthesised paragraph.

Comparison queries. “WordPress vs Shopify for small business” or “in-house marketing vs agency” involve weighing trade-offs specific to the searcher’s situation. Google is far more cautious about resolving these inside an Overview because the “right” answer genuinely depends on context that Google doesn’t have.

Transactional and commercial queries. Anything with clear buying intent, “SEO services pricing”, “hire a content marketing agency”, “book a website audit”, tends to route to the standard organic and ad results rather than a generative answer, because the searcher’s next action is a business decision, not a knowledge gap.

This is the segmentation worth applying to any site’s own query list. Pull the queries driving each page’s traffic and roughly sort them into the “explains something” and “helps someone decide or find something” buckets. The first bucket is the one losing clicks. The second, for most business sites, is holding up reasonably well.

The Strategic Response, Part One: Becoming a Cited Source Inside the Overview

If a query is going to trigger an AI Overview regardless of what a business does, the sensible response isn’t to fight that outcome. It’s to make sure the business’s own content is the source the Overview draws from and, ideally, cites by name.

Being cited inside an Overview still carries value even without a click. It builds brand association at the exact moment someone is researching a topic, and it’s the closest thing available right now to appearing at position zero on a query a business could never have ranked for outright. Getting there consistently comes down to a handful of practical habits rather than any single trick:

Answer the question in the first two or three sentences of the relevant section. Google’s generative systems pull from passages that state a clear answer immediately and then follow it with supporting details. Content that builds up to its point over several paragraphs is far less likely to be lifted into a summary.

Structure content so each section stands on its own. A section that only makes sense after reading three paragraphs before it is hard for a generative system to extract cleanly. Self-contained sections of roughly 100 to 300 words, each answering one clear sub-question, are much more citable.

Use original data and named specifics wherever genuinely available. Statistics, named examples, and first-hand project details are cited more often than general statements because they add something a generic summary of the topic wouldn’t otherwise contain. Ciaran Connolly, founder of ProfileTree, puts it plainly: “If your page just repeats what every other page says, the AI has no reason to quote you specifically. It’ll quote whoever said something more precisely.”

Cover the sub-questions a searcher would ask next. Overviews increasingly draw from pages that address multiple related angles of a topic rather than a single narrow point. A page on local SEO that also touches Google Business Profile, review management, and citation consistency gives Google more opportunities to lift a relevant passage.

Keep technical fundamentals in order. Structured data, clear headings, and a genuinely fast, crawlable page still matter. A brilliant answer buried in a page Google struggles to parse or render quickly is far less likely to surface anywhere, Overview or otherwise. A website performance audit is a reasonable first step for any business unsure where its technical foundations currently stand.

Tracking whether any of this is actually working requires watching AI citation activity specifically, which behaves differently from standard ranking data. ProfileTree has covered how to measure AI citations in more depth, including which reports actually show this kind of activity today.

The Strategic Response, Part Two: Rebalancing Content Investment Toward Decision-Stage Queries

The second half of the response is less about winning back the queries that Overviews now dominate and more about deliberately shifting investment toward the queries where clicks still convert reliably.

This doesn’t mean abandoning informational content. Top-of-funnel pages still build topical authority, still support internal linking, and still occasionally send traffic on queries that haven’t yet triggered an Overview (Google’s rollout is uneven and continues to expand, so today’s unaffected query can become tomorrow’s Overview trigger). The point is proportion. A content plan weighted heavily toward pure definitional content is now weighted toward the exact query type most likely to lose direct traffic, while under-investing in the query types that still reliably produce enquiries.

Practically, that means giving more planning time and production budget to:

Local and location-specific pages. A dedicated page addressing a service in a specific area, built with genuinely local detail rather than a templated swap of place names, continues to perform well because these queries rarely resolve neatly inside a generative summary. ProfileTree’s own SEO services in the Belfast page are treated internally as exactly this kind of asset.

Comparison and decision-support content. “How to choose” pages, framework-based comparisons, and criteria-driven guidance fall into the category that Google is most cautious about summarising away, because the right answer depends on the reader’s specific circumstances. A content marketing strategy guide built around decision criteria, rather than a generic overview of the topic, fits this pattern well.

Service and pricing pages. Pages describing what a service includes, how it’s delivered, and what it costs sit at the transactional end of the query spectrum. Investment here (clearer process explanations, more specific inclusions, genuine pricing guidance) tends to yield a better return than an equivalent amount of effort devoted to a broad definitional guide. ProfileTree’s digital marketing strategy service page and website design services page both fall into this category.

Case studies and proof assets. These rarely get summarised because they’re inherently specific to one business’s circumstances, and they carry disproportionate weight in a searcher’s actual decision once they’ve narrowed down their options.

This rebalancing also has a knock-on benefit for the citation strategy above. A page written to help someone decide, with real criteria and specific detail, naturally produces the kind of original, non-generic content that generative systems are more likely to cite in the first place. The two halves of this response reinforce each other rather than pulling in different directions.

Beyond written content, video continues to hold up well against this shift, since Google’s generative summaries don’t replace the experience of watching a demonstration or a walkthrough. Businesses weighing up where to redirect budget might reasonably consider video production or YouTube marketing services as part of that rebalancing, alongside adjustments to written content.

Measuring the Difference: Overview, Problem, or Ranking Problem?

Search Console is still the right place to start, but it needs to be read with a specific question in mind rather than a general “is traffic up or down” check.

Check the average position first. If position has held steady or improved while clicks have fallen, that’s the clearest early signal of an Overview effect rather than a ranking issue. A genuine ranking drop appears as a position moving in the wrong direction, usually accompanied by a corresponding decline in impressions for the affected queries.

Check impressions alongside clicks, not clicks alone. A page can show flat clicks and think nothing has changed, while impressions have actually grown. If impressions are up and clicks are flat, the click-through rate has fallen even though the raw click number looks unchanged, and that’s still an Overview signal worth investigating.

Segment by query, not just by page. A single page often ranks for a mix of informational and transactional queries. Aggregate page-level data can hide the fact that one query type is losing clicks while another on the same page holds steady or grows. The query-level report inside Search Console is the only place this becomes visible.

Compare CTR against the position for that query type. Google publishes rough expected click-through rate benchmarks by position for standard organic results. A page holding position 3 with a click-through rate far below what would normally be expected for that position, specifically on an informational query, is a reasonable indicator that an Overview is sitting above it and absorbing the click.

Watch for the query in the “search appearance” filters, where available. Search Console has expanded its AI-related reporting over time, and checking whether a specific search appearance filter is available for a property is worth doing periodically, since Google continues to add visibility into where impressions come from. Google’s own documentation on AI features and your website is the most reliable place to check what’s currently reportable, since this area changes faster than most other parts of Search Console.

A genuine ranking problem, by contrast, usually shows up as a broader pattern: multiple queries on a page losing position together, often alongside a specific event such as a core update, a technical change, or a competitor’s content genuinely improving. If that broader pattern isn’t present and the position line is flat, the more likely explanation is sitting one layer up on the results page itself.

What This Means for Ongoing Reporting

Businesses handling this shift well tend to split reporting into two buckets: “informational query traffic” and “decision-stage query traffic.” The first is genuinely more exposed as Google expands Overview coverage. The second remains a reliable indicator of whether the underlying SEO work is doing its job.

This changes what a traffic drop should trigger. A decline concentrated in informational queries with stable rankings points to the citation strategy above, not a rewrite. A decline in transactional or local queries, especially with a position change, is a real ranking issue and warrants the usual review.

Businesses unsure which bucket their decline sits in can talk to our team for a walkthrough of what the numbers mean for their site.

FAQs