Here is how AI Overviews affect organic traffic: they answer the question at the top of the page, so fewer people scroll down and click your link. The deflection is not uniform. Informational queries with a clean answer lose the most clicks, while comparison, transactional, and deep how-to pages hold up far better. If your budget is built around ranking number one for definitional terms, the click you used to win is now going to Google's answer box, not your site.
What "traffic deflection" means: how AI Overviews affect organic traffic
Here is the mechanic, with no hand-waving.
For years, the deal with Google was simple. You ranked, you got the click, the click came to your site. AI Overviews break that deal. Google now reads the top results, writes its own answer, and puts that answer above everything else. The user reads it. The question is solved. There is no reason to click.
That is traffic deflection: the search still happens, the demand is still real, but the click that used to land on your page never leaves the results page. The industry word for this is the zero-click search, and AI Overviews poured gasoline on a trend that was already growing. The difference now is that the answer isn't a snippet you wrote; it's a synthesis Google assembled from several sources, often without the user ever registering which sites it came from. (If you want the formal definition of the feature itself, our glossary entry on Google AI Overviews lays it out.)
The part most agencies get wrong is treating this as one big number. "AI Overviews cut traffic by some scary percentage" is the kind of headline stat that sounds alarming and tells you nothing useful. The deflection is uneven across query types, and the uneven part is where the actual strategy lives. A site that's 80% definitional content and a site that's 80% comparison content can sit under the same headline percentage and be having completely different years.
The click-through patterns that matter
Click-through behavior under AI Overviews splits along the intent of the query. Three patterns show up again and again, and they map cleanly to where a search sits in the buying journey.
Pattern 1: Definitional and simple-answer queries get gutted
"What is what you sell." "How many this in another." "What time does what you sell happen." These are queries where the AI Overview is the complete answer. The user needed one fact, the box gave it to them, done. There was never a second step.
If your organic traffic leans on this kind of content, you are the most exposed. The page may still rank in the classic blue links underneath, but the click-through rate on those positions has dropped sharply when an AI Overview sits on top. Public studies put the drop in real numbers: Ahrefs found that the presence of an AI Overview correlates with roughly 58% lower click-through rate for the top-ranking page (Ahrefs, 2025), and Seer Interactive measured a 61% organic CTR drop on queries that show one (Search Engine Land, 2025). You did the work to rank. The answer box collected the click. The cruel part is that the better your answer was, the more likely Google was to lift it into the box and keep the visit for itself.
Pattern 2: Comparison and "best of" queries hold up better
"Product A vs Product B." "Best-in-class for your use case." "Service near me with a requirement." Here the user wants to weigh options, read opinions, and often buy. An AI Overview can summarize the landscape, but it cannot finish the decision for them. People still click through to compare specs, read real reviews, and see live pricing they don't trust a summary to have right.
This is why comparison content and bottom-of-funnel pages have been more durable. The click survives because the answer is not a single fact, it is a judgment the buyer wants to make themselves. It's the same reason our own comparison pages keep earning clicks: nobody outsources a real decision to a paragraph they skimmed.
Pattern 3: Deep, experience-driven, and transactional pages keep their clicks
Original research, first-hand teardowns, calculators, pricing pages, product pages, and genuinely deep how-to guides retain clicks at a much higher rate. The Overview cannot replace a tool you use, a price you pay, or a perspective only a practitioner could write. In a lot of cases the Overview cites these pages, and the citation itself sends qualified traffic, often better-qualified than the generic ranking did, because the reader has already absorbed the basics from the answer and is clicking through for the depth.
The throughline across all three patterns: the thinner and more commodity the content, the harder it gets deflected. The deeper and more decision-shaped it is, the more it survives. That is not a coincidence. It is the whole point now, and it's why "is SEO dead" is the wrong question. Our take on whether SEO is still worth it with AI search goes deeper on that, but the short version: SEO isn't dying, the easy clicks are.
AI Mode raises the stakes on the same mechanic
Google's AI Mode is the more aggressive version of the same idea. Instead of a box above the results, the entire experience becomes a conversation. The user asks, gets an answer, asks a follow-up, gets another answer, and may never see a traditional results page at all.
Two things follow from that. First, the surface area for a classic blue-link click shrinks further, because each follow-up keeps the user inside the conversation instead of bouncing them back to ten organic results. Second, citation placement inside the AI answer becomes the new shelf space. Being the source the model pulls from and names is the position that matters, because in a conversational answer there is no "page two" to climb to later and no second SERP to rank on. You are either named in the answer or you are invisible.
This is not only a Google story. Your buyers are also asking ChatGPT and Perplexity, and the rules differ by engine. Perplexity answers with organic citations and sends a referral when it names you, with no paid placements to buy your way in (it exited advertising in early 2026), so earning the citation is the only lever there. ChatGPT has started running ads, but those appear as labeled sponsored cards alongside or below the generated answer rather than inside it, which means the organic citation inside the answer is still the prize; our breakdown of how ChatGPT ads work for brands covers where the paid layer sits. Different engines, same underlying shift: the answer is the product, and you want to be inside it.
What this means for an SEO-heavy budget
If most of your marketing dollars are pointed at SEO, this is the part to read twice. We are not telling you to cut SEO. SEO is still the foundation that earns the citation in the first place, because the AI answer is built from content that ranks. The move is to change what the SEO budget is aimed at.
Three honest reallocations to consider.
Stop funding content that the answer box now eats. Audit your traffic by query type. If a large share of your organic sessions come from definitional or simple-answer pages, those are depreciating assets. Maintaining a wall of "what is X" posts that get deflected is spending to lose. Redirect that effort. This is also a common (and fixable) cause of the slow bleed people panic about; if your numbers are sliding, our guide to why your website traffic is dropping helps you separate Overview deflection from algorithm updates, seasonality, and self-inflicted technical issues before you reallocate a dollar.
Fund the content that retains clicks. Comparison pages, original data, calculators, pricing transparency, and deep practitioner how-tos. These survive deflection and tend to convert better anyway, because they meet the buyer closer to a decision. This is a better use of the same dollar, not a bigger one.
Add answer engine optimization as its own line, not a rebrand of SEO. Getting cited inside AI Overviews, AI Mode, ChatGPT, and Perplexity is a related but distinct discipline. It rewards clean answer structure, strong entity signals, schema, and content that is genuinely the best source on the question. The practical playbook for the Google side lives in our guide to Google AI Overviews and the tactical companion on how to get cited in Google AI Overviews. The discipline itself is what we run under answer engine optimization, and it sits right alongside the SEO work rather than replacing it.
You also can't reallocate what you can't see. Before you move budget, you need a baseline for how often you're surfacing in AI answers versus how often a competitor is. That measurement is its own thing, and a quick AI visibility check is a reasonable place to start.
The blunt version: a budget that was 90% "rank for keywords" needs to become a budget that ranks, gets cited, and points its content effort at the pages a machine cannot answer for the user. Same total spend, very different target.
Where MoonSauce comes in
If your organic traffic is softening and you are not sure how much of it is AI Overviews versus everything else, that is a fixable question. We will pull your traffic apart by query type, show you which pages are getting deflected and which are holding, and lay out exactly where your current SEO spend is working against you.
No guaranteed-rankings theater, no jargon wall. Just a clear read on where you stand in AI search and what to do about it. Book 30 minutes with us. No pressure and no jargon, just the real picture.
