![\"Get](\"https:\/\/no-cache.hubspot.com\/cta\/default\/53\/9dd5e54b-fbef-4dd0-bc44-1689feb1ea18.png\")
<\/a><\/p>\nAs of December 2025, AI Overviews chop organic click-through rate (CTR) for position-one content by an average of 58%<\/a>, and that\u2019s no coincidence. We\u2019re in the middle of a huge shift in how search engines surface information, and it\u2019s rewriting the rules for marketers and content teams across every industry.<\/p>\n First, Google\u2019s AI Overviews<\/a> are answering queries directly on the results page, intercepting searches that previously drove clicks to websites. And second, a growing portion of searchers are skipping Google entirely and turning to answer engines like ChatGPT and Perplexity for answers.<\/p>\n Both trends slice the traffic search engines<\/a> send to sites, but it\u2019s not gone entirely. I\u2019ve spent the last year navigating the ebbs and flows of traffic with HubSpot, and we\u2019re learning how to balance AI behavior and website traffic expectations. Here\u2019s what businessess need to know.<\/p>\n Table of Contents<\/strong><\/p>\n <\/a> <\/p>\n AI Overviews change how users interact with search results by reducing CTR for some informational queries and redistributing clicks rather than eliminating all website traffic. Simple fact-based queries are more likely to trigger zero-click results, while more detailed, branded questions like comparisons are more likely to earn clicks when users need depth and validation.<\/p>\n Marketers and brands that invest in AEO to help capture AI overviews rather than ignoring them are the brands that will stay competitive. Original research improves citation potential in AI answers, structured data improves machine readability of page content, and concise Q&A sections help answer engines extract and cite content. Tools like HubSpot AEO<\/a> are designed to help marketers operationalize this shift, making it easier to optimize content specifically for AI-generated results and track performance over time.<\/p>\n Learn more about how to improve AI search performance in <\/strong>HubSpot\u2019s free AEO guide.<\/a><\/strong><\/p>\n <\/a> <\/p>\n AI Overviews are generated summaries that appear at the top of Google\u2019s search results, above<\/em> both paid ads and organic listings. When one appears for a brand\u2019s target query, it answers the user\u2019s question, pushing all of the blue links farther down the page.<\/p>\n And marketers all know what happens the farther down they appear on a SERP.<\/p>\n If a brand is the site cited in the overview, impressions stay up (or grow), but clicks drop, and even if its website ranks well, clicks drop because users likely already got their answer in the overview.<\/p>\n In the example below, \u201cWhat is Bollywood?\u201d notice how even big names like Masterclass and popular mediums like YouTube video can be pushed multiple scrolls below the fold.<\/p>\n According to McKinsey<\/a>, half of Google\u2019s results already feature AI-powered features like overviews, and trends predict that number will reach 75% by 2028.<\/p>\n If marketers are looking at their traffic reports and asking, \u201cWhy did my website traffic drop after AI search?\u201d \u2014 this is the \u201czero-click\u201d reality.<\/p>\n A study by Seer Interactive<\/a> found that organic CTR for AI Overview queries dropped by 61% from June 2024 to September 2025. Even more alarming: The CTR of queries without<\/em> AI Overviews also fell by 41% in the same period.<\/p>\n This suggests broader behavioral changes are at play. In other words, users are turning to search engines less frequently as search behavior on social media and answer engines increases.<\/p>\n Pro tip:<\/strong> Use HubSpot\u2019s free AEO Grader<\/a> to check how visible your brand is in AI-powered search engines. For ongoing optimization, HubSpot AEO helps teams continuously improve their visibility and benchmark performance against competitors.<\/p>\n <\/a> <\/p>\n The problem of how to measure AI Overviews\u2019 impact on web traffic is real. Google Search Console<\/a> currently does not offer a direct way to isolate or filter data for AI Overviews (AIO).<\/p>\n All performance metrics from AI Overviews are aggregated with standard web search data. For instance, when a brand\u2019s content is cited in an AI Overview, Search Console doesn\u2019t tell them. Their impressions and clicks are logged, but merged with everything else.<\/p>\n HubSpot recently added \u201cAI Referrals\u201d<\/a> to its list of traffic sources which refers to AI assistants and chatbots like ChatGPT, Claude, and Perplexity. It also includes visitors who click links provided in AI-generated responses. In Marketing Hub Pro and Enterprise<\/a>, AEO features give teams a more direct view into how their content performs in AI-powered search, helping track visibility and identify optimization opportunities across answer engines.<\/p>\n Marketing and content teams can also make educated predictions with third-party data. For example, Ahrefs provides estimates on which keywords have AI Overviews, whether a brand was cited, and how much traffic that equates to, approximately.<\/p>\n Source<\/em><\/a><\/p>\n I spoke with Amanda Sellers, HubSpot\u2019s blog growth manager, about the best ways to forecast traffic under AI overviews.<\/p>\n She recommends using linear regression, a mathematical method that uses past data to simulate a trend into the future. A linear regression assumes that nothing big \u2014 like an algorithm update or increase in SERP features like AI Overviews \u2014 will disrupt that trend.<\/p>\n \u201cYou and I both know that Google likes to throw a wrench into things,\u201d explains Sellers.<\/p>\n \u201cAt one point, AI Overviews showed up for less than 10% of the HubSpot blog\u2019s keywords, most of them being informational definition intent. Today, nearly 50% of the keywords the HubSpot blog ranks for have an AI Overview at the top.\u201d<\/p>\n For this reason, Sellers frequently checks AI Overview exposure in Ahrefs and performs CTR curve analysis using data from Google Search Console. That way, multiple scenarios can be forecasted on top of the baseline linear regression, such as \u201cwhat if AI Overviews increase by 20%\u201d or \u201cwhat if we get impacted negatively by an algorithm update.\u201d<\/p>\n Linear regressions also allow marketers to quantify seasonal changes, determining patterns in historical data.<\/p>\n For example, there might be a historical pattern of low traffic in December compared to November due to holiday seasonality. A linear regression can help marketers and SEO strategists create seasonality modifiers that adjust the traffic baseline according to the average pattern.<\/p>\n Sellers continues, \u201cIf we take the baseline traffic, December usually lands 65% below the baseline because fewer people are searching. January tends to be one of our stronger months at around 135% above the baseline. Adding these fluctuations into our model can help us understand if there is unexpected performance in one direction or another.\u201d<\/p>\n If a traffic forecast already factored in seasonality<\/a> in this way, any performance anomalies in one way or another would mean seasonality is not the culprit. From there, an SEO strategist can use Ahrefs to determine whether Google increased the visibility of AIOs or whether another factor was at play. However, it\u2019s not always that simple.<\/p>\n \u201cKeywords rise and fall, AIOs appear and disappear, algorithm updates come and go\u2026 and there are internal technical factors that can impact performance. In reality, attributing performance is so much more complex.\u201d<\/p>\n For instance, after a particularly tough algorithm update, Sellers found 46.7% of a subsection of HubSpot\u2019s keywords lost positioning and <\/em>gained an AI Overview. It\u2019s much more difficult to attribute how much of the performance change was the AI Overview siphoning traffic versus a decrease in CTR from simply a lower SERP position.<\/p>\n For this reason, it\u2019s best to let the data speak for itself. Sellers split the keywords into different buckets:<\/p>\n By comparing the performance of these buckets against each other and swapping CTRs, Sellers was able to get an estimate of how much performance change came from positioning changes versus AIOs.<\/p>\n (Spoiler alert: AIOs were the much bigger culprit.)<\/p>\n By comparing, Sellers found that even keywords where HubSpot didn\u2019t lose positioning still had significant CTR losses. This means there was less traffic, even when we were performing well. Meanwhile, by swapping CTAs and multiplying by impressions, we could estimate the traffic decline.<\/p>\n <\/a> <\/p>\n Not all queries are affected by AI Overviews. Thankfully, the data is becoming clearer about which types feel the greatest zero-click impact and which can still drive website traffic for businesses.<\/p>\n In 2025, Semrush reported that nearly 95%<\/a> of keywords triggering AI Overviews have little to no paid ads or commercial value. In other words, Google seems to be deploying AI summaries mainly for informational searches, with transactional content (i.e., pricing pages, demo pages) staying in the traditional SERP format.<\/p>\n That means the website traffic most at risk is top-of-funnel educational content that typically grabs a lot of clicks for businesses and builds brand awareness.<\/p>\n Think simple right-or-wrong lookups (\u201cwhat is [concept],\u201d \u201chow to\u201d explainers, definition queries, and single-source informational questions), like this example: \u201cWho is Shahrukh Khan?\u201d<\/p>\n This question is answered by Google in an AI overview so there\u2019s less need to continue on to the other results.<\/p>\n The same study found that transactional keywords like \u201cbuy,\u201d \u201ccompare,\u201d and \u201cnear me\u201d tend to have higher CTRs because AI typically doesn\u2019t complete transactions. Continuing our example, look at the results of \u201cBuy Shahrukh Khan DVD.\u201d (A DVD, for my younger folks, is a \u201cdigital video disc,\u201d what we used to watch movies before streaming.)<\/p>\n Comparison queries like \u201cX vs. Y for [use case]\u201d also continue to drive clicks, because users want depth and validation that a two-paragraph AI summary can\u2019t fully provide. The same is true for queries that require local, real-time, or highly specific information.<\/p>\n Overall, the best content for generating clicks and website traffic is currently bottom-funnel content (pricing pages, comparison guides, case studies), local service queries, niche technical queries, and original research that AI can\u2019t synthesize from elsewhere.<\/p>\n <\/a> <\/p>\n Ok, so here\u2019s where the picture shifts from bleak to nuanced: Being cited in an AI Overview may slash a brand\u2019s top-of-the-funnel, awareness website traffic, but those who do<\/em> visit are arguably more qualified.<\/p>\n A study from Dataslayer<\/a> found that brands cited in AI Overviews earn 35% more organic clicks and 91% more paid clicks than brands not cited in the same queries. Whether this is due to greater awareness or other factors is hard to say, but it\u2019s still encouraging.<\/p>\n Sure, a brand can\u2019t control whether an AI Overview appears for their target query, but they can work to earn the citation when it does.<\/p>\n To improve a brand\u2019s chances of securing AI overviews, marketers need to learn how to write for AI search<\/a> and invest in answer engine optimization<\/a> (AEO). Here\u2019s what that entails:<\/p>\n HubSpot Content Hub<\/a> can help content writers templatize these patterns and schema, streamline content briefs, and maintain editorial governance at scale as their team produces more AEO-optimized content. HubSpot\u2019s AEO capabilities in Marketing Hub Pro and Enterprise<\/a> extend this by helping teams see their performance.<\/p>\n Even Google aside, a growing share of users are starting their search journey with AI through ChatGPT, Perplexity, or other answer engines.<\/p>\n BrightLocal research shows that Google still drives 61% of all general searches<\/a>, but more importantly, AI referral traffic tends to convert at a dramatically higher rate<\/a>.<\/p>\n To earn that high-intent traffic, teams need Answer Engine Optimization (AEO)<\/a>:<\/p>\n HubSpot AEO<\/a> helps marketers track AI citation performance and optimize content for visibility across answer engines \u2014 giving teams insight into a channel that traditional analytics platforms still struggle to measure.<\/p>\n <\/a> <\/p>\n Google Search Console does not surface this natively, and other tools group things into a general \u201cAI referral\u201d bucket.<\/p>\n The best approach is to manually search a brand\u2019s top target queries in an incognito browser and note whether its site appears as a cited source in the AI Overview. Then, use a linear regression to simulate a trend into the future.<\/p>\n For systematic tracking at scale, third-party tools like Semrush, Ahrefs, and Authoritas can monitor which of a business\u2019s URLs appear in AI Overviews and track citation frequency over time. HubSpot AEO also helps teams monitor their presence in AI-generated results and uncover which pages are earning visibility, making it easier to prioritize optimization efforts.<\/p>\n Yes, significantly. Non-branded informational queries are where AI Overviews most commonly appear and where CTR losses are steepest. Branded traffic tends to be more resilient because navigational and branded queries trigger AI Overviews at a lower rate.<\/p>\n Try using Google Search Console\u2019s new branded\/non-branded filter<\/a> to track both segments independently.<\/p>\n Partially, but don\u2019t abandon informational content entirely. Factual, educational content is still valuable for building topical authority and earning AI citations. But marketers should rebalance their investment toward comparison content, bottom-funnel queries, and original research that AI can\u2019t fully synthesize.<\/p>\n The goal is to be the source AI cites, not to avoid the queries AI covers. Shift success metrics from pure click volume to share of voice, citation frequency, and branded search growth.<\/p>\n At the risk of sounding dramatic: now<\/em>. If more than 50% of a business\u2019s web traffic currently comes from non-branded organic search, it is overexposed.<\/p>\n Email lists, communities, newsletters, and direct audience relationships are immune to AI Overview cannibalization, algorithm updates, or shifts in Google\u2019s rendering. The value of owning an audience compounds over time; it\u2019s the one distribution channel where a business\u2019s results are entirely its own.<\/p>\n Publishers with high branded and direct traffic, like the Daily Mail (whose over 60% of traffic<\/a> is direct) have proven significantly more resilient to AI Overview disruption than sites reliant on non-branded organic search.<\/p>\n\n
Is AI killing web traffic?<\/h2>\n
What AI Overviews Change on the SERP<\/h2>\n
![\"ai](\"http:\/\/fliegewiese.org\/wp-content\/uploads\/2026\/04\/ai20overviews20answering20questions20like20E2809Cwhat20is20bollywoodE2809D20are20potentially20killing20website20traffic.png\")
<\/p>\nHow to Measure AI Overviews\u2019 Impact on Your Traffic<\/h2>\n
![\"is](\"http:\/\/fliegewiese.org\/wp-content\/uploads\/2026\/04\/is20killing20website20traffic3B20ai20referrals20as20traffic20source20in20hubspot.png\")
<\/p>\n![\"is](\"http:\/\/fliegewiese.org\/wp-content\/uploads\/2026\/04\/is20killing20website20traffic3B20ai20overviews20being20tracked20in20ahrefs-1.jpg\")
<\/p>\nWhat is the best way to forecast traffic under AI Overviews?<\/h3>\n
How do you attribute changes to AI Overviews vs seasonality?<\/h3>\n
\n
Is AI killing web traffic more for certain queries?<\/h2>\n
Queries Most Vulnerable to Zero-Click<\/h3>\n
![\"is](\"http:\/\/fliegewiese.org\/wp-content\/uploads\/2026\/04\/is20killing20website20traffic3B20ai20overviews20being20tracked20in20ahrefs.png\")
<\/p>\nQueries That Still Earn the Click<\/h3>\n
![\"is](\"http:\/\/fliegewiese.org\/wp-content\/uploads\/2026\/04\/is20killing20website20traffic3B20who20is20shahrukh20khan20answered20in20ai20overview.png\")
<\/p>\nIs AI killing web traffic, or do you get traffic from AI citations?<\/h2>\n
Optimizing for AI Overviews<\/h3>\n
\n
Optimizing for Answer Engines (AEO)<\/strong><\/h3>\n
\n
FAQs About AI Overviews and Web Traffic<\/h2>\n
How can I tell if my pages are being used as sources in AI Overviews?<\/h3>\n
Do AI Overviews affect branded and non-branded traffic differently?<\/h3>\n
Should I change my keyword strategy because of AI Overviews?<\/h3>\n
When should you shift budget toward owned channels?<\/h3>\n
![\"\"](\"https:\/\/track.hubspot.com\/__ptq.gif?a=53&k=14&r=https%3A%2F%2Fblog.hubspot.com%2Fmarketing%2Fis-ai-killing-web-traffic&bu=https%253A%252F%252Fblog.hubspot.com%252Fmarketing&bvt=rss\")
<\/p>\n","protected":false},"excerpt":{"rendered":"