Author name: aftabkhannewemail@gmail.com

Understanding the Shift in Google Search Google has fundamentally transformed the search experience with the roll-out of AI Overviews, formerly known as the Search Generative Experience (SGE). Powered by advanced large language models, including Google’s Gemini, AI Overviews aim to provide users with direct, synthesized answers to complex queries right at the top of the search engine results page (SERP). However, as search engine optimization (SEO) professionals and digital marketers analyze the behavior of these AI-generated summaries, a clear pattern has emerged: AI Overview data is not uniform. In fact, it looks vastly different when analyzing commercial queries compared to informational or transactional searches. For organizations relying on organic search traffic to drive leads and sales, understanding these discrepancies is critical. Tracking tools and industry studies often report conflicting statistics regarding how frequently AI Overviews appear, which sites they link to, and how much space they occupy on the screen. The reality is that AI Overview tracking can tell very different stories depending on the prompts, query types, and specific geographic markets included in the analysis. To build a resilient SEO strategy, marketers must look beyond aggregate data and analyze how Google’s AI handles commercial search intent. The Diversity of Search Intent and the AI Overview To understand why AI Overview data fluctuates so dramatically, it is necessary to examine how Google categorizes search queries. Traditional SEO breaks search intent down into four primary categories: Informational: Queries where the user wants to learn something (e.g., “how does photosynthesis work”). Navigational: Queries where the user is looking for a specific website (e.g., “Netflix login”). Commercial: Queries where the user is researching products, services, or brands with the intention of buying in the future (e.g., “best enterprise CRM software” or “top-rated running shoes”). Transactional: Queries where the user is ready to make an immediate purchase (e.g., “buy iPhone 15 Pro Max online”). Google’s AI engine handles these intents differently. Informational queries are highly conducive to text-heavy AI Overviews that synthesize definitions, history, and step-by-step guides. Commercial queries, however, present a unique challenge and opportunity for Google. These searches involve high-value intent, where users are actively comparing options. Consequently, the AI Overviews generated for commercial queries are often highly structured, featuring comparison tables, product carousels, pricing information, and pros and cons lists. Because the layout and sourcing of these overviews are so complex, the underlying data tracked by SEO platforms differs wildly from informational benchmarks. Why AI Overview Tracking Data Varies Across Tools Many SEO professionals have noticed that prominent search tracking tools—such as Semrush, BrightEdge, Moz, and others—frequently publish conflicting data regarding AI Overview penetration. One tool might report that AI Overviews appear on 15% of all searches, while another claims the number is closer to 40% or even higher. This divergence is not due to inaccurate tracking, but rather to the composition of the keyword datasets being monitored. Keyword Sample Bias If a tracking tool monitors a keyword set heavily weighted toward conversational, long-tail, informational queries, it will naturally report a much higher frequency of AI Overviews. Google’s AI is highly active in answering “how-to” questions and explaining complex concepts. Conversely, if a tracking tool focuses on head terms, branded keywords, or purely transactional product searches, the trigger rate for AI Overviews will appear much lower. Because commercial queries sit right in the middle—requiring both synthesis of information and product listings—the trigger rates for these terms are highly sensitive to algorithmic tweaks. The Impact of Search Prompts and Conversational Language The phrasing of a search query significantly affects whether an AI Overview is displayed. Traditional, short-form keyword searches (e.g., “running shoes”) may return standard search results dominated by Google Shopping ads and organic e-commerce category pages. However, if the user inputs a conversational prompt (e.g., “what are the best running shoes for someone with high arches who runs on concrete?”), Google’s AI is far more likely to generate a custom overview. Tracking systems that only monitor traditional keywords will miss the massive footprint of AI Overviews generated by these longer, conversational queries. Geographic and Market Differences Google does not roll out AI features globally in a single day. New updates, UI elements, and algorithmic thresholds are tested extensively in specific markets, primarily the United States, before being expanded to other regions like the United Kingdom, Canada, or Australia. Additionally, regulatory environments—such as the Digital Markets Act (DMA) in the European Union—impact how Google can present search results and integrate its own services. Consequently, tracking data for commercial queries in the US will show a much higher integration of Google Merchant Center data and interactive shopping modules compared to data tracked in European markets. Anatomy of an AI Overview for Commercial Queries When Google does trigger an AI Overview for a commercial query, the presentation is drastically different from a standard text response. Marketers must understand these unique structural elements to optimize their content effectively. Integration of Google Merchant Center and Product Feeds For commercial product searches, Google often pulls real-time inventory, pricing, images, and reviews directly from the Google Merchant Center. Instead of simply citing articles that list “the best products,” the AI Overview may construct an interactive product grid. This means that having a highly optimized website is no longer the only requirement for visibility; businesses must also maintain accurate, up-to-date product feeds within Google’s shopping ecosystem to be featured in these AI-driven comparison blocks. The Sourcing of Recommendations Unlike informational AI Overviews, which frequently source data from authoritative educational sites, wikis, and top-tier news publications, commercial AI Overviews rely heavily on user-generated content and independent review sites. Google looks for authentic, experiential data. It synthesizes consensus opinions from platforms like Reddit, Quora, and specialized forums, alongside editorial reviews from trusted publishers. As a result, commercial tracking data shows a highly diversified set of linked sources, making digital PR and forum-based brand sentiment more important than ever. Interactive Comparison Modules Commercial searchers want to compare features, pros and cons, and pricing. To satisfy this intent, Google’s AI frequently generates

Placement exclusions on the Google Display Network (GDN) have long been treated as basic account hygiene. For years, media buyers approached them with a simple checklist mindset: identify spammy domains, flag low-converting pages, and block poor-quality inventory to preserve budget and protect brand safety. It was a digital defense strategy designed to keep your banner ads away from clickbait forums, low-tier mobile apps, and controversial content. However, the rapid expansion of automated bidding, artificial intelligence, and broad matching algorithms has transformed the role of display exclusions. In today’s advertising ecosystem, exclusions do much more than just block bad placements. They serve as critical training signals that guide Google’s machine learning models, helping the algorithm understand where to look—and, more importantly, where not to look—for high-intent buyers. To maximize the return on your ad spend (ROAS) in an AI-driven environment, digital marketers must rethink how they approach placement exclusions. Shifting from a purely defensive hygiene tactic to a strategic data-sculpting method will help you steering automated campaigns toward high-quality conversions while avoiding budget-draining learning loops. The legacy blueprint: Hygiene and budget conservation To understand the strategic shift taking place today, we must first look at why blocking placements mattered in traditional PPC. Historically, placement exclusions served two primary business functions: protecting brand integrity and conserving financial resources. Brand safety and alignment No brand wants its messaging displayed alongside extreme political rants, adult content, pirated media, or sensationalist clickbait. In the early days of programmatic advertising, brand safety was a manual battle. Advertisers spent hours analyzing where their banner ads appeared, manually adding offensive or irrelevant sites to shared exclusion lists to prevent brand dilution or public relations issues. Direct cost control The Google Display Network spans more than two million websites, videos, and mobile apps, reaching over 90% of global internet users. While this scale is impressive, a massive share of this inventory consists of high-click, zero-conversion zones. Classic examples include flashlight apps, utility tools, and children’s mobile games. In these spaces, users often click on banner ads by accident while trying to navigate the app’s interface. Even premium, highly reputable publications like The New York Times, CNN, or major financial portals can become budget killers for direct-response advertisers. While these sites offer brand-safe environments and high-quality traffic, they often carry high cost-per-click (CPC) rates. For a business focused on immediate sales or lead generation rather than broad brand awareness, a single premium placement can consume thousands of dollars in ad spend with very little conversion intent behind the traffic. The traditional, static approach The traditional solution to these issues was simple but labor-intensive. Digital marketers built massive, static master lists of 70,000+ excluded URLs, blocked all mobile app categories entirely, and pulled “Where Ads Showed” reports every week or month to manually eliminate outlier placements. While these legacy tactics are still necessary as foundational account hygiene, they only scratch the surface of how modern, AI-powered advertising platforms process data. How AI changed the rules of the GDN In modern Google Ads setups, machine learning handles the heavy lifting of audience targeting and bidding. Smart Bidding algorithms—such as Target Cost Per Acquisition (tCPA) and Target Return on Ad Spend (tROAS)—are built to find customers within your target parameters at a predictable cost. When you combine these automated bidding strategies with broad targeting or optimized targeting, Google’s AI does not just passively wait for your instructions. Instead, it actively hunts for positive user signals across the web. The algorithm constantly analyzes who clicks your ads, who converts on your landing pages, and where those actions take place. It then builds complex predictive models to identify and target placements that match those successful behaviors. This machine learning feedback loop is incredibly powerful when fueled by accurate data, but it can quickly backfire when bad data enters the system. If your display campaigns do not have clear strategic guardrails, Google’s AI will naturally gravitate toward the cheapest and highest-volume inventory available to test its hypotheses. For example, a flood of accidental clicks from mobile puzzle games or low-quality click-fraud sites can initially look like highly positive signals to the algorithm because of their high click-through rates (CTR) and low CPCs. Believing it has found a goldmine of engaged users, the Smart Bidding algorithm may double down on these low-quality placements, consuming your daily budget before discovering that none of these clicks lead to actual sales or qualified leads. By the time the AI realizes these placements are underperforming, your budget for the month is already gone, and the machine learning model has been trained on low-value data. This dynamic shifts the purpose of Google Ads placements from a simple list of where your ads can show to a critical set of guardrails that define the boundaries of your AI’s sandbox. Moving from hygiene to strategy: Guardrails for the algorithm Strategic exclusions are no longer just about deciding where your ads should not appear. They are about guiding the automated engine away from low-quality data pools and toward high-intent traffic sources. By proactively shaping the environments where Google’s AI is allowed to operate, you inject human intent, business context, and strategic direction back into automated campaigns. Campaign intent mapping Instead of applying one generic, account-level exclusion list to every single campaign, you should use tailored exclusions to match the specific strategic intent of each campaign: Top-of-funnel brand awareness campaigns: For these campaigns, keep premium placements like major news outlets, industry-leading publications, and popular media sites active. Exclude niche directories, forums, and low-quality blogs. This pushes the AI to focus its budget on high-visibility, highly reputable environments that build long-term brand equity. Bottom-of-funnel direct-response campaigns: For campaigns focused on immediate sales or lead generation, take the opposite approach. Exclude broad-reach, high-cost premium sites that consume large portions of your budget without driving immediate action. Force the machine learning model to focus on long-tail, content-rich blogs, product review sites, and highly specific niche pages where users are actively researching products with strong purchase intent. Preempting Smart

The Hidden Cost of Marketing Silos In the vast majority of digital marketing organizations, search engine optimization (SEO) and affiliate marketing operate in completely separate universes. The SEO team is laser-focused on keyword research, technical site health, on-page optimization, and capturing organic search real estate. Meanwhile, the affiliate marketing team focuses on nurturing partner relationships, recruiting new publishers, negotiating commission rates, and tracking conversion links. Because these channels are treated as distinct budget lines with separate key performance indicators (KPIs), these two teams rarely share information, let alone coordinate on a unified search strategy. This lack of alignment is a major missed opportunity. When SEO and affiliate teams operate in silos, they don’t just miss out on potential growth—they often actively work against one another, driving up customer acquisition costs and diluting brand equity. Breaking down these internal walls is crucial for sustainable business growth. Stepping out of the isolated “SEO bubble” allows search professionals to align their efforts with broader corporate initiatives. By collaborating with affiliate teams, SEOs can gain a clearer understanding of what business success looks like across the entire conversion funnel. Conversely, the affiliate team can leverage deep search data to make more informed partnership decisions. Closer integration between these two channels helps protect your brand, improves visibility in artificial intelligence and large language model (LLM) search engines, resolves technical indexation issues, and maximizes total profit margins. Protecting Brand Search Terms and Reclaiming Your Revenue One of the most immediate financial hazards of a disjointed marketing strategy is the cannibalization of branded search traffic. When third-party affiliate sites rank for search queries directly tied to your brand name, they intercept high-intent customers who were already searching for your products. This results in the business paying commission fees for traffic and sales that should have been captured organically at zero marginal cost. To protect your bottom line, the SEO team must take ownership of any search query that impacts the company’s organic performance. High-intent, transactional branded keywords are particularly vulnerable to being hijacked by external affiliate sites. These terms typically involve consumer searches for deals, such as: [Your Brand Name] + discount code [Your Brand Name] + promo code [Your Brand Name] + coupon [Your Brand Name] + vouchers When a customer reaches the checkout page on your site, sees a discount code field, and opens a new tab to search for a promo code, they are already convinced and ready to buy. If an affiliate website ranks first for that search query, the customer clicks their link, grabs a code, and returns to complete the purchase. The brand then pays a hefty commission to the affiliate publisher for a conversion that was already secured. If your internal SEO team successfully ranks your own website for these queries, you capture that customer directly, saving substantial affiliate payouts and preserving your overall profit margins. Case Study: Trainline and the Opportunity of Branded Promo Codes To understand the real-world impact of this phenomenon, consider the digital ticketing platform Trainline. In the United Kingdom, the search query “trainline promo code” attracts approximately 17,000 monthly searches. This represents a massive pool of consumers who are on the verge of purchasing a train ticket. While Trainline has created a dedicated landing page specifically designed to target promo and discount codes, the page has historically suffered from suboptimal on-page SEO. Because the copy, meta titles, and heading tags were not fully aligned with search intent, the page’s rankings fluctuated significantly. This allowed external voucher and coupon code directories to rank above Trainline’s own landing page. By failing to consistently secure the top organic spot for its own branded search term, the brand has repeatedly lost high-intent traffic to third-party affiliates, resulting in unnecessary affiliate commissions. The solution to this problem does not require a massive development project. By executing basic on-page adjustments—such as optimizing the H1 tags, updating meta titles to match user intent, and adding relevant, high-quality body copy to the page—brands can reclaim these positions from their own affiliates. Securing Share of Voice (SOV) Success Reclaiming search positions yields dramatic results. In a similar case involving an enterprise brand’s discount page, the brand’s share of voice (SOV) for high-intent branded voucher queries had plummeted due to aggressive competition from their own affiliate network partners. The SEO team implemented a strategic content update to align the page directly with the search intent of users looking for active, verified promo codes. Within 24 hours of publishing the optimized updates, the brand’s Share of Voice for these competitive queries surged from 14% to 31%. This single intervention had a massive compounding effect on the entire business: Increased Organic Revenue: Direct, non-paid channels captured a higher percentage of bottom-of-funnel transactions. Reduced Affiliate Payouts: The business stopped paying unnecessary commissions on transactions that occurred within a single user session. Improved Company Profitability: Overall marketing margins improved, demonstrating that SEO insights can directly drive corporate financial success. The New Search Landscape: Driving LLM and AI Visibility The search landscape is undergoing a massive shift. Traditional search engines are increasingly integrating generative AI features, and consumers are turning directly to Large Language Models (LLMs) like ChatGPT, Claude, and Google Gemini to research products and make purchasing decisions. This shift requires brands to rethink how they establish authority online. Unlike traditional algorithms that rely heavily on backlink profiles and technical on-page optimizations, LLMs build their recommendation models by synthesizing vast quantities of web data, looking specifically for consistent reputational signals across authoritative sources. The Power of Reputational Signals in AI Models When an LLM answers queries like “What is the best CRM software for small businesses?” or “What are the top-rated running shoes for marathon training?”, it does not simply pull a random URL from its index. It analyzes patterns in how brands are discussed across the web. If your brand is consistently mentioned, reviewed, and recommended across dozens of reputable third-party platforms, generative AI models recognize your brand as an industry leader. These repeated associations act as

The New Mental Model of Search Intent For more than two decades, search engine optimization operated on a fundamental, predictable law: user intent dictates user behavior on the search engine results page (SERP). If a user typed a brand name into Google, they were executing a navigational search. They knew exactly where they wanted to go, resulting in a rapid, friction-free exit from the SERP. If they typed a broad query like “how to repair a running toilet,” they were in informational mode, settling in for a slower, multi-page discovery process. That paradigm has officially broken. The catalyst for this shift is the widespread integration of Google’s AI Overviews (AIO). By rendering a cohesive, synthesized block of text at the very top of the organic search results, Google has fundamentally altered how users interact with information. Instead of treating the SERP as a mere launchpad of links, users are now treating it as the destination itself. Search has transformed into a reading session. To understand the depth of this transformation, Eric Van Buskirk of Clickstream Solutions analyzed anonymized clickstream data from approximately 846,000 U.S.-based Google search sessions. The findings reveal a stark divergence in user behavior depending on whether an AI Overview is present on the SERP. The Flattening and Compression of Intent In a traditional search environment without an AI Overview, time-on-page scales predictably with user intent. The differences between how different searchers interact with the page are distinct: Navigational searchers are highly efficient. After 21 seconds, only 12% of these users remain on the search results page. They find their target link immediately and click away. Local searchers linger much longer. Because local results are densely packed with maps, reviews, operational hours, and address details, 32% of these searchers are still evaluating their options on the SERP after 21 seconds. Informational searchers sit comfortably in the middle, scanning organic snippets before committing to a click. However, when an AI Overview is present, this behavioral spread completely collapses. The distinctive signatures of different search intents disappear, compressing into a tight, uniform cluster. With an AIO active on the page, the percentage of users still on the SERP after 21 seconds across all five primary search intents—informational, local, navigational, transactional, and video—concentrates remarkably between 41.9% and 48.5%. The variation between a fast-paced navigational search and a complex local search shrinks to a mere six percentage points. This means that search sessions on pages featuring an AI Overview are, on average, nearly four times longer for quick-intent queries. The presence of the AI block arrests the user’s journey, turning passive scrollers into active readers regardless of their initial goal. Why Searchers Stay: The Grounding of Answers This dramatic expansion of SERP dwell time is driven by the density of the information provided. Rather than forcing users to click through to three different websites to compare facts, Google’s AI Overview performs the heavy lifting of aggregation, synthesis, and summarization directly on the search page. The searcher stays because they are reading a custom-generated answer. This shift from indexing pages to synthesizing answers represents a structural evolution in how search engines work. Bing outlined this transition clearly in an industry publication titled “Evolving the role of the index: From ranking pages to supporting answers.” As search engines evolve from simple indexers to proactive answer engines, the core engineering constraint changes: “Grounding an AI-generated answer introduces a fundamentally different constraint: The system is no longer just pointing to information, it is using it. The goal shifts from ‘fetch the best documents’ to ‘fetch the best information to synthesize into a reliable, verifiable answer.’” In the classic search model, search engines bore little responsibility for the accuracy of individual landing pages. They merely provided a ranked list of “ten blue links” and monitored user clicks to refine those rankings. If a user clicked a link and found poor information, they bounced back to the search page, signaling to the algorithm that the destination was low quality. Under the new AI-driven model, the responsibility for accuracy shifts to the search engine. To generate a synthesized answer, the engine must extract factual statements from underlying web documents, verify their accuracy against known entities, and present them cohesively. The engine is no longer just a signpost; it is an author. Consequently, the user spends their time scrutinizing the engine’s output before—or instead of—navigating to an external source. This behavioral change is not opt-in. Most Google users are not actively seeking out AI tools; they are simply using the default search interface as they have for decades. As Google continuously rolls out these formats, users are guided into AI-centric reading patterns. This passive, forced adoption has met some resistance, as evidenced by a 30% surge in installations of privacy-focused alternatives like DuckDuckGo. Nevertheless, with Google reporting that over 1.5 billion people actively interact with AI Overviews, this is the standard operating environment for modern organic search. Winning the “Second Impression” If users are spending more time reading the SERP, how do brands secure their attention and drive traffic to their websites? The answer lies in optimizing for the “second impression.” The second impression occurs during the back-scroll. When a user lands on a SERP containing an AIO, they naturally focus on the large AI-generated block at the top. They read the summary, absorb the primary takeaways, and then begin scrolling down to view the traditional organic results. If they do not find an immediate click, or if they seek to verify the AIO’s assertions, they scroll back up. This second pass is the equivalent of a consumer re-evaluating options on a retail shelf. It is the moment where credibility, structural clarity, and visual cues determine which link earns the click. To win this crucial micro-moment, websites must optimize their search listings based on the specific page templates they are trying to rank. The strategy requires a tailored approach across three primary page types: Product Detail Pages (PDP), Category Detail Pages (CDP), and Blog Content. Product Detail Pages (PDP)

For several years, the search marketing industry has been operating in a state of high anxiety. The introduction of generative artificial intelligence into search engines—most notably Google’s AI Overviews and AI Mode—fundamentally transformed the traditional search landscape. Suddenly, answers that used to require a click to a publisher’s website were displayed directly on the search engine results page (SERP). Publishers and SEO professionals found themselves facing a double-edged sword: they wanted the visibility that came with being cited in AI-generated answers, but they feared the loss of traffic from zero-click searches, and they lacked any real way to measure or control how their content was being used. Google is finally addressing these dual concerns with two major updates rolling out in Google Search Console (GSC). First, the search giant is introducing a dedicated Search Generative AI performance report, designed to offer publishers visibility into how often their content appears in AI-driven search experiences. Second, Google is testing a new direct control—a simple “toggle” within Search Console—that allows website owners to block their content from appearing in generative AI search features entirely. These features represent a massive shift in how search engines negotiate value with content creators. However, like many Google rollouts, the details reveal a complex compromise between publisher demands, regulatory pressure, and Google’s own technological ambitions. The Search Generative AI Performance Report: What Data Do We Get? For a long time, tracking the impact of Google’s AI search features was a guessing game. SEOs relied on third-party tracking tools, manual searches, and anecdotal evidence to see if their pages were being cited in AI Overviews. With the new Search Generative AI performance report, Google is bringing native data to Search Console, as many in the industry had expected. According to Google’s official announcement, these new insights are designed to help website owners evaluate their footprint in generative search features. The report tracks the appearance of website pages across generative AI features in both Search and Discover. As detailed on the Google Search Central Blog, the metrics included in this new report cover several essential data points: Impressions: This metric measures how often URLs from your website appeared as citations, sources, or links within generative AI features in Search and Discover. Pages: A breakdown of the specific URLs that Google’s AI models chose to include inside generative search responses. This is highly valuable for understanding which pieces of content Google considers authoritative enough to ground its AI models. Countries: Geographic data showing where your generative AI visibility is strongest, allowing for regional analysis. Devices: Insights into whether users are seeing your content in AI responses on mobile devices or desktop computers (currently available for Search results). Dates: Over-time tracking that allows webmasters to monitor their AI-driven visibility trends on an hourly, daily, weekly, or monthly basis. For webmasters who want to dive deeper into the technical implementation and categorization of these metrics, Google has published a comprehensive help center document detailing how the data is aggregated and displayed. The Elephant in the Reporting Room: No Click Data While the introduction of an AI performance report is a step forward, it comes with a massive, highly controversial catch: the report does not include click data. For digital publishers, marketers, and SEO analysts, impressions tell only half the story. Knowing that your content was displayed in an AI Overview is useful, but without click metrics, it is impossible to calculate CTR (click-through rate) or understand the direct financial impact of being featured in AI search. Without clicks, webmasters cannot easily prove the ROI of creating content optimized for AI grounding. When asked directly about the absence of click data, a Google spokesperson gave a standard, forward-looking statement: “We’re continuing to work with website owners to understand what insights will be most helpful to inform their strategies, and we’ll introduce additional metrics over time.” This omission has not surprised industry veterans. Google has historically been protective of granular click-through data when introducing new search features, often citing privacy concerns or technical limitations. Furthermore, showing publishers the exact click-through rates of AI Overviews might confirm their worst fears: that generative AI is indeed cannibalizing organic traffic. For now, SEOs will have to settle for impression data and use advanced internal analytics and referral traffic tracking to try and piece together the rest of the puzzle. The AI Blocking Control: A Toggle to Opt Out of Generative Search Alongside the performance report, Google is testing a groundbreaking new control panel inside Google Search Console. This feature comes in the form of a simple toggle that allows website owners to opt out of having their content used in Google’s generative AI features, including AI Overviews, AI Mode, and AI Overviews in Discover. According to Google, this control is designed to put choice back into the hands of publishers: “website owners can decide if they want their site to appear in and help ground responses in our generative AI Search features.” For publishers who choose to toggle this feature off, the consequences are straightforward. Google noted that “sites that opt out will not receive traffic or impressions from our generative AI features.” Crucially, Google has explicitly confirmed that toggling off your content for AI features will not be used as a ranking signal for standard search results. If you opt out of AI Overviews, your site should still rank normally in the core organic search listings. This is a critical distinction, as publishers feared they would be penalized across the entire Google ecosystem if they refused to participate in Google’s AI ambitions. Why Would a Publisher Choose to Opt Out? While visibility in search is usually the ultimate goal of SEO, generative AI presents a unique dilemma. When Google uses a publisher’s content to generate a direct answer, the searcher has less incentive to click through to the actual website. The publisher bears the cost of creating, hosting, and researching the content, while Google captures the user’s attention and keeps them on the search page. Prior

Pay-per-click (PPC) advertising is one of the most effective ways to drive targeted traffic to a business, but it has always carried an inherent financial risk: invalid traffic. From competitor click fraud and malicious botnets to accidental double-taps on mobile devices, advertisers have long battled the reality of paying for engagements that have zero chance of converting. To address these concerns and provide more transparency, Google has published new documentation highlighting its Invalid Activity Credit Report. This move brings renewed attention to an incredibly valuable, yet often underutilized, tool designed to help digital marketers track refunds issued for invalid clicks and campaign interactions. This documentation offers a clearer view of how Google credits back ad spend across Search and Performance Max (PMax) campaigns, providing advertisers with the data they need to audit their traffic quality and verify Google’s internal fraud prevention systems. Understanding Invalid Traffic (IVT) in Google Ads Before diving into the specifics of the newly highlighted report, it is essential to understand what Google classifies as “invalid activity.” In the world of digital advertising, not every click is a genuine customer showing interest. Google categorizes invalid traffic (IVT) into several buckets, ranging from simple user mistakes to sophisticated, malicious attempts to drain advertising budgets. Common examples of invalid traffic include: Accidental Clicks: Double-clicks on ad elements, or clicks on mobile ads that occur because a user was trying to scroll past a banner. Manual Click Fraud: Competitors manually clicking on your search ads to exhaust your daily budget and temporarily remove your business from the search results. Automated Bot Traffic: Automated scraping tools, web crawlers, and malicious botnets designed to simulate human behavior, inflating click metrics on search results pages or display networks. Deceptive Ad Placements: Clicks generated through unethical publisher tactics, where ads are hidden or placed in a way that forces accidental user interaction. Google employs an extensive, multi-layered system to detect and filter out these low-quality interactions. However, because detection methods must constantly evolve to keep up with sophisticated ad fraud techniques, not all invalid traffic can be caught in real-time. This is where retroactive credits come into play. The Mechanics of Google’s Detection and Refund System Google’s invalid traffic protection operates in two primary phases: proactive real-time blocking and retroactive post-billing analysis. In the first phase, Google’s automated algorithms analyze every single click and interaction as it happens. If a click is deemed highly suspicious or clearly accidental, it is filtered out immediately. In these cases, the advertiser is never charged, and the invalid interaction does not impact the campaign’s billing metrics. This real-time detection handles the vast majority of invalid traffic. The second phase involves deep forensic analysis. Some sophisticated invalid activity can only be identified after the billing cycle has concluded, as patterns of coordinated bot behavior or click fraud often require days or weeks of data to emerge. When Google’s offline analysis confirms that invalid clicks slipped past the initial real-time filters, the system automatically calculates the cost of those clicks and issues a credit to the advertiser’s billing account. Historically, finding and reconciling these refunds was a frustrating experience. Advertisers could see “Invalid Activity” credits on their monthly billing statements, but these credits were typically presented as lump-sum adjustments. There was no straightforward way to tie those refunds back to specific campaigns, dates, or performance trends. The Invalid Activity Credit Report bridge this gap by offering granular transparency. A Deep Dive into the Invalid Activity Credit Report The newly highlighted help documentation makes it clear that the Invalid Activity Credit Report is designed to provide a highly detailed, campaign-level view of how invalid traffic impacts your performance and budget. Rather than guessing which campaigns were targeted by invalid traffic, digital marketers can now pinpoint exactly where the adjustments occurred. When you generate the report for Search and Performance Max campaigns, you gain access to several critical data columns: Credited Clicks: The exact number of clicks that Google determined to be invalid after billing had already occurred, which have now been refunded. Credited Interactions: This extends beyond standard search clicks to include other ad engagement types, such as swipe-ups, video views, and local action clicks, particularly on diverse inventory networks like Performance Max. Credited Spend: The exact dollar amount refunded to your account for the associated invalid clicks and interactions. Campaign-Level Impact: A breakdown showing precisely which campaigns were affected by invalid traffic, allowing you to see if specific targeting options, keywords, or asset groups are attracting higher levels of non-converting traffic. Adjusted Performance Metrics: Perhaps the most valuable aspect of the report, this feature recalculates your campaign performance metrics (such as Click-Through Rate, Cost-Per-Click, and Conversion Rate) after removing the invalid traffic. This gives you a highly accurate view of your actual marketing ROI. Why This Report is Crucial for Advertisers and Agencies The release of updated documentation is a welcome development for the PPC community. As ad budgets rise and automation plays a larger role in modern campaigns, transparency is more important than ever. There are several key reasons why advertisers and digital marketing agencies should pay close attention to this report. 1. Eliminating Manual Billing Reconciliation For decades, agency account managers and in-house finance teams have spent hours trying to reconcile monthly Google Ads invoices with actual platform performance. If an invoice showed a credit for invalid activity, it was incredibly difficult to determine which client campaign originally incurred the waste. This report drastically reduces the need for manual reconciliation by directly mapping billing credits to individual campaigns and performance metrics. 2. Verifying the Transparency of Performance Max Campaigns Performance Max (PMax) campaigns utilize Google’s advanced machine learning to serve ads across Search, YouTube, Display, Discover, Gmail, and Google Maps. While PMax campaigns are highly effective at driving conversions, they have also faced criticism for their “black box” nature. Advertisers have consistently asked for deeper reporting on where their ads are appearing and what kind of traffic they are generating. The integration of PMax data into the

Google begins testing healthcare ads in AI Mode The landscape of digital advertising is undergoing a massive shift as generative artificial intelligence becomes deeply integrated into search engines. In its latest move to monetize these advanced search environments, Google has officially confirmed that it is beginning a small-scale test of healthcare-related ads within its new “AI Mode.” This development represents a significant step forward for Google’s search monetization strategy. Healthcare has historically been one of the most strictly regulated, sensitive, and carefully monitored advertising categories on the internet. Testing ad placements in an AI-driven interface for this sector indicates that Google is confident in its AI’s ability to maintain brand safety and compliance—or is at least ready to put those boundaries to the test. The Details of the AI Mode Healthcare Ad Test The confirmation of this pilot program follows weeks of observations and speculation among digital marketing professionals. Industry analysts and search engine marketers had begun noticing healthcare-related promotional materials appearing within AI-generated search results, prompting questions about whether Google had quietly expanded its ad inventory. Responding to these inquiries on LinkedIn, Google Ads Liaison Ginny Marvin officially confirmed the program’s existence. According to Marvin, Google is “beginning a small test of ads in AI Mode for the healthcare vertical.” As with many of Google’s early-stage feature tests, this pilot has highly specific parameters designed to control quality and gather clean performance data. The current constraints of the test include: Geographic Limitation: The test is strictly limited to healthcare advertisers targeting users within the United States. Language Restriction: Only English-language search queries are currently eligible to trigger these ads in AI Mode. Controlled Access: Only a select group of advertisers and specific campaign types are participating in this initial phase. The test was first spotted in the wild by Ben Goldman, a Senior Strategist, who noticed the placements and raised the question in a reply to Ginny Marvin’s recap of the Google Marketing Live (GML) 2026 event on LinkedIn. Marvin’s subsequent response cleared up the speculation, providing the industry with concrete details on which campaigns can participate. Eligible Campaign Types in AI Mode For healthcare brands that meet the criteria to participate in the test, Google has opened up several of its most popular automated and AI-driven campaign types. According to Google, the same campaign models that serve ads in traditional AI Overviews are also eligible for the new AI Mode test. These include: Performance Max (PMax) Campaigns Performance Max has become the cornerstone of Google’s automated advertising ecosystem. By leveraging machine learning to optimize bids and placements across YouTube, Display, Search, Discover, Gmail, and Maps, PMax allows advertisers to find converting customers wherever they are. Its inclusion in the AI Mode test suggests that Google is using its algorithm to dynamically match user intent in conversational AI searches with relevant product or service offers. AI Max with Search Term Matching AI Max represents Google’s deeper push into machine-learning-driven campaign architecture. By utilizing advanced search term matching, this campaign type allows Google’s AI to interpret the nuances of natural language queries. Rather than relying solely on exact keywords, it maps semantic intent to an advertiser’s offerings, which is crucial in a conversational environment like AI Mode where users speak or type in long, complex queries. Shopping Campaigns Product listing ads are also part of the mix. For e-commerce-enabled healthcare brands, such as those selling over-the-counter wellness products, medical devices, or health supplements, Shopping campaigns in AI Mode could allow users to purchase products directly referenced in an AI-generated answer. Broad Match Campaigns Broad match keywords allow advertisers to reach a wider audience by matching queries that are related to—but not necessarily containing—the exact keyword. In the context of AI Mode, broad match provides the flexibility needed for Google’s systems to connect conversational, highly detailed health queries with relevant commercial solutions. The Fine Print: Creative Restrictions and Compliance Because healthcare is subject to stringent legal regulations, consumer safety laws, and internal platform policies, Google has introduced several creative restrictions for this initial pilot. These boundaries are designed to prevent misleading claims and ensure that AI-generated spaces remain trustworthy for users seeking medical information. Marvin highlighted that the initial phase of this test is restricted to ad creatives that do not require: Pinned Assets: Advertisers cannot pin specific headlines or descriptions to fixed positions in these ad placements. Google’s AI must have the flexibility to assemble and present the ad copy dynamically to fit the context of the AI Mode response. Text Disclaimers: Ads that rely heavily on lengthy text disclaimers, such as those often required for prescription pharmaceuticals (often referred to as “fair balance” statements), are excluded from this early testing phase. These limitations significantly narrow the pool of eligible healthcare advertisers. Large pharmaceutical companies promoting prescription medicines, for instance, may have to sit out of the initial test due to their regulatory requirements for detailed warnings and disclaimers. Consequently, the test is likely dominated by hospitals, local healthcare providers, wellness brands, telehealth platforms, and manufacturers of general medical supplies who can run compliant ads without heavy text disclaimers. Why This Test Matters for the Digital Marketing Industry The introduction of healthcare ads to AI Mode is more than just a minor update to Google’s ad inventory; it is a major indicator of where the search industry is heading. There are several reasons why digital marketers, SEO specialists, and paid media managers are watching this roll-out closely. Monetizing the AI Search Experience As users transition from traditional search results to interactive, AI-driven summaries, search engines risk losing traditional ad revenue. By introducing ads directly into AI Mode, Google is demonstrating how it plans to protect its primary revenue engine. If successful, this format will change how brands allocate their budgets between traditional search engine marketing (SEM) and AI-targeted advertising. A Blueprint for Other Regulated Industries If Google can successfully navigate the complexities of healthcare advertising in an AI interface, it paves the way for other highly regulated sectors. Industries

Understanding the July 2026 Google Ads Terms of Service Update The landscape of pay-per-click (PPC) advertising is undergoing its most significant structural shift in a generation. As artificial intelligence and automated campaign types become the default operating model for search engine marketing, Google is updating its legal framework to match this new reality. Google is rolling out a comprehensive update to the Google Ads Terms of Service, set to take effect on July 1st. This update explicitly codifies how Google’s machine learning models and automated systems can utilize advertiser-provided inputs. Crucially, the revisions clarify the legal relationship between the automation generating the campaigns and the human advertisers who ultimately pay for them. While the updated terms require no immediate opt-in action from account administrators prior to the rollout date, they carry massive implications for brand governance, campaign control, and legal liability. It is important to note at the outset that these changes apply strictly to Google Ads accounts. Other enterprise services, such as Google Workspace, Google Cloud Platform, and Cloud Identity, remain unaffected by this specific legal update. However, for search engine marketers, agency partners, and in-house digital teams, the upcoming terms represent a pivotal moment in the evolution of digital advertising. The Core Changes: AI Training, Crawling, and System Integration The updated Terms of Service reflect a platform that is transitionally moving away from traditional keyword-and-bid mechanics toward fully automated, intent-based systems. To power these systems, Google’s algorithms require a continuous loop of data. The revised terms focus heavily on how Google accesses, processes, and utilizes the assets, data, and information that advertisers provide. 1. Expanded Use of Advertiser-Provided Inputs Under the new terms, Google has expanded the language surrounding how advertiser-provided inputs may be utilized across the broader Google Ads ecosystem. These inputs include creative assets, copy, product feeds, and customer lists. The updated terms clarify that these materials can be leveraged by Google’s automated systems to optimize performance, refine bidding strategies, and train internal machine learning models to improve overall campaign delivery. 2. Integration of Conversational AI Tools With the introduction of conversational campaign creation tools, Google has integrated natural language chat interfaces directly into the ad creation flow. The new Terms of Service explicitly state that any information, data, or prompts entered into these conversational experiences can be captured and utilized by Google’s underlying systems. This means that proprietary business information, brand guidelines, or target demographic details shared during a chat setup are legally integrated into Google’s database for system optimization. 3. Automated Crawling of Web Properties For automated campaigns like Performance Max and AI-driven Search campaigns to function effectively, Google’s bots must dynamically understand an advertiser’s website content. The updated terms clarify and expand Google’s authorization to access, crawl, and analyze the URLs and accounts provided by the advertiser. This authorization is designed to streamline the automated setup of landing page destinations, asset generation, and ad group targeting without requiring manual verification for every new page or asset. The Language Shift: From Tools of Assistance to Full Authorization To fully grasp the magnitude of this update, one must look at how the contractual relationship between Google and the advertiser has evolved. Historically, Google Ads Terms of Service framed automation as an optional convenience. Previous terms generally stated that Google could provide optional tools to assist advertisers in generating keywords, target audiences, or ad copy, while preserving clear mechanisms to opt-out of these features. The upcoming terms remove much of this optional framing. The revised language introduces a sweeping authorization clause: “Customer authorizes Google and its affiliates to serve ads, including through the use of automated program features to format, select, or generate targets, ads, or destinations on Customer’s behalf.” This single sentence represents a monumental shift. By agreeing to the new terms, advertisers grant Google’s machine learning systems the explicit legal authority to write ad copy, choose search query targets, format creative placements, and select the specific landing pages to which users are sent. What was once an optional optimization feature is now codified as a foundational element of how Google operates its advertising network. Why the PPC Community is Raising Concerns The response from the digital marketing community has been a mix of caution and criticism. Many veteran search marketers view these changes as a continuation of Google’s long-term strategy to reduce manual controls in favor of a black-box approach to ad buying. Anthony Higman, founder of the specialized legal marketing agency AdSQUIRE, has been one of the most vocal critics of the updated terms. Higman argues that the revision systematically erodes two of the historical pillars of search engine marketing: relevance and control. According to Higman, previous iterations of the Google Ads platform succeeded because advertisers could precisely control the exact search queries their ads appeared on, the exact copy displayed to the user, and the specific landing page of the destination URL. By shifting this authority to automated program features that format, select, or generate these core components, the advertiser is largely relegated to a passive supervisory role. Critics also point out the asymmetrical nature of this arrangement. While Google’s algorithms gain broader permission to automatically generate and serve ad variations, the advertiser continues to carry the full financial and brand safety risk if those automated systems perform poorly, display inaccurate information, or target irrelevant audiences. The Liability Dilemma: Automation vs. Accountability One of the most critical aspects of the upcoming Terms of Service is the reaffirmation of advertiser responsibility. Despite Google’s systems taking a highly active role in generating targets, writing ad copy, and selecting landing destinations, the legal liability remains entirely on the customer. Under the new terms, advertisers must guarantee that they possess all necessary rights, trademarks, copyrights, and permissions for any content, assets, URLs, or data inputs provided to Google Ads. Furthermore, the contract makes it clear that the advertiser is solely responsible for: Reviewing and approving any auto-generated ad assets, text variations, or asset groups. Monitoring, editing, or removing campaigns that are automatically generated by Google’s

The landscape of search is undergoing its most profound transformation since the invention of the commercial web browser. For nearly three decades, search engines have been built for humans. They index pages, evaluate ranking signals, and present a list of blue links for human fingers to click and human eyes to read. However, as artificial intelligence transitions from conversational chatbots to autonomous agents, this human-centric search model is starting to show its limitations. Recognizing this fundamental shift in how the internet is navigated, Microsoft has officially released Web IQ. This new grounding API suite, powered by Bing’s massive web index, is designed specifically for how AI agents—rather than human users—interact with, retrieve, and synthesize real-time data from across the web. Announced via the official Microsoft announcement, Web IQ is built from the ground up to serve as the informational backbone for the next generation of artificial intelligence. What is Web IQ? Grounding the Agentic Era At its core, Web IQ is a suite of AI-native grounding APIs. In the context of large language models (LLMs), “grounding” is the process of linking abstract AI systems to real-world, verified, and up-to-date information. While base LLMs are frozen in time based on their training cutoff dates, grounding allows them to fetch live information, reducing the risk of “hallucinations” and ensuring that output is accurate, timely, and contextually relevant. Web IQ connects AI systems and autonomous agents to fresh intelligence spanning the entire digital ecosystem. This includes standard web pages, real-time news articles, images, and videos. Because it is powered by Bing’s index, Web IQ inherits decades of search technology, web crawling infrastructure, and semantic understanding. However, the way Web IQ processes and delivers this information is entirely different from traditional search. While the APIs behind Web IQ represent the next step in Microsoft’s developer ecosystem, the underlying infrastructure is already battle-tested. Web IQ utilizes the same API infrastructure that powers Microsoft Copilot and some of the world’s most sophisticated AI systems, including OpenAI’s ChatGPT. It is the engine that allows ChatGPT to browse the web for specific queries and enables Bing to generate synthesized Copilot answers directly at the top of its search engine results pages (SERPs). Why Traditional Search Fails AI Agents To understand why Microsoft developed Web IQ, it is essential to analyze how AI agents use the web compared to humans. According to Jordi Ribas, President of Search & AI at Microsoft, traditional search is optimized for human browsing habits, which are fundamentally different from agentic workflows. When a human searches for “best enterprise CRM software,” they typically enter a single query, scan the first page of results, click on two or three promising links, and manually synthesize the information. For this behavior, traditional search ranking is highly critical. Being the number-one result on Google or Bing can make or break a business because humans rarely venture past the first few options. AI agents do not behave this way. An agent tasked with “evaluating and recommending a CRM software based on our company’s specific budget, user count, and integration requirements” does not just click a link and stop. Instead, the agent engages in what is known as “fanning out” or multi-hop searching. It will: Execute an initial search to identify the top five CRM players. Simultaneously launch five secondary searches to pull pricing, integration documents, and API limitations for each of those players. Execute tertiary searches to find user reviews on specific platforms like G2 or Reddit. Synthesize hundreds of pages of raw data into a structured report. For an AI agent, traditional human-centric search results pages are cluttered with unnecessary elements, such as ads, layout code, navigation menus, and engagement bait. Agents do not need beautifully formatted web pages; they need clean, structured, and highly relevant data passages. They do not care about ranking as much as they care about raw information extraction, speed, and contextual accuracy. Re-Architecting the Stack: From Indexing to Orchestration Because agents search deeply, rapidly, and continuously, Microsoft had to rebuild its search infrastructure from the ground up to support Web IQ. This meant redesigning every single layer of the search stack to align with the requirements of inference-time grounding. 1. Indexing and Retrieval Traditional search indexing prioritizes page speed, authority, and visual presentation. Web IQ’s indexing focus is shifted toward semantic data extraction. The system indexes web content in a way that allows AI models to quickly parse the semantic meaning of a page, rather than just matching keywords or evaluating classic backlink profiles. 2. Passage Selection and Extraction Rather than returning a full HTML document or a simple meta description snippet, Web IQ is optimized to locate the exact passages within a document that answer an agent’s query. This reduces the work the LLM has to do to find the needle in the haystack, saving processing power and time. 3. Orchestration Because agentic workflows require multiple search steps, Web IQ’s orchestration layer is built to handle complex, multi-turn queries. It allows agents to perform parallel searches, refine queries on the fly, and pull diverse media types (like videos and images) to support multi-modal reasoning. The Critical Bottlenecks: Speed, Tokens, and Costs In the developer world, building agentic applications can quickly become prohibitively expensive and sluggish. When an agent has to search the web multiple times to complete a single user request, two major bottlenecks emerge: latency and token usage. Every piece of text sent to or received from an LLM is measured in tokens. If a search API returns a massive, unoptimized webpage full of boilerplate code and irrelevant text, the agent must process all of those tokens. This drives up the cost per API call and slows down the system’s response time. Microsoft built Web IQ specifically to solve these economic and performance challenges. The system is designed to use the fewest tokens possible while delivering the highest quality answers. The operating philosophy behind Web IQ is simple: “fewer tokens in, better answers out, lower cost per call.” Furthermore, speed is

The Next Evolution of AI-Driven Search The landscape of search engine optimization is shifting beneath our feet at an unprecedented pace. As Google continues to integrate generative artificial intelligence into its core search product, digital marketers, SEO professionals, and business owners must constantly adapt to new features, algorithmic shifts, and user behaviors. The transition from traditional blue links to interactive, conversational search experiences is no longer a future projection—it is our current reality. Recently, Google introduced major updates that further solidify the role of generative AI in everyday search journeys. Among these updates are the expansion of “Preferred Sources” within AI Overviews and the brand-new AI Mode, fresh insights from Alphabet CEO Sundar Pichai regarding the viability of AI-driven search, and fascinating data from iPullRank detailing how Gmail activity directly influences brand visibility in AI-driven search environments. Alongside these organic developments, Google is also rolling out highly targeted ad formats designed specifically for AI Overviews, signaling a new era for paid search marketing. Understanding these developments is crucial for any business aiming to maintain or grow its digital footprint. Let us dive deep into the mechanics of these changes, what they mean for your SEO strategy, and how you can position your brand to thrive in an AI-first search ecosystem. Google Expands Preferred Sources to AI Overviews and AI Mode One of the most significant updates to Google’s search interface is the expansion of the “Preferred Sources” feature. Originally designed to give users more control over the types of content they see, Google is now integrating this feature directly into AI Overviews and the dedicated AI Mode. Preferred Sources allow searchers to actively designate specific websites, publishers, or platforms as trusted authorities. When a user conducts a query, Google’s generative AI models prioritize information from these chosen domains to construct the AI Overview response. This represents a monumental shift from purely algorithmic retrieval to a hybrid system where user preference plays a defining role in content delivery. The Mechanics of Preferred Sources in AI Mode In AI Mode, which offers a highly conversational and iterative search experience, the inclusion of Preferred Sources changes how answers are synthesized. If a user frequently relies on a particular tech blog, culinary site, or financial news outlet, Google’s Gemini-powered algorithms will heavily weight content from those specific URLs when compiling real-time summaries. This personalization layer means that two users searching for the exact same query in AI Mode may receive completely different, highly customized summaries based on their individual Preferred Sources. For SEOs, this emphasizes the importance of brand loyalty and direct user engagement. It is no longer enough to rank for a keyword; you must convince your target audience to explicitly designate your site as a trusted resource within their Google ecosystem. What This Means for Organic Traffic The expansion of Preferred Sources could lead to a bifurcation of organic search traffic. Brands that successfully secure a spot in a user’s Preferred Sources list will enjoy highly consistent, high-intent traffic, as their content will be continuously surfaced in AI-generated answers. Conversely, websites that rely solely on transactional, one-off visits from generic search queries may see a decline in visibility as personalized AI Overviews take center stage. Sundar Pichai Addresses the Future of AI Overviews As AI Overviews continue to roll out globally, concerns from the publishing and digital marketing communities have reached a fever pitch. Many content creators fear that zero-click searches will skyrocket, starving publishers of the traffic required to sustain their business models. Addressing these concerns, Alphabet and Google CEO Sundar Pichai has shared crucial insights regarding the health, monetization, and user reception of AI Overviews. The Economics of AI Search Historically, one of the biggest hurdles for Google in deploying massive generative AI models was the computing cost. Serving a generative AI answer is exponentially more expensive than serving a standard list of indexed links. However, Pichai has noted that Google has dramatically optimized its infrastructure, slashing the cost of serving AI Overviews by significant margins since their initial testing phase. This cost reduction ensures that AI Overviews are not a temporary experiment; they are financially viable for Google to run at scale. User Engagement and CTR Trends Pichai has also defended the impact of AI Overviews on the broader web ecosystem. According to Google’s internal data, users who interact with AI Overviews actually show higher search satisfaction and tend to conduct more complex, long-tail queries than they did previously. Crucially, Pichai emphasized that AI Overviews are designed to drive high-quality traffic to publishers. Instead of replacing the need to click, Google asserts that the context provided by AI summaries makes users more confident in the links they do choose to click, leading to higher-quality referral traffic. While the SEO industry remains cautious, Google’s official stance is clear: AI Overviews are intended to coexist with, and actively support, the open web ecosystem. How Gmail Shapes Brand Visibility in AI Mode: The iPullRank Study Perhaps one of the most eye-opening recent discoveries in the SEO space comes from the technical SEO agency iPullRank. In a groundbreaking study, researchers measured the impact of a user’s Gmail footprint on their personalized AI Mode results. The findings suggest that Google is leveraging data from across its ecosystem—specifically Gmail—to tailor brand visibility within AI-generated search results. The “Gmail Pull” Phenomenon The iPullRank study revealed a strong correlation between a user’s email interactions and the brands surfaced in their AI Mode queries. When a user regularly receives, opens, or interacts with emails from a specific brand (such as newsletters, order confirmations, shipping updates, or promotional offers), Google’s underlying user profile notes this affinity. When that same user enters AI Mode to ask a broad, non-branded question related to that brand’s industry, Google’s AI is significantly more likely to feature that specific brand in its generated response. For example, if you frequently receive emails from a particular outdoor gear retailer, a generic query in AI Mode about “the best hiking boots for rainy weather” is highly