Key takeaways
- Google AI Overviews now appear on 48% of all tracked queries (up 58% year-over-year), and click-through rates have dropped nearly 30% as a result.
- Only 30% of brands maintain consistent visibility from one AI-generated answer to the next, according to AirOps data.
- Google's AI Overviews are 44% more likely to speak negatively about a brand than ChatGPT, making sentiment monitoring as important as presence tracking.
- Brands winning in AI search tend to share specific traits: structured content, strong third-party citations, and active presence on sources AI models trust (Reddit, YouTube, review sites).
- Tracking and optimizing for AI visibility requires different tools and strategies than traditional SEO -- monitoring alone isn't enough.
The click is no longer the point
For two decades, the click was everything. You ranked, you got traffic, you measured success in sessions and pageviews. That model is breaking down fast.
When Pew Research Center analyzed 68,879 Google searches in March 2025, users who encountered an AI-generated summary at the top of their results clicked on a traditional search link only 8% of the time -- roughly half the 15% click rate for searches without one. Even more striking: just 1% of visits resulted in someone clicking a cited link inside the AI summary. A full quarter of users who saw an AI Overview ended their browsing session entirely, satisfied with the answer they got.
That was over a year ago. Things have moved faster since.
By early 2026, BrightEdge's data shows AI Overviews appearing on 48% of all tracked queries, up 58% from a year earlier. Google search impressions climbed 49% over the same period, but click-through rates dropped nearly 30%. More searches, fewer clicks. The math is brutal for brands that haven't adjusted.
Google CEO Sundar Pichai tried to frame this positively during the company's Q2 2025 earnings call, arguing that AI Overviews were driving a 10% increase in search usage overall. That may be true. But more searches with fewer clicks per search doesn't automatically help brands that built their business on organic traffic.
The real question isn't whether clicks are declining. They clearly are. The question is: which brands are still winning, and how?
The visibility gap is wider than most brands realize
Here's a number that should make any marketing team uncomfortable: only 30% of brands stay visible from one AI answer to the next, according to data shared by AirOps. When you extend that to five consecutive queries on the same topic, just 20% of brands remain present throughout.
That means for most brands, AI visibility is inconsistent at best and nonexistent at worst. You might appear when someone asks "best project management software for remote teams" but disappear entirely when the same person follows up with "which project management tools integrate with Slack." Same buyer, same intent, different answer.
GrowthOS's analysis of AI brand visibility across 10 industry categories found something counterintuitive: size doesn't guarantee AI visibility. Specialized brands with strong authority signals routinely outperform category leaders in AI-generated recommendations. A niche cybersecurity firm with deep, well-cited content can outrank a Fortune 500 competitor in AI answers, even if that competitor dominates traditional search.
Promptwatch tracks this kind of visibility across 10 AI models simultaneously, and their data (over 1.1 billion citations, clicks, and prompts processed) tells a similar story: the brands winning in AI search aren't necessarily the biggest -- they're the ones that AI models have the most reason to trust.
What Google AI Overviews actually reward
Google's AI Overviews pull from a different set of signals than traditional organic rankings. Understanding what those signals are is the starting point for any visibility strategy.
Structured, answer-ready content
AI Overviews are built to answer questions. Content that's structured around specific questions -- with clear headings, concise answers near the top, and supporting detail below -- gets cited more often than content that buries the answer in narrative prose. This isn't about keyword stuffing. It's about making it easy for a language model to extract a clear, accurate answer.
FAQ sections, comparison tables, step-by-step guides, and definition-style content all tend to perform well. Long-form content that meanders before getting to the point tends to get skipped.
Third-party validation
AI models don't just read your website. They read everything written about you. Reddit threads, YouTube reviews, G2 and Trustpilot ratings, industry publications, and news coverage all feed into how AI models perceive and recommend brands.
This is a significant shift from traditional SEO, where you had more control over your own content. In AI search, your reputation on third-party platforms matters as much as your own site. A brand with 200 five-star G2 reviews and active Reddit discussions is going to appear in more AI answers than a brand with a polished website and no external footprint.
Topical authority, not just keyword coverage
AI models reward brands that comprehensively cover a topic, not just brands that rank for individual keywords. If you sell project management software, having content that covers team collaboration, task dependencies, sprint planning, and resource allocation -- not just "project management software" -- signals to AI models that you're a genuine authority.
The GrowthOS benchmark data found that brands covering 70%+ of the subtopics within their category appeared in AI answers at roughly 3x the rate of brands with narrow content coverage.
Citation quality over citation quantity
Not all backlinks are equal in traditional SEO, and not all citations are equal in AI search. Being cited in a Wall Street Journal article carries more weight than being mentioned in 50 low-authority blog posts. AI models appear to weight citations from authoritative, frequently-crawled sources more heavily.
The sentiment problem nobody's talking about
Here's the finding that caught me off guard: according to a study published in Fortune in March 2026, Google's AI Overviews are 44% more likely to speak negatively about a brand than OpenAI's ChatGPT. The study was based on hundreds of millions of prompts.
That's not a small difference. It means a brand could be getting mentioned frequently in Google AI Overviews while simultaneously being framed in a negative light -- and if you're only tracking presence (whether you appear) rather than sentiment (how you're described), you'd miss it entirely.
This matters especially for brands in competitive categories where AI Overviews might compare products, highlight complaints, or surface negative reviews. Being visible in AI search isn't automatically good. Being visible and being recommended positively are two different things.
Monitoring sentiment across AI platforms -- not just presence -- is now a core part of any serious brand visibility strategy.
Who's winning and why
Across the data available, a few patterns emerge consistently among brands that perform well in Google AI Overviews.
They publish content that directly answers buyer questions
The brands showing up most consistently in AI Overviews aren't necessarily the ones with the most content. They're the ones whose content most directly answers the questions buyers actually ask. This sounds obvious, but most brand content is still written to persuade rather than to inform -- and AI models are much better at extracting informational content than promotional content.
They're active where AI models look for social proof
Reddit and YouTube have emerged as surprisingly important signals for AI visibility. Go Fish Digital's 2026 analysis found that brands with active, authentic presence on Reddit -- through community participation, not just self-promotion -- appeared more frequently in AI-generated recommendations than brands with no Reddit footprint.
This makes sense when you think about how AI models work. They're trained on the web, and Reddit is one of the most heavily-crawled, high-trust sources on the web. If people are discussing your brand positively on Reddit, that signal feeds into AI recommendations.
They track visibility at the page level, not just the brand level
Winning brands know which specific pages are getting cited by AI models, how often, and in response to which prompts. This lets them double down on what's working and identify gaps where competitors are appearing but they're not.
Tools like Promptwatch make this possible with page-level citation tracking across multiple AI models. The brands that are improving fastest are the ones treating AI visibility as a measurable, optimizable metric -- not a vague aspiration.
Who's losing and why
The flip side is equally instructive. Brands struggling with AI visibility tend to share a few common characteristics.
They're optimizing for keywords, not questions
Traditional SEO trained us to think in terms of keywords. AI search thinks in terms of questions and intents. A brand that ranks #1 for "CRM software" but has no content addressing "what CRM is best for a 10-person sales team" or "how do I migrate from Salesforce to HubSpot" is going to be largely invisible in AI answers, even if their traditional rankings look great.
They have no third-party presence
If your brand only exists on your own website and paid channels, AI models have very little reason to recommend you. The brands disappearing from AI answers are often the ones that never invested in building presence on the platforms AI models trust most: industry publications, review sites, community forums, and video platforms.
They're not monitoring, so they don't know what they're missing
The most common reason brands are losing ground in AI search is simply that they don't know it's happening. Without visibility tracking, you can't see that a competitor is appearing in 80% of AI answers for your core category while you appear in 15%. You can't see which prompts you're missing. You can't see whether AI models are describing you accurately.
This is the gap that tools built specifically for AI visibility monitoring are designed to close.
Tools worth knowing about
The market for AI visibility tools has grown significantly in 2026. Here's a quick comparison of the main approaches:
| Tool | Best for | Monitoring | Content gap analysis | Content generation | Sentiment tracking |
|---|---|---|---|---|---|
| Promptwatch | Full-cycle AI visibility | Yes (10 models) | Yes | Yes | Yes |
| Profound | Enterprise monitoring | Yes | Limited | No | Partial |
| Otterly.AI | Budget monitoring | Yes | No | No | No |
| BrightEdge | Enterprise SEO + AI | Yes | Partial | No | Partial |
| AthenaHQ | Monitoring-focused teams | Yes (8 models) | No | No | No |
| Peec.ai | Multi-language monitoring | Yes | No | No | No |
Most tools in this space are monitoring dashboards -- they show you where you stand but leave you to figure out what to do about it. The more useful tools close the loop between visibility data and content action.
For brands that want to go beyond tracking and actually improve their AI visibility, Promptwatch is worth a look. Its Answer Gap Analysis shows exactly which prompts competitors are visible for that you're not -- and the built-in content generation tools are trained on citation data, so the content it produces is engineered to get cited by AI models, not just to read well.
What to actually do about it
If you're a marketing or SEO team trying to improve your brand's visibility in Google AI Overviews, here's where to start:
Audit your current AI visibility. Before you can improve, you need to know where you stand. Run your core category prompts through Google AI Overviews, ChatGPT, and Perplexity. Note which competitors appear and which prompts you're absent from. This gives you a baseline.
Map your content to buyer questions. Take your top 20-30 category prompts and check whether you have content that directly answers each one. Gaps here are your highest-priority content opportunities.
Build your third-party presence. Identify the review sites, forums, and publications that AI models cite most often in your category. Prioritize getting coverage there -- whether through PR, community participation, or customer review campaigns.
Monitor sentiment, not just presence. Set up tracking that tells you not just whether you appear in AI answers, but how you're described. A negative mention is worse than no mention in many cases.
Track at the page level. Know which specific pages are driving your AI citations. This tells you what content formats and topics are working, so you can replicate them.
Measure the traffic impact. AI visibility without traffic attribution is incomplete. Connect your AI citation data to actual traffic and conversions so you can demonstrate ROI and prioritize accordingly.
The brands that figure this out in 2026 will have a significant head start. AI Overviews aren't going away -- they're expanding. The question is whether your brand will be part of the answer, or invisible while competitors are.