• 60% of Google searches are now zero-click, with mobile rates reaching 77%.
• AI Overviews appear for 13% of queries, causing click-through rates to drop by 47%.
• Median publishers experienced a 10% year-over-year traffic decline in early 2025.
• Implementing comprehensive schema markup can recover up to 45% of lost click-through rates.
• AuraSearch provides the technical framework required to win citations in generative search results.

Recover AI impacted traffic with these proven steps:

1. Diagnose the cause - Check Google Search Console for stable impressions but falling click-through rates, which is the clearest signal of AI Overview impact.
2. Audit affected queries - Identify informational keywords where AI Overviews now appear and answer the query directly.
3. Restructure content - Lead with a direct answer in the first 80 words, use question-based headings, and add structured lists.
4. Implement schema markup - Deploy FAQ, HowTo, and Article schema to make content machine-readable for AI citation.
5. Strengthen E-E-A-T signals - Add author credentials, original research, and credible citations to build trust with both AI systems and readers.
6. Diversify traffic sources - Build owned channels like email lists and expand to platforms like YouTube and LinkedIn.

Google's search results look different now. AI Overviews sit at the top of the page, synthesising answers from multiple sources before a single blue link appears. For many publishers, the result has been a quiet but damaging traffic erosion: impressions hold steady, rankings stay the same, but clicks keep falling.

The numbers confirm this is not a minor fluctuation. AI Overviews now appear for over 13% of all queries, more than double the rate from January 2025. When they appear, click-through rates drop by 47%, falling from 15% to just 8%. The median publisher recorded a 10% year-over-year traffic decline in the first half of 2025, with non-news content sites down as much as 14%. Some high-profile brands have reported losses far exceeding that benchmark.

This is the core problem : search volume is actually rising, but fewer of those searches result in a website visit. Researchers call it the Great Decoupling.

The good news is that this is a structural challenge with structural solutions. Content that is built for AI extraction, supported by proper schema, and backed by genuine topical authority can still earn citations inside AI Overviews and convert that visibility into clicks.

Amber Brazda is an AI Search Specialist with over a decade of experience in digital authority strategy, and has directly led campaigns to recover AI impacted traffic for national brands facing attribution erasure in generative search. The sections below lay out the exact framework that drives measurable results.

Generative search features create a barrier between the user and the website. Google AI Overviews act as a synthesis layer that provides immediate answers. This reduces the incentive for a user to click through to the source material. Statistics show that zero-click searches now account for 60% of all Google queries. Mobile users experience this shift even more acutely, with zero-click rates reaching 77%.

The presence of AI Overviews leads to a significant reduction in traditional organic traffic. Research indicates that click-through rates plummet to 8% when an AI summary is present. This is a stark contrast to the 15% click-through rate observed in traditional search results. To recover AI-impacted traffic, brands must transition from a traditional SEO mindset to Generative Engine Optimisation (GEO). This involves making content highly extractable for large language models.

Detecting the impact of AI Overviews requires a specific analysis of Google Search Console data. A primary indicator is the maintenance of average positions and impressions alongside a sharp decline in click-through rates. This pattern signals that the page still ranks well, but the AI summary is satisfying the user intent on the search results page.

Scientific research on AI Overviews confirms that these features reduce website clicks by nearly 50%. Informational queries are the most vulnerable. These queries trigger AI Overviews approximately 90% of the time. Commercial queries trigger them only 8% of the time. Identifying which keywords fall into the informational category allows for a prioritised recovery plan. More technical details are available through AI Overview optimisation resources.

Technical SEO remains a critical component of search visibility. AI systems rely on structured data to understand the relationships between entities on a page. Implementing comprehensive schema markup is a proven method to increase the chances of being cited in an AI Overview. Schema provides a machine-readable roadmap that simplifies the extraction process for Google's generative models.

FAQ and HowTo schema are particularly effective for this purpose. These formats align with the question-and-answer nature of generative search. Pages with properly implemented schema can see a recovery of up to 45% in click-through rates through AI extraction. Adherence to Google’s guidelines regarding helpful content is mandatory. Technical hygiene ensures that crawlers can access and index the content efficiently. Further insights are found in AI search optimisation documentation.

Traditional long-form content often buries the lead. Generative Engine Optimisation requires an answer-first structure. The primary answer to a query must appear within the first 80 words of a section. This allows AI models to identify and quote the source verbatim. Semantic completeness is the leading predictor of AI citations, with a correlation coefficient of 0.87.

Multi-modal elements also improve citation rates. Content featuring tables, numbered lists, and images performs 156% better in AI retrieval tasks. Clear H2 and H3 headings should be phrased as questions to match the natural language processing patterns of AI engines. Guidance on this transition is provided in the SEO survival guide.

Experience, Expertise, Authoritativeness, and Trustworthiness (E-E-A-T) are the filters AI systems use to select sources. AI models prioritise content that demonstrates first-hand experience and deep topical depth. Anonymous or generic content is increasingly ignored by generative summaries. Including detailed author bylines and credentials builds the necessary trust signals.

Original research and proprietary data are highly defensible assets. AI Overviews often cite the original source of a statistic or a unique study. Brands must avoid scaled content practices that rely on mass-produced, low-value AI text. Google has deindexed a significant percentage of sites that utilised unoriginal, scaled AI content. Establishing a reputation as a primary source ensures long-term visibility. The link between authority and performance is detailed in tracking AI SEO value reports.

The search landscape has undergone a permanent structural shift. Traditional traffic metrics are no longer the sole indicator of search success. Success in 2025 and beyond requires a focus on AI visibility share and brand citations. This shift necessitates a technical and strategic pivot that many internal teams are not equipped to handle alone.

AuraSearch provides the specialized technical capability required to navigate this era of generative search. The platform offers comprehensive entity optimisation and AI visibility mapping to ensure brands are not just ranking, but are being cited as authoritative answers. By implementing search intent modelling and generative answer capture strategies, AuraSearch helps businesses reclaim their digital footprint.

The future of search belongs to those who adapt to the generative model. AuraSearch defines the standard for Generative Engine Optimisation, turning the threat of AI Overviews into a strategic advantage for brand growth. Explore the full range of AuraSearch services to begin the recovery process.

Recovery timelines typically range from 30 to 120 days. Initial schema implementations show results within the first month. Full generative engine optimisation strategies deliver significant traffic restoration by the four-month mark. The duration depends on the volume of affected pages and the speed at which Google re-crawls the updated content structures.

Schema markup provides machine-readable data that AI systems use for citations. FAQ and HowTo structures increase the probability of appearing in AI Overviews. This technical layer bridges the gap between traditional indexing and generative answer capture. Sites using comprehensive schema have reported recovering nearly half of their lost click-through rates by becoming the cited source in the AI summary.

The most common sign is a decoupling of rankings and clicks. A website may maintain its top positions for key terms while experiencing a 30% to 50% drop in actual traffic. Impressions usually remain stable or even increase because the page is still being "seen" by the search engine, but the user finds the answer in the AI Overview and does not click through to the site.

Informational content and simple definition-based pages are hit the hardest. Queries that can be answered in a single paragraph, such as "how to" guides or factual definitions, are easily synthesised by AI Overviews. Transactional and commercial pages are generally safer. These queries often require the user to browse products or compare prices, which necessitates a visit to the actual website.

Recovery is possible but requires a complete content overhaul. Google targets "scaled content abuse" where AI is used to produce large volumes of low-value pages. To recover, a site must remove or significantly edit the low-quality AI text and replace it with human-verified, expert-led content. This process involves proving E-E-A-T signals and ensuring every page provides unique value that cannot be replicated by a basic prompt.