One Egg Strategy Team
An e-commerce operator we work with watched his keywords and organic traffic fall roughly 50% year-over-year. It looked like a disaster. Except his sales were flat-to-up, and his Google conversion rate had risen sharply. The keywords she lost were informational searches. They were never buyers. The ones still arriving were further down the funnel and ready to transact.
His words to us: “not all stats are what they look like at face value.“
That captures the entire 2026 organic-search story. The headline number is misleading. Traffic is down, but revenue is holding or climbing, because AI search is reshuffling which visitors you receive. The work is no longer to stop the decline. It’s to understand what’s actually happening underneath, and build for where the qualified traffic is going.
Key Takeaways
- AI Overviews drop organic CTR by 34.5% to 58% depending on the study window (Ahrefs, 2025-2026).
- Visitors from AI search experiences convert 4.4x better than classic organic visitors (Similarweb, 2025).
- 53% of sources cited in Google’s AI Overviews don’t rank in the top 10 organically; 27% aren’t even in the top 100 (ArXiv, October 2025).
- AI platforms generated 1.1 billion referral visits in June 2025 alone, up 357% year-over-year (Similarweb, 2025).
The clicks are disappearing. Ahrefs found AI Overviews slashed CTR by 34.5% across a 300,000-keyword study in April 2025. By February 2026, that drop had deepened to 58% (Ahrefs, 2026). Seer Interactive analysed 25 million organic impressions across 42 organisations and found organic CTR dropped from 1.76% to 0.61% on queries with AI Overviews, a 61% fall. Queries without AI Overviews still lost 41% (Seer Interactive, 2025). Gartner forecasts traditional search volume will drop 25% by 2026 as buyers shift to AI chatbots (Gartner, 2024).
The platforms those buyers moved to are big now. ChatGPT hit 800 million weekly active users by October 2025 (TechCrunch, 2025). Perplexity processed 780 million queries in May 2025 alone, more than triple its volume nine months earlier (TechCrunch, 2025).
Here’s what the panic misses. Observed organic traffic across the web was down only 2.5% year-over-year as of late 2025 (Search Engine Land, 2025). AI platforms generated 1.1 billion referral visits in June 2025, up 357% YoY (Similarweb, 2025). That growth is off a smaller base, and AI referral volume is still a fraction of Google’s total traffic. But the trajectory is the point. The traffic didn’t vanish. It moved. And the traffic that arrives through AI answers converts at materially higher rates than the informational traffic it replaced.
We’ve watched this pattern repeat across dozens of client accounts over the past 24 months. Overall organic traffic drops because most of it was top-of-funnel (TOFU) informational queries. The “what is a SMSF” and “how does contract law work” traffic is collapsing into AI chat sessions. The bottom-of-funnel (BOFU) traffic, the people ready to transact, is not only holding steady but converting harder.
The numbers back up the pattern. Similarweb found visitors who land on a site via an AI search experience convert 4.4x better than classic organic visitors (Similarweb, 2025). Brands cited inside Google’s AI Overviews see 35% more organic clicks and 91% more paid clicks than uncited competitors on the same SERP (Seer Interactive, 2025).
Why the lift? AI-referred visitors arrive pre-qualified. They’ve already compared options, asked follow-ups, and narrowed their shortlist inside the AI conversation before they click through. The click itself is the last step of a consideration process that used to happen across five separate Google searches and three tabs of browser comparison.
The strategic response writes itself. Go wide and deep on BOFU content. Use paid, social, and email to cover TOFU and MOFU instead of chasing those queries through SEO. The old playbook of ranking for every informational query in your category was already low-margin. AI search just made it uneconomic.
This is the leapfrog opportunity almost nobody is talking about at the right volume. A peer-reviewed study published on arXiv in October 2025 analysed where Google’s AI Overviews actually pull their citations from. The findings reshape how you should think about “ranking.”
The Leapfrog Effect
More than half the domains Google’s AI cites don’t rank in the top 10 organically. For challenger brands, this is the first meaningful shift in search economics since PageRank. Ranking in the top 10 is no longer a prerequisite for being the authority AI quotes.
The median domain rank of a source cited by ChatGPT-style AI tools sits around 1,124, well outside the universe most SEO platforms even track. Translation: smaller, sharper sites are appearing next to major brands inside AI answers. Companies that have ranked well in Google for years can be invisible in AI responses. And the inverse holds. A well-structured site that never broke into the top 10 organically can become the source AI cites across its category.
For challenger brands, this is the first meaningful shift in search economics since Google rolled out PageRank. For incumbents, it’s a threat that requires a different playbook than the one that got them to page one.
After running optimisation work across hundreds of pages for our clients, the patterns are consistent. AI systems cite content that looks a specific way.
What doesn’t get cited: keyword-stuffed paragraphs, long-winded introductions, sales-heavy copy, and unstructured walls of text. The old on-page SEO checklist wasn’t wrong so much as it was optimising for the wrong reader.
One of our clients, a financial advisory firm, went from invisible on AI platforms to hearing “ChatGPT recommended you” from new prospects inside 82 days. The changes weren’t glamorous. We cleaned up entity signals so machines could read the business correctly (consistent NAP, interconnected schema, authoritative third-party mentions). We added Article, FAQPage, and Speakable schema across priority pages. We restructured existing content for citation rather than ranking, with answer-first paragraphs and question H2s. No hacks. Just the unsexy fundamentals done deliberately.
Related: the AEO and AI SEO blueprint we use with clients.
AI search moves faster than traditional SEO ever did. We published a post on a new ChatGPT feature and were showing up in Gemini and Perplexity within 24 hours, plus ranking #1 on Bing for the feature name. A podcast agency we work with published a piece on their pricing model and it surfaced on Perplexity the day after publishing.
Freshness is a real ranking factor for AI systems. Ahrefs analysed 17 million AI citations and found AI-cited content is on average 25.7% fresher than organic Google results (Ahrefs, 2025). ChatGPT specifically cites URLs that are hundreds of days newer than Google’s top organic results for the same query.
The implication for content strategy: frequency of fresh, well-structured content matters more now than it did in traditional SEO. The old model of publishing a cornerstone page once and maintaining it for three years doesn’t earn citations in AI systems. Regular publishing against questions your buyers actually ask, structured for extraction, compounds faster than any ranking strategy we’ve used in the last decade.
The measurement problem is the single biggest reason most marketing teams haven’t acted on this yet. Traditional rank tracking doesn’t show AI citations. Google Analytics doesn’t cleanly separate ChatGPT referral traffic. Here’s the practical stack we use.
The query fan-out trick. Perplexity shows you what it searched to answer a prompt. Open any Perplexity response and expand the “sources” or “steps” panel. Those searches are the queries your content needs to rank for to get cited next time. Reverse-engineer from Perplexity answers in your category, then build content against those specific query patterns.
AI visibility platforms. Dedicated tools like Profound, Peec AI, and Otterly monitor brand mentions across ChatGPT, Perplexity, Gemini, and Claude. Paid. Worth it once you’re actively optimising. A handful of free scanners exist for one-off checks.
The attribution workaround nobody talks about. Ask every new customer “how did you hear about us?” at booking or sale. Pair it with GA4 events on high-intent pages (pricing, services) and set up first-user source tracking for ChatGPT, Perplexity, and Claude referral domains. It’s imperfect, but combined with the direct question it gives you a defensible picture of AI-assisted customer acquisition.
A practical checklist for the next 90 days, ordered by impact.
sameAs attribute to your Organization JSON-LD, linking to your LinkedIn, Facebook, YouTube, Crunchbase, Wikipedia (if applicable), and Google Business Profile. This tells LLMs exactly which profiles belong to your business.sameAs to LinkedIn. AI systems weight content with clear authorship and author-organization connection far more than anonymous copy.Related: generative SEO pricing for Australian firms.
No, but it’s changed. Traditional organic search is down roughly 2.5% year-over-year globally (Search Engine Land, 2025), while AI-referred visits are growing triple digits. Traditional SEO still drives the majority of qualified traffic for most businesses. What’s changed is that ranking alone no longer guarantees clicks, and citation inside AI answers is now its own ranking goal.
Three levers in order of impact. First, publish content structured for extraction: answer-first paragraphs, question H2s, short paragraphs, FAQPage schema. Second, build entity signals so AI systems can identify your business correctly (consistent NAP, third-party citations, interconnected schema). Third, publish consistently. AI systems prefer fresh content, so regular publishing on relevant topics compounds faster than one-time optimisation.
Significantly higher than classic organic. Similarweb found AI-referred visitors convert 4.4x better than traditional organic visitors (Similarweb, 2025). That’s because AI-referred clicks are late in the buyer journey. The visitor has already compared options and narrowed their shortlist inside the AI conversation before clicking through.
Faster than traditional rankings. We’ve published content that appeared in Perplexity within 24 hours, and it’s common to see new pages cited in ChatGPT within a week if the content is well-structured and the domain has reasonable entity signals. AI systems weight freshness, so recently published content has a citation advantage that doesn’t exist in traditional Google rankings.
The question isn’t whether AI becomes the primary way buyers find you. It’s when. And whether you’ve done the quiet work today to be the undisputed answer in your category when that day arrives.
The data looks scary on the surface. The AI Overview click drops are real. So are the AI-referral gains and the 4.4x conversion uplift. Read together, they tell you the game is shifting, not ending. The businesses that will come out ahead are the ones that understood “not all stats are what they look like at face value” early, and built for where qualified traffic is actually going.
It’s not too late to start. It’s just late enough that starting now matters.
Traffic down 50%? Keywords collapsing? Before you panic, look at conversion. AI search is reshuffling who lands on your site, and the visitors still arriving convert 4.4x better than classic organic. Our read on the 2026 search shift.
Legal is the most expensive vertical in paid search and the most AI-disrupted in organic search. Our take on the PPC vs SEO mix that actually works for Australian law firms in 2026.
After six months using Claude for SEO, we’ve identified its strengths and weaknesses. It excels in trend analysis but often hallucinates, causing phantom canonical issues and missed schema detections. While it speeds up tasks, it can’t fully replace SEO professionals. Our review emphasizes the need for real context and output verification. If you’re a marketing manager aiming to enhance your team’s efficiency, read on to learn how to effectively use Claude while keeping a human verification step.
Subscribe to receive exclusive industry
insights & updates
Subscribe to receive exclusive industry insights & updates