Project Title
Organic Traffic Decline – July 2025
Introduction
In early 2025, we noticed a puzzling trend in our analytics: impressions were rising, but traffic was falling. At first, we assumed it was a content or ranking issue — but the data told a different story. This post is the result of that investigation. What we uncovered was part of a much larger shift happening across the web, driven by Google’s rollout of AI Overviews and the more recent launch of AI Mode. These changes are fundamentally altering how users interact with search — and what that means for content creators, publishers, and digital teams like ours.
The Challenge
The initial trigger was a sharp decline in organic traffic to our site, despite stable or even improved rankings. Our impressions were up — in some cases, dramatically — but click-through rates were falling. This pattern wasn’t unique to us. As we dug deeper, we found similar stories across the industry. The culprit? A new era of “zero-click” search, where AI-generated answers satisfy user intent directly on the search results page, leaving little reason to visit the source.
My Role
As part of our digital strategy team, I contributed to the analysis of our traffic data and helped contextualise our findings within broader industry trends. Together, we reviewed external research, benchmarked our performance against peers, and shaped internal conversations about how we measure success and adapt our strategy. This post is part of that shared effort — to document, reflect, and share what we’re learning in real time.
Tools & Setup
- Google Search Console: To track impressions, clicks, and CTR across key queries
- GA4: For behavioural insights post-click
- Internal Dashboards: Custom visualisations of RHS traffic patterns
- Industry Research: Insights from Ahrefs, BrightEdge, BBC Future, and others
What We Observed
We didn’t set out to run a test — the test came to us. But we tracked and analysed the following:
- A year-on-year doubling of impressions for many RHS pages
- A simultaneous decline in clicks, especially for informational content
- A drop in CTR, even when rankings remained stable
- An increase in AI Overview citations, but with negligible referral traffic
What I Found
- Insight 1: AI Overviews now generate two impressions per query — once as a traditional link, and again as a citation — inflating visibility without increasing engagement.
- Insight 2: Our experience mirrors what Ahrefs calls “The Great Decoupling” — a growing negative correlation between impressions and clicks.
- Insight 3: With the rollout of AI Mode — currently live in the US and expected in the UK later this year — we anticipate this trend may accelerate. AI Mode replaces traditional search results with AI-generated summaries, further reducing the need for users to click through to websites 2.
What I Learned
- What worked well: Proactively investigating the data and validating our experience against broader industry trends helped us avoid internal misdiagnosis.
- What we’re considering: We’re exploring a shift in focus — from raw traffic to qualified engagement — and evaluating alternative discovery channels like social, email, and partnerships. At the same time, we’re closely monitoring how these changes evolve, especially as AI Mode becomes more widely adopted.
- What I’d do differently: Start tracking AI-triggered SERPs earlier and diversify our content distribution strategy sooner.
- Surprises: AI traffic, while lower in volume, may convert better — Ahrefs reports 23x higher conversion rates from AI search visits compared to traditional search.
Visuals
Clicks vs. Impressions: The Crocodile Chart
This line graph illustrates the growing disconnect between impressions (purple) and clicks (blue) from July 2023 to June 2025. While impressions have steadily increased — particularly after May 2024 — clicks have declined or remained flat, creating a widening gap. This pattern, often referred to as the “crocodile mouth,” reflects the impact of AI Overviews on user behaviour in search results.
Source: The Great Decoupling – Ahrefs
Google AI Mode: The Timeline
A timeline of key milestones in Google’s AI evolution — from the launch of its AI division in 2017 to the release of AI Mode in 2025. This visual contextualises the rapid rollout of AI-powered search features, helping explain the broader shift in user behaviour and organic traffic patterns discussed in this post.
Source: What Is Google AI Mode and How Does It Work? – Neil Patel
“You Can’t Un-Toast a Crumpet”: AI Explains the Unexplainable
A screenshot of a Google AI Overview attempting to define the made-up idiom “you can’t un-toast a crumpet.” The AI interprets it as a metaphor for irreversible decisions — a surprisingly thoughtful explanation for a completely fictional phrase. This visual highlights the quirks (and occasional charm) of AI-generated search results in the new era of machine-led information.
Source: Google AI Overview
Feedback
Have you seen similar patterns in your data? How are you adapting your strategy in response to AI-driven search?
Leave a Reply