Google traffic is falling. AI-generated answers are siphoning clicks before users ever reach a website. For our clients, this isn’t a distant threat—it’s happening right now, and we’re already restructuring how we approach search visibility. Here’s what the latest industry data tells us and exactly how we’re responding.

Key Takeaways

• Organic traffic from Google is in measurable decline, and businesses need diversified visibility strategies immediately.
• Schema markup has shifted from a nice-to-have to a critical requirement for AI citation eligibility.
• AI visibility gap analysis is now a standard part of competitive auditing.
• Advanced SEO in 2025–2026 means optimising content specifically for AI extraction, not just traditional rankings.
• Keyword cannibalisation doesn’t just hurt rankings—it actively undermines your chances of being cited by AI systems.

Google Traffic Is Dropping—Diversify or Fall Behind

The numbers are clear. Click-through rates from traditional Google results are shrinking as AI Overviews, featured snippets, and zero-click answers expand. We’ve been tracking this across our client portfolio for months. As outlined in this detailed breakdown on marketing success in the age of AI, the businesses that survive this shift are those building owned audiences—email lists, communities, and direct brand demand—alongside their organic search efforts.

Our recommendation to every client: stop treating Google as your sole traffic source. We’re investing in content that earns direct visits, newsletter sign-ups, and brand recall. Organic search remains valuable, but it’s no longer the only game worth playing.

Schema Markup Is Now Non-Negotiable for AI Discoverability

AI systems like Google’s Gemini, ChatGPT, and Perplexity need structured data to understand what your pages actually contain. Without schema markup, you’re invisible to them. We’ve integrated structured data audits into every new project and maintenance cycle, following the practical standards detailed in this guide on what schema markup is and how to add it to your site.

At minimum, every client site we manage now carries:

• Organisation and LocalBusiness schema
• FAQ and HowTo schema on relevant content pages
• Article schema with author and publisher markup for E-E-A-T signals

If your site doesn’t have structured data in place, AI engines have no reason to cite you.

AI Visibility Gap Analysis Is Part of Every Competitive Audit

Traditional keyword gap analysis tells you where competitors rank and you don’t. AI visibility gap analysis tells you where competitors get cited in AI-generated responses and you don’t. That’s a fundamentally different dataset, and we’re now running it as standard. The methodology we follow mirrors the process set out in this walkthrough on finding AI visibility gaps with Semrush.

The insight is actionable. When we identify prompts where a competitor is cited but our client isn’t, we create or restructure content to directly address those queries with clear, extractable answers.

Optimising for AI Extraction Requires a Different Content Architecture

Writing for humans still matters. But content also needs to be machine-readable in a way that goes beyond basic on-page SEO. We’re restructuring client content to use concise, factual answer blocks, clear heading hierarchies, and definitive statements that AI models can pull directly. These are among the advanced SEO techniques and strategies for 2026 that we’ve already adopted across active campaigns.

Crawl depth improvements, internal linking overhauls, and content consolidation are all part of this work. If a page is buried four clicks deep, neither Googlebot nor an AI crawler will prioritise it.

Keyword Cannibalisation Kills AI Citations

When multiple pages on your site target the same keyword, search engines struggle to pick a winner. AI systems face the same problem—and they simply move on to a competitor with a single, authoritative page. We’ve seen this pattern repeatedly. Our fix follows the structured approach in this resource on how to find, fix, and prevent keyword cannibalisation: audit, consolidate, redirect, and monitor.

Every client content audit we run now flags cannibalisation issues as a top-tier priority, not an afterthought.

The shift from traditional SEO to generative engine optimisation isn’t theoretical. It’s measurable, it’s accelerating, and the agencies and businesses that act on it now will hold the visibility advantage for the next several years. We’re building every client strategy around this reality.

Frequently Asked Questions

What is generative engine optimisation (GEO)?

GEO is the practice of optimising your website content so that AI-powered search engines and chatbots cite and reference it in their generated responses. It builds on traditional SEO but focuses specifically on structured data, clear answer formatting, and authoritative content that AI models can extract.

How do web designers make sites visible to AI Search engines?

We implement schema markup, build clean heading hierarchies, and structure content into concise, factual answer blocks. These technical foundations make it straightforward for AI crawlers to parse, understand, and cite your pages in generated results.

Why does keyword cannibalisation hurt AI visibility?

When multiple pages compete for the same query, AI systems can’t identify a single authoritative source on your site. They default to citing a competitor with one clear, consolidated page instead.

How can I check if my competitors are being cited by AI and I’m not?

AI visibility gap tools compare your site’s citation presence against competitors across specific prompts and queries. Running these audits regularly reveals exactly which topics need new or restructured content to earn AI citations.

What is schema markup and why does it matter for SEO in 2025?

Schema markup is structured code added to your site that tells search engines and AI systems precisely what your content represents. In 2025, it’s essential because AI models rely on structured data to decide which sources to cite in their generated answers.