What ‘Google Zero’ might mean for visitor attractions

In Disney-Pixar’s Monsters, Inc., doors fly through the Scare Factory, each one a gateway to a different world.

For twenty years, search has worked a bit like that factory: predictable gateways; reliable systems that manage the journeys.

But AI has slowly – and then very quickly – changed the game. 

The contract was once simple: create value and search engines would send you the audience). But search engines are now answer engines. 

Blue links, languishing below personalised answers from AI Overviews, are starting to feel like the ‘dead’ doors as kids outgrow being scared by the monsters. 60%  of searches are already ‘zero click’.

So, on the precipice of this profound shift, we gathered with digital leaders from visitor attractions and charities at London Zoo. We discussed the shared challenges and opportunities presented by AI Search.

Here’s what we learnt.

The reality check

Some reported organic search traffic drops of 30-50% since the rollout of AI Mode. This change has been more prevalent following the Google Core update rolled out in March 2025. 

Following the update, the SEO community reported a higher percentage of keywords in the UK showing AI Overviews (approximately 18%) which was reflected in a decrease in organic traffic.

Others shared frustrations about AI overviews that confidently state incorrect information about their services – like telling families that dogs are welcome at venues when they're not. When visitors arrive expecting something different, they don't blame the AI. They blame you.

Interestingly, some noticed more direct traffic, with visitors entering the sales funnel much later than usual. This shows how the entire marketing funnel is shifting. People now find and access information in increasingly complex and scattered ways.

Digital transformation has fractured consumer journeys into unpredictable, nonlinear patterns” (BCG.com) with AI only speeding this up. The solution? Understand what your audiences actually want and deliver that information wherever they happen to be looking.

The technical picture is messy too. Google Search Console doesn’t distinguish between traditional search clicks and AI Overview interactions, and GA4 only provides a limited view through source/medium insights – leaving everyone guessing about what's actually working.

There’s also confusion around whether JavaScript can be rendered by AI crawlers (Gemini, yes; ChatGPT, Perplexity and Claude, no). This means any content hidden behind client-side rendering is invisible to a significant portion of the AI landscape. Organisations are moving to ensure content is accessible in the initial HTML response.

Most striking was how few organisations have clear AI strategies in place. Teams are having to improvise approaches to these fundamental changes in how audiences discover them. That's not sustainable.

The way forward

We also uncovered some potential opportunities beneath the concerns.

One participant mentioned a 1.3% revenue increase from visitors referred through ChatGPT compared to traditional journeys. Another noted more engagement with their longer-form content since March – suggesting people still value depth when AI can only provide surface-level answers. Podcasts, in particular, seem to be holding their ground.

We’ve also heard that one participant is working on a chat functionality on their site to provide the same (and arguably improved) experience AI platforms provide to users on their own platform.

There’s promise in experimenting with content that AI can't authentically replicate (yet), and mirroring what AI does better than traditional websites. 

We know that, at present, the AI models use traditional websites to access information and so contemporary SEO practice is still relevant. E-E-A-T guidelines and structured data is more important than ever. 

And so, rather than scrapping the good work you’ve been doing, one of the key parts in responding to SEO is evolving your SEO strategy. 

Here are a few ways to start:

1. Create modular content: Create paragraphs/sections that can stand alone and be used by LLMs.

2. Focus on context (not query): Use semantically rich content focused on intent and written in natural language.

3. Prioritise E-E-A-T: Create content that shows expertise and is authoritative, providing original insight, verifiable data and citations. 

4. Focus on clear structure: Prioritise UX, page speed and clear structure that AI can easily understand. Use clear heading tags and apply Schema Markup.

5. Backlinks matter (again): Ensure links to your site from trustworthy sites, as well as referencing yourself on Wikipedia and Reddit.

The aim of your content should always be to provide value to users, aligning with their intent and needs. We should ensure that we’re meeting our audiences at the right time, with the right content and in the right place.

And, most importantly, this mission needs approaching with the agility and ingenuity that have served organisations well in previous transformation efforts.

After all, the doors to your world are still flying around the factory floor. They just work differently now.

Antony and Dominika work in digital consulting at manifesto, helping purpose-led organisations make big advances through digital transformation.