Google unveiled something at I/O 2026 that most marketers glossed over. Not AI Mode. Not Gemini 3.5 Flash. Something smaller – but in some ways more consequential.
They redesigned the search box.
They’re calling it the “Intelligent Search Box.” It sounds like a product name someone came up with in a branding meeting. But the mechanics behind it are worth paying close attention to – especially if you run a tourism website, a hotel, or a destination that depends on organic search traffic.
What Actually Changed? (And What Didn’t)
The old Google search box was a single-line input field. You typed 3-5 keywords, hit enter, got a list of results. That interaction model hadn’t meaningfully changed since 1998 – nearly three decades of the same basic interface.
The new one is different in three specific ways.
- First, it expands dynamically as you type. It’s designed for longer, conversational prompts – full sentences, questions, context – not just keywords. The box itself is a visual invitation to say more.
- Second, it has an AI-powered suggestion layer. Google’s Head of Search Liz Reid described it as going “beyond autocomplete.” Traditional autocomplete predicts the end of your query based on what others have searched. This goes further – it’s prompting you toward richer, more complex requests that happen to work better in AI Mode than in traditional search.
- Third, you can now input more than text. Images, files, videos, and even your open Chrome tabs can go directly into the search field. You can point at a hotel you’re looking at in another tab and ask a question about it. You can upload a photo of a dive site and ask about conditions. The query interface is no longer just a text field.
All of it powered by Gemini 3.5 Flash.
Google Search box
This Isn’t a UX Change. It’s a Behavior Change Mechanism.
Here’s what most of the coverage is missing: Google didn’t redesign the search box to make it look nicer. They redesigned it to change how people search.
Longer, more conversational queries trigger AI Overviews. Short keyword queries often don’t. Google knows this. The Intelligent Search Box is engineered to push users toward the query style that feeds the AI-first experience – and away from the query style that produces ten blue links.
Think about it from a user behavior perspective. Someone planning a diving trip in the Maldives used to type “Maldives diving” and scan a results page. Now the search box is visually inviting them to type “best dive sites in the Maldives for beginners in November with liveaboard options.” That query doesn’t produce a list of websites. It produces an AI Overview. And if they ask a follow-up, they’re automatically moved into AI Mode.
Moving from AI Overviews to Full AI
That seamless AI Overviews → AI Mode transition? Google also made that globally live this week on both desktop and mobile. It had been in testing since January. Now it’s everywhere. When someone gets an AI Overview and asks a follow-up question, they’re automatically moved into AI Mode – no extra click, no opt-in, it just happens.
The full funnel now looks like this:
Short keyword → Intelligent Box nudges toward longer query → longer query triggers AI Overview → follow-up question → automatic AI Mode → Google completes the task → no click to your website
Every single step in that chain reduces the probability of a user reaching your pages.
What Sundar Pichai Actually Said
At Google I/O, Sundar Pichai made a statement that deserves more attention than it got:
“Search is evolving – from individual queries to ongoing conversations and now to agentic workflows.”
Most people stopped at “ongoing conversations” – that’s been the story for a while. The phrase to focus on is “agentic workflows.”
An agentic workflow isn’t just a conversation. It’s a task completed. Google isn’t positioning Search as the place you go to find information anymore.
They’re positioning it as the place you go to get things done – without leaving Google. Find a hotel, compare prices, check availability, make a booking. Research a destination, build an itinerary, find flights. All inside Google.
That’s the roadmap. The Intelligent Search Box is the front door to it.
What This Means for Tourism Websites Specifically
The impact isn’t uniform across all content types. Let’s be specific about what’s at risk and what isn’t.
Content that’s most at risk:
Informational content – “best time to visit Bali,” “what to pack for a safari,” “diving certification requirements,” “Maldives vs Thailand for snorkeling” – is the most directly exposed. These are exactly the queries that the Intelligent Search Box is designed to pull into AI Mode. Google has the information to answer them. They will.
General destination guides with no unique angle are the next category. If your “Complete Guide to Scuba Diving in Koh Tao” covers the same ground as every other guide on the topic, there’s no reason for Google to send a user to your page when it can answer the question directly.
Content that’s less at risk:
First-party content built on your own data and experience is harder to replace. Case studies, audit findings, real numbers from your clients, dive site conditions you’ve observed personally – Google cannot synthesize that from the web because it doesn’t exist on the web until you publish it.
Transactional content remains relatively protected. Booking pages, availability searches, price comparisons, itinerary builders – these require the user to interact with your system, not just consume information. Google can describe them but can’t complete them on your behalf. Yet.
Local intent queries still send traffic. “Dive centre Koh Tao” or “boutique hotel Hvar” still requires a local destination. Google will show a local pack, a GBP listing, review scores – but the booking still happens on your site or through a direct channel.
Content built around specific expertise – your perspective on something – holds value longer than content that’s just an aggregation of facts. “Why I recommend liveaboards over land-based diving for first-timers in Raja Ampat” is a point of view. AI can paraphrase it, but it can’t replace the authority that comes from you having actually dived there.
Google Intelligent Search Box (Source: Google)
The Citation Layer: The New SEO Game
Here’s a strategic point that most hotel marketers and DMO teams aren’t thinking about yet.
Even when Google answers a question in an AI Overview, it cites sources. Those citations get clicks. Not as many as ranking position one used to get – but they get clicks from users who are already engaged and want to go deeper.
Optimizing to be cited in AI Overviews is different from optimizing to rank. It requires:
- Structured, factual content – clear statements that can be extracted and attributed
- Entity clarity – Google needs to understand unambiguously who you are, where you are, what you do
- Consistent presence across the web – mentions, citations, reviews, GBP data all contribute to whether Google considers you authoritative enough to cite
- Original data or perspectives – AI systems prefer to cite something that adds to the answer, not just restates it
For a dive center, this might mean publishing actual visibility data for your dive sites by month. For a DMO, it might mean publishing original research on visitor behaviour. For a hotel, it might mean a genuinely detailed guide to your local area written by someone who actually lives there – not a generic neighbourhood overview.
A Practical Framework for What to Do Now
This isn’t a crisis. It’s a transition. The sites that treat it as a crisis will make reactive decisions. The ones that treat it as a transition will adapt methodically.
Here’s how to think about it:
- Step 1: Know your exposure. Open Google Search Console. Look at your top 50 informational queries by impressions. Cross-reference them with your click-through rates. If you have queries getting hundreds of thousands of impressions and CTRs below 2%, you’re already experiencing AI Overview suppression. Understand the size of the problem before you react to it.
- Step 2: Identify what you have that Google doesn’t. This is the most important question. What data, experience, or perspective lives in your organisation that can’t be scraped from the web? Client results. Operational knowledge. Local expertise. First-hand experience. That’s your content moat. Build into it.
- Step 3: Shift your content mix toward tasks, not topics. “What is X” is now Google’s job. “Here’s how to do X in a way that leads to booking with us” is yours. Practically: more buyer journey content, more comparison content with a conversion endpoint, more content that exists at the bottom of the funnel rather than the top.
- Step 4: Optimise for citation, not just ranking. Structured data matters more than it used to. Clear, factual, well-attributed content matters more than it used to. Being a recognised entity in Google’s knowledge graph matters more than it used to. This is GEO – Generative Engine Optimisation – and it’s now a real part of the job.
- Step 5: Watch your branded search closely. When AI Mode recommends your hotel or tour operation by name, users will often do a follow-up branded search to validate. Your ability to control what appears on that branded SERP – your GBP, your reviews, your own site – is now a conversion layer, not just an SEO metric.
The Honest Take
The search box change is not going to destroy organic traffic overnight. Google Search still processes billions of queries a day and the majority of them will still route users somewhere. The economic model of web publishing hasn’t collapsed yet.
But the direction is clear. Google is systematically building a product that answers questions rather than routes users to answers. The Intelligent Search Box is the latest piece of that – it’s the input layer designed to make AI-first search feel natural rather than like a mode you have to switch into.
The New Panel and the Tourism Industry
For tourism specifically, the good news is that the industry has structural advantages: local intent, transactional complexity, emotional decision-making, real-world physical experiences. These are all things that resist full AI displacement. You can’t book a feeling through a chat interface.
But the informational layer – the part of tourism content that explains, describes, compares and informs – that layer is going to get thinner. The sites that have built their traffic on being the best at explaining things are going to find it harder to hold position.
The sites that have built their traffic on being genuinely authoritative, locally expert, and connected to the actual booking decision? Those are the ones that come out of this transition in reasonable shape.
The search box hasn’t looked like this in 25 years. Make sure your content strategy doesn’t still look like it’s from 25 years ago either.
Questions about how this affects your specific situation – hotel, DMO, tour operator, dive centre? Let’s talk.
About the Author
I’m Peter Sawicki, a Destination SEO Strategist helping tourism brands and DMOs grow their online presence through SEO, technical audits, and creative digital strategies. Over the years I’ve worked across multiple countries and markets, which gives me a global perspective on every project I take on. When I’m not optimizing websites, you’ll most likely find me underwater. Scuba diving is where my two biggest passions meet.
