Google I/O is Google’s annual developer conference, held in Mountain View, California, where the company unveils its latest innovations in software, hardware and artificial intelligence. The 2025 edition showcased significant advancements, particularly in AI technologies that are set to transform search and user interaction.
To help make sense of the more than 100 Google I/O announcements, we asked Amanda Tietjen (Associate Director of Marketing – Search) and Tom Duffy (Head of AI Strategy) to break down what these shifts mean for brands, marketers and product teams.
From the end of the “ten blue links” to the rise of AI agents as customers, they outline how to adapt your marketing strategy for a search experience that is increasingly conversational, contextual and action-oriented.
What Google I/O 2025 Means for Organic Search Marketing
(Editor’s Note: These are key Google I/O 2025 insights from Amanda Tietjen, Associate Director of Marketing – Search)
This year’s Google I/O wasn’t just a showcase of AI bells and whistles. It was a blaring siren that the foundation of search is shifting fast. If your business relies on organic visibility, it’s time to take notice. The traditional “ten blue links” are no longer the centerpiece of the experience, and here’s the reality we all need to accept: we are never ever ever getting back together with the old SERP. Google is moving toward something far more dynamic, personal and task-driven. A new era. Your strategy needs to move with it.
Here’s what you need to know and what your brand should start doing to stay ahead.
AI Mode: Bringing AI Into Search
Google announced the rollout of AI Mode to all users in the U.S. You may have tried it out as a part of Google’s Search Labs in March. But now, it’ll start rolling out nationwide – no more beta programs or Search Labs opt-ins needed. Although it will roll out initially as a new tab within search results, Google will gradually integrate the capabilities into the core search experience.
Here’s Googe CEO Sundar Pichai revealing this feature:
The new interface is conversational and built for longer, more complex queries. Instead of showing a list of links, it generates synthesized answers drawn from multiple sources. And it’s far more powerful than the AI Overviews we’ve gotten used to over the last year. This new AI Mode has the ability to process more complex, multi-step, multi-modal queries, providing hyper-personalized responses.
The most fascinating “under-the-hood” part of this, especially to those of us in the content strategy game, is the “query-fan-out” method that it uses. When a user types a query, AI Mode doesn’t just do one simple search. Instead, it systematically breaks that question down into its core subtopics and then performs a multitude of related searches simultaneously.
Here’s a visualization of the “query-fan-out” method:
For nearly a decade, we at Noble Studios have advocated for a topic clustering model. (Peep this topic clustering blog post from 2018). And we’ve seen it result in 5X organic growth for clients. In a topic clustering model, you build a central “pillar page” on a broad topic and then support it with numerous detailed “cluster pages” on specific subtopics, all interlinked. This approach directly aligns with AI Mode’s new “query fan-out” behavior.
A site with well-executed topic clusters offers the AI a rich, interconnected content library that’s easy to explore and gather information from across many related facets of a query. This makes your site a highly authoritative and valuable source, increasing the likelihood of your content being chosen for AI-generated answers or referenced in search results.
Deep Search: Taking Research to the Next Level
Building on AI Mode, Google also introduced Deep Search, a feature specifically designed for more thorough research. Deep Search is integrated into AI Mode and uses the same query fan-out technique, but takes it to the next level. Instead of diving into a few queries, it issues hundreds of searches and “reason[s] across disparate pieces of information” to generate a detailed report, complete with citations and data to back it up.
Again, for SEOs, the implication is that content needs to be discoverable and provide unique, valuable information that this deep search process can utilize. You want your content to become the source material. That means your content needs to be well-structured, up-to-date and supported by signals that demonstrate trust. We’ve been talking about E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) for years. Now, it’s a prerequisite.
Hyper-Personalization: Persona is King (and Queen)
Perhaps one of the most profound, and a little unnerving, announcements from I/O 2025 is Google’s intensified ability to weave together AI-generated search answers with a user’s broader digital footprint across Google. It will use insights gleaned from Gmail, Google Maps history, YouTube watch history, and other products associated with your Google profile to personalize your search responses.
The AI will be crafting responses that are not just relevant to the query, but deeply personalized to the individual asking it.
- Example 1 – B2B SaaS: Imagine a B2B buyer researching CRM systems. If their Gmail reveals frequent correspondence with a specific software vendor, or if their Google Drive contains documents on sales pipeline optimization, Google’s AI could highlight CRM features that integrate with their existing tools and workflows. This leads to hyper-relevant recommendations that address their operational efficiency needs.
- Example 2 – Travel & Tourism: Consider a traveler whose Google Photos are full of recent beach getaways, and whose Google Maps history shows a strong preference for unique, boutique accommodations. When they search for a new destination, the AI tailors its recommendations to coastal adventures and off-the-beaten-path stays, making your content about those specific experiences critically important.
This means your content needs to be so finely tuned to your audience that it anticipates their personalized needs, even before they explicitly state them. This re-emphasizes that authentic, empathetic, and truly helpful content crafted for specific personas will be the most valuable asset in the AI-first search landscape.
If you haven’t already, it’s time to invest in deeply understanding your target audience and crafting your content strategy around them.
Agentic Search: From Informational to Transactional
AI Mode is gaining agentic capabilities, including the ability to book reservations, purchase event tickets or complete checkout with Google Pay, all within the search interface. Google is working directly with partners like Ticketmaster, StubHub, Vagaro and Resy to make this possible.
This means Google’s AI is moving from being a responsive chatbot to an active assistant, capable of understanding complex, multi-step intentions and taking actions on your behalf – possibly without you even leaving the search interface.
What could this look like in practice? For B2B software companies, this could mean an AI finding and comparing solutions, initiating demo requests, or even filling out lead forms on behalf of a user who has granted permission. For travel organizations, an AI could plan detailed itineraries, compare flights and hotels in real-time, and then proceed with booking actions directly within search.
What’s changing is that search is no longer just a place to find information. It is becoming the place where actions happen.
If your business wants to stay on the conversion path, you will need to explore ways to integrate with Google’s evolving agent model through APIs, schema or strategic partnerships. Ensure your site has structured product and service offerings, clear pricing and comparison data, and crawler-accessible forms. The goal is to be the preferred choice that Google’s agentic AI presents and acts upon.
Search Live: Combining Visual & Voice Search in Real Time
Another major announcement was Search Live, which allows users to point their phone at an object or scene, speak a question and get an answer in real time. This represents the full realization of multimodal search: visual, spatial and conversational.
Imagine holding up your phone to a broken air conditioner or a unique historical landmark in a new city and asking the AI questions about it – and getting immediate, contextual answers.
For content creators, this means not only optimizing for text, but also for the visual world and the specific, immediate questions users might have when observing something. Optimizing for camera- and voice-driven search means investing in high-quality imagery and videos, accurate metadata and robust schema markup.
Strategic Shifts to Follow
Search is evolving into a dynamic, personalized, and action-oriented experience. From AI Mode’s “query fan-out” and Deep Search’s research-grade synthesis to agentic capabilities and real-time visual queries via Search Live, the way users interact with information—and the expectations they bring—are shifting rapidly. The brands that will thrive in this new landscape are those that embrace AI-aligned strategies now: building high-quality, structured, persona-driven content and optimizing for every modality of search. The future of SEO isn’t about gaming the algorithm—it’s about becoming the best, most relevant answer in a radically reimagined search journey.
What Google I/O 2025 Means for AI
(Editor’s Note: These are key Google I/O 2025 insights from Tom Duffy, Head of AI Strategy)
I usually tune out the hype of tech keynotes, but Google I/O 2025 caught my attention. The most important thing I noticed wasn’t how frequently AI was discussed, it was how literally everything presented and announced was AI or had an AI component.
I try to focus on what’s practical, transformative and forward-moving. I/O 2025 definitely delivered. AI is becoming the new operating system, and it feels like Google will be using Gemini 2.5 as the foundation for all things.
Here’s what hit me the hardest from Google I/O 2025:
Project Astra: The Universal Assistant
Project Astra aims to build an AI assistant that understands its surroundings. It speaks, remembers and interacts with the world. In the Travel & Tourism industry, DMOs will need to pay close attention to this. Visitors will increasingly rely on their personal AI assistants (like Astra) to ask questions, plan trips, discover activities and make bookings.
Here’s a demo of Astra:
Astra, using its advanced understanding and personalization, will be the one curating and presenting options. DMOs need to ensure their destination and its offerings are what tools like Astra finds and deems most relevant and appealing for its user.
Project Mariner: AI That Acts
Mariner lets AI multitask and learn by doing. It marks a step toward travel planners that handle bookings and rebookings on their own. B2B SaaS platforms should be considering a new customer class for their ICP: AI Agents. Agentic platforms like Mariner need access to tools, so if you’re a software company, you need to start looking beyond humans in your ICP.
Local Models and AI on Edge Devices
Edge AI matters when speed and privacy matter. Google unveiled several new versions of DeepMind’s Gemma 3 model. Gemma 3n is particularly interesting. It uses a new architecture that allows for high intelligence on low powered hardware. If privacy is key, local is the way. Get the power of a large language model with none of the data privacy and security concerns. SaaS apps can process data locally, helping with security and cutting server costs.
Gemini Live: Better Conversations
Gemini Live makes voice interactions smoother and more natural. The long-promised and previously overhyped real-time voice assistant concept is finally being realized.
Travel brands can use it for guided tours. If the assistant is not only able to interact with you in real-time through voice, but is also aware of what you are seeing, it becomes the ultimate tour guide. You will be able to just ask questions about what you’re looking at and Gemini has all of that as context, can access the internet and its training data for answers, and can be customized by a destination with key facts, brand information and specific details about a location that might not be easily found. SaaS companies can build helpful voice-based support or onboarding assistants.
Veo 3: Smarter Video Creation
Veo 3 adds sound and dialogue to AI video. This might not seem like a big deal, but it is. For anyone who uses AI video generation regularly, the process of having to split your workflows to handle audio separately is annoying. It was only a matter of time before someone released a model that does both and Veo 3 is really impressive.
The ability to quickly create immersive stories means that a lot more time and attention can be spent on the part that matters most. The story itself. AI still can’t hold a candle to humans when it comes to taste and creativity. The less execution work humans have to do, the more they can focus on the part that makes them amazing.
Flow: AI for Scene Creation
Flow lets creators build scenes with text. It lowers the bar for making great content. Travel and SaaS brands can quickly test new video ideas or explainers. Not every video needs to be pulled into Adobe Premiere by a professional editor. Sometimes you just need something quick to make a point or prove a concept. That’s where Flow really shines.
Live Translation in Google Meet
Real-time translation helps teams work across borders. Since the early days of ChatGPT, we’ve been promised real-time voice translation and every attempt has been lackluster. Watching Live Translation in Google Meet is the closest we’ve come to having this promise fulfilled. Teams with international members can now just speak in their native language comfortably and the meeting will speak the translated version in their voice. It even matches emotional inflections pretty well.
Here’s is Google CEO Sundar Pichai revealing Live Translation:
Strategic Shifts to Follow
Agentic AI doesn’t just answer questions. It gets things done. This will be extremely impactful in almost every industry out there. We are getting closer and closer to a world where most people have a digital agent that curates products, services and travel recommendations for them. These agents will only look for brands that truly align with the person they are working for. Brand will be more important than ever. Every business will need to make sure they are doing everything it can to align with its ideal customers.
Google didn’t just show off new tools at I/O 2025. It showed us that AI isn’t an add-on anymore. It’s the frame everything now sits on. If you’re building a brand or a product, the message was plain: the screen isn’t the main event. The AI is. It’s how people will find you, talk to you and remember you. Start there.
Time to Hug Change
Google I/O 2025 made one thing clear: AI is no longer an add-on to search. It’s the core experience.
From AI Mode to Deep Search to task-completing agents, Google is changing how people find, evaluate and act on information. If your strategy still relies on traditional rankings or outdated content tactics, it’s time to shift.
At Noble Studios, we pride ourselves on being Change Huggers. We lean into what’s next, not what’s comfortable. We keep our knees bent, and find our sea legs fast. If your brand is ready to evolve with the future of search and stay competitive in an AI-first world, let’s talk.
