10 June 2026

Google is building a search engine that already knows you

school of fish

From queries to context

Now that we have had time to digest Google I/O, the picture is clearer than it was on keynote day.

Most of the headlines fixated on Gemini 3.5 Flash and a reimagined search box, which is fair enough. Those are big announcements. But, the thread running underneath almost every announcement was something quieter and potentially more consequential.

Google is moving from a search engine that answers your question to one that understands your context before you finish asking it.

The word the company kept using was personalisation. The thing it actually unveiled was something closer to a search engine that arrives at the conversation already knowing who you are.

That matters for anyone working in paid media, because the surface we have optimised against for two decades is being rebuilt around the individual rather than the keyword.

We’ve already explored how Google’s latest Search updates are changing the way discovery happens online. What I/O revealed is the bigger picture behind those changes: why Google is redesigning Search around context, personalisation and AI-assisted decision making.

Here’s what changed, and why one of the more interesting takes came from a YouTuber rather than an analyst.

The biggest change to the search box in 25 years

Liz Reid, Google’s head of Search, called the new intelligent search box “the biggest upgrade to our iconic search box since its debut over 25 years ago.”

That is not marketing puff.

The box now dynamically expands as you type, anticipates intent, and accepts text, images, files, videos or Chrome tabs as inputs, reasoning across all of them at once.

The shift in behaviour behind this is the real story.

Google says Search has become less about individual queries and feels more like an ongoing conversation. The numbers back it up. AI Mode has passed one billion monthly users just a year after launch, with queries more than doubling every quarter and Search queries hitting an all-time high last quarter. Gemini 3.5 Flash is now the default model powering AI Mode globally.

For us, the takeaway is that keyword-shaped intent is being replaced by paragraph-shaped intent.

People are describing what they want in full sentences, with images and context attached, and Google is reasoning across the lot.

The query is no longer a string. It is a brief.

Personalisation is the actual product

This is where I/O got genuinely interesting.

Google announced it is expanding Personal Intelligence in AI Mode to nearly 200 countries and territories across 98 languages, with no subscription required.

Personal Intelligence lets you securely connect apps like Gmail and Google Photos, and soon Google Calendar, so Search can blend your personal context with general web knowledge.

The framing was all about control. Google stressed that connecting apps is opt-in and that you choose if and when you connect them. The intent, in Google’s words, is that for AI to be most helpful, it should not just know the world’s information, it should understand your context too.

Sit with what that means in practice.

Gmail is full of confirmation emails, bookings, receipts and the quiet record of your actual life. Once Search can read that context, a question like “what should I pack for my trip next month” stops being a generic answer and becomes an answer built around your flights, your destination and the weather on your travel dates.

Google itself describes this as going a step further than the usage-history personalisation that already adjusts logged-in results.

For paid media, this is the part to watch.

The personalisation that powers the experience runs on first-party data the advertiser never sees. Google is building a higher-resolution understanding of the shopper than the brand selling the product has.

The targeting advantage is no longer just intent signals in an auction.

It is the user’s own inbox.

Information agents and mini apps

Two more announcements pushed the personalisation theme into action rather than just answers.

Information agents are personalised AI agents you set up to work in the background, 24/7, to find what you need at exactly the right moment and help you take action.

The apartment-hunting example was the clearest one.

You brain dump your exact requirements, and the agent continuously scans for you, notifying you when something matches.

Google framed these as the next evolution of Google Alerts, except they run continuously in the cloud and deliver synthesised briefings rather than keyword matches.

Then there are mini apps.

For ongoing tasks, Search can now code custom dashboards or trackers that pull from real-time sources like reviews, live maps and weather data.

If you ask for a mortgage calculation, Google builds the calculator in front of you.

This is generative UI, and it is a direct shot at every utility site and comparison tool that has relied on Google for traffic.

Commerce is becoming a background task

On the commerce side, the same logic shows up as Universal Cart.

It lets users add products from Search, Gemini, YouTube and Gmail into a single persistent cart with deal tracking and compatibility checks.

Pair that with the updated Agent Payments Protocol, where agents can make purchases on your behalf within pre-set limits, including Human Not Present payments for things like limited-release tickets the moment they go on sale.

The path from discovery to purchase is being compressed to almost nothing, and Google wants to own every step of it.

The counterweight, from someone who watches this closely

Here is the counterweight.

I am an avid follower of Linus Sebastian of Linus Tech Tips, and his video on 20 May, titled Google’s Most-Hated Announcement Ever, captured the mood better than most of the coverage I read.

His framing was simple.

Google did not just add AI to Search. It put AI into everything.

Your search engine, your phone, your email, your calendar, your documents, your photos. Everywhere.

Whether you asked for it or not. That is the part worth sitting with.

The keynote was pitched as helpfulness. A lot of people experienced it as something being installed into their daily tools without an off switch.

Linus walked through the announcements one by one, from AI Search to Spark, Omni, Docs Live, shopping and the XR glasses, and the running thread was that the user is increasingly being moved from the centre of the experience to the passenger seat.

The AI does the searching, the drafting, the shopping and the deciding.

You approve. I think that lands for a reason.

There is a real difference between AI you choose to use and AI you cannot avoid.

When the personalisation is genuinely useful, people love it. When it shows up uninvited across every surface they touch, the same feature reads as intrusion.

The negative reaction to I/O was not really about the technology being bad. It was about consent and control.

People want to feel like they are still driving.

What this means for the rest of us

Put the pieces together and a clear direction emerges.

Google is making Search more personal, more conversational and more capable of acting on your behalf.

The query is becoming a brief. The result is becoming an experience.

And the purchase is becoming a background task an agent handles inside a mandate you set once.

For anyone managing paid budgets, three things are worth holding onto.

First, intent is moving away from keywords and towards context. The new ad formats announced at Google Marketing Live already place commercial content inside AI-generated answers. The unit of optimisation is shifting from the search term to the synthesised answer, and we will need to learn how to show up inside it.

Second, the value increasingly lives in first-party data. Personal Intelligence runs on the user’s own Gmail and Photos. Brands do not get that signal. The strategic response is to build your own first-party relationship that an agent has to read, because under agentic checkout the brand encoded into the mandate is the brand that wins the reorder.

Third, consent and control are becoming a real factor in how people respond to AI. The backlash that Linus captured is not about the technology failing. It is about the AI-everywhere default arriving without an off switch.

Some users want less AI, not more, at least until they choose it themselves.

That sentiment is something to plan around rather than dismiss, because the people who push back first tend to be the ones everyone else follows.

Google built its empire on a simple promise: you type a question and it returns a list of links.

At I/O 2026, it declared that promise is no longer enough.

For now, the safest assumption is that the search box is no longer a box. It is becoming an assistant that already knows you.

And marketers need to start planning for that reality.

Louder’s recommendations

  • Focus on building first-party relationships, not just first-party data: As AI agents increasingly influence discovery and purchasing decisions, brands need stronger direct relationships that create preference before an agent enters the process.
  • Prepare for a world beyond keywords: Search strategy should evolve from keyword targeting alone toward understanding audiences, context, intent signals and the information AI systems use to generate answers.
  • Strengthen measurement before journeys become less visible: As more activity happens within AI-powered environments, robust measurement frameworks, data collection and attribution foundations become increasingly important.
  • Prioritise trust, consent and transparency: Consumers are showing they value control over how AI uses their information. Brands should ensure personalisation delivers genuine value and remains grounded in clear consent practices.
  • Monitor how AI changes the path to purchase: Agent-assisted shopping, persistent carts and automated recommendations may fundamentally alter how consumers discover, evaluate and buy products. Brands should start testing for these behaviours now rather than waiting for adoption to mature.

Get in touch

Get in touch with Louder to discuss how we can assist you or your business and sign up to our newsletter to receive the latest industry updates straight in your inbox.



About Anmol Kumar

Anmol Kumar is a Paid Media Consultant at Louder. In his spare time, you’ll likely find him dancing Bachata, traveling, or soaking up the sun at the beach.