Google I/O & Google Marketing Live: What's Really Changing for Retailers and Shopping Ads

Google is no longer just changing its search engine. It is progressively redefining the entire shopping journey : discovery, product recommendation, advertising, landing page… all the way to payment.

Between Google I/O and Google Marketing Live, one thing became clear: we are shifting from a keyword-based search to a model of agentic commerce, driven by AI.

For retailers and Paid Media teams, the question is no longer “should we pay attention to these announcements?”.

The real question is:

What does this actually change for my Shopping campaigns and media profitability?

Spoiler: probably more than you imagine.

Why Google is no longer just talking about Search, but about the shopping journey

For a long time, Google separated the stages:

• Search for discovery
• Ads for acquisition
• The retailer's site for conversion

This boundary is becoming blurred.

At Google I/O, Google confirmed a major evolution: Search is becoming conversational, multimodal, and agentic.

Specifically, users no longer just type: “men’s running shoes”

They ask: “I’m looking for a lightweight pair for running 3 times a week, decent cushioning, under €150, fast delivery.”

This changes everything.

Because Google no longer just has to match keywords. It must now:

• understand the intent,
• contextualize the need,
• select relevant products,
• explain its recommendations,
• and sometimes directly facilitate the purchase.

For retailers, this means one simple thing:

Visibility increasingly depends on Google's ability to understand your products.

And no longer just on your bid.

Search is becoming conversational: why this changes the rules of Shopping

One of the most important messages from Google I/O probably went a bit under the radar:

queries are becoming much longer and more complex.

With AI Mode, Google is already observing conversational searches that are several times more detailed than before.

A user no longer expresses a keyword.

They express a need.

This poses an immediate problem for retailers:

What happens if your catalog doesn't contain the necessary signals?

Simple example.

A user searches for: “light dress for a June wedding, budget €150, small size, bohemian style”

If your feed only contains:

Title:
“Women's beige dress”

Google lacks signals.

Conversely, a catalog enriched with: cut, occasion of use, material, seasonality, body type, style, color nuances, provides much more data for AI systems.

In an AI-first world: The product feed gradually becomes a semantic understanding engine.

Ads become agentic: the gradual end of keyword-driven management?

Google Marketing Live has formalized a development already visible for several months:

automation is no longer a “plus”.

It becomes the norm.

AI Max: towards more intelligent query matching

With AI Max, Google takes the Performance Max logic applied to Search and Shopping even further.

Objective:

to find distribution opportunities that manual targeting would never have captured.

In other words:

Google no longer simply matches a query to a keyword.

It interprets:

• the context,
• the intent,
• the behavior,
• product signals,
• conversion probabilities.

For Paid Media teams, this changes the game.

The focus is gradually shifting:

Before:
campaign structure → keywords → bids.

Tomorrow:
data quality → business signals → AI arbitration.

The real question becomes:

Are my products understandable by the algorithm?

Ads in AI Mode: when advertising becomes an answer

Google is also testing new ad formats directly integrated into AI conversations.

Advertising is becoming less "banner-like."

And more contextual.

Example:

A user asks:

“Which sofa is suitable for a small living room with a dog?”

Tomorrow, Google could:

1. understand the need,
2. recommend products,
3. explain why they are relevant,
4. integrate a native ad within the response.

This means one thing:

The creative role of the ad is changing.

The product, its attributes, and its contextualization become more important than the advertising mechanism itself.

Ask Advisor: the new Paid Media copilot

Google is also pushing an agentic assistant capable of:

• analyzing performance,
• detecting opportunities,
• recommending optimizations,
• automating certain decisions.

For acquisition teams, this doesn't mean the disappearance of the profession.

But probably a shift in value.

Less time spent on:

• manual optimizations,
• campaign micro-adjustments.

More time for:

• product strategy,
• profitability,
• business signals,
• the quality of assets and feeds.

Commerce goes native: towards frictionless purchasing

This is probably the most underestimated announcement.

Google is closely integrating:

Search + recommendation + purchase

with:

• Universal Cart,
• Direct Offers,
• simplified checkout,
• integrated Google Pay.

The idea is simple:

if Google already knows:

• your need,
• your budget,
• your preferences,
• your history,

why make you go through 8 steps?

Purchasing is becoming increasingly native.

We search.

We compare.

We validate.

Without leaving the Google environment.

For retailers, this creates a paradox:

More potential conversion… but less control

Catalog quality becomes critical.

Because a poor product signal can now break an entire transactional opportunity.

Some frequently observed examples:

Out-of-sync prices

The advertised price no longer matches the website.

Inconsistent variants

Google promotes an unavailable size.

Incomplete attributes

The AI doesn't truly understand product usage.

Irrelevant merchandising

Recommendations lack context.

In an agentic world:

Incorrect product data can potentially cost more than a bad bid.

What's the practical impact for retailers?

Specifically, three major consequences are emerging.

1. The product feed becomes a business infrastructure

For a long time, the product feed was seen as a technical or media activation topic.

In the future, it will become a decision-making infrastructure.

It powers:

• Google Shopping
• PMax
• AI Mode
• AI recommendations
• shopping agents
• conversational interfaces

A poor feed inherently reduces your ability to be recommended.

2. Generic PDPs become a problem

If Google generates an ultra-contextual response…

Why send a user to a generic page not aligned with their needs?

Example:

Search:

“beginner running shoes fragile knees”

Landing page:

standard product grid without context.

Immediate disconnect.

The more personalized the ad, the more consistent the post-click experience must be.

Contextual continuity becomes a conversion driver.

3. Paid Media becomes more strategic

Acquisition teams will likely optimize campaigns less...

and more optimize decision systems :

• catalog quality,
• merchandising logic,
• business signals,
• margins,
• product segmentation,
• post-click consistency.

The profession isn't disappearing.

It's evolving.

What Paid Media Teams Need to Prepare Now

Not in 3 years.

Not "when Europe catches up."

Now.

1. Audit the Depth of Product Attributes

Ask yourself a simple question:

Does an AI truly understand my product?

Or just its name?

2. Treat Merchant Center as a Business Driver

Merchant Center is gradually becoming an AI-powered visibility hub.

It's no longer just about compliance.

It's about product understanding.

3. Rethink the Post-Click Experience

The more Google contextualizes...

the more the landing page must remain consistent with that promise.

Otherwise, friction increases.

4. Prepare for a World Without Dominant Keywords

The future of Shopping will likely be less driven by:

"which keywords to buy?"

And more by:

“what product signals should be transmitted to AI?”

Key takeaways

• Google is transforming Search into a conversational and agentic engine.
• Shopping Ads will rely more on the quality of product data.
• AI Max accelerates the end of keyword-centric management.
• Ads are becoming contextual and integrated into AI conversations.
• Universal Cart and native checkout bring Google closer to the final transaction.
• Retailers will need to invest as much in product understanding as in media.

One thing seems increasingly clear: the brands that will perform best tomorrow won't just be those that invest the most in acquisition.

They will primarily be those whose products are best understood, contextualized, and activated by AI systems.