Search Will Never Be the Same for Commerce Marketers (Or Anyone Else)

After watching the Google I/O 2025 keynote, it was clear that something fundamental had shifted. Google is signaling a new era—where shopping campaigns become more visual and dynamic, YouTube becomes shoppable, and AI Mode redefines how discovery works. Keywords aren’t going away, but they’re no longer the center of the strategy. This moment got us thinking: the search landscape is being re-architected in real time—and it’s not just a Google shift. It’s a catalyst for retail marketers to rethink how discovery is being shaped across every touchpoint.

How should retail marketers rethink discovery in the new search moment?

For commerce teams, discovery now starts well before a typed query and spans social, retail media, and AI assistants.
Treat discovery as intent-building across surfaces.
Winning programs align structured product data, helpful content, and measurement so AI systems can surface you earlier in the journey.

Search doesn’t just start on Google anymore—and even Google is signaling that the old rules don’t apply.

For more and more shoppers, it starts with a swipe on TikTok, a scroll through Instagram, or a typed question in ChatGPT. According to Pew Research 2025, social platforms (including TikTok) are rapidly becoming entry points for information—especially for under-30s—shifting where brands must “earn” visibility. And when it comes to product discovery, those moments are starting to matter more than the browser. Generative AI tools like ChatGPT aren’t just experiments—they’re quickly becoming trusted commerce companions, helping users make shopping decisions earlier in the journey. Traditional search engines aren’t dead—they’re just no longer the default starting point for many shopping journeys.

That shift has big implications for retail marketers.

We’ve been trained to work in structured, measurable environments—campaigns defined by keywords, CPCs, feed quality, and ROAS. But today, those levers don’t tell the whole story. Consumers aren’t waiting to search. They’re discovering products earlier, in ways that don’t always include a click—or even an ad. Which means visibility can’t just be something you buy. It has to be something you build across a wider surface area.

Paid and organic search can no longer be defined solely by Google and Bing because the consumer journey now begins wherever attention lives—on TikTok, Amazon, ChatGPT, Instagram, emerging AI interfaces, and even voice assistants like Alexa and Siri. These platforms are not just discovery channels. They’re intent engines, surfacing product answers before users ever open a browser.

And they’re changing the rules faster than most teams can keep up. If your team manages paid search at scale, anchor that work in an enterprise search optimization motion that treats Google/Microsoft as part of a broader discovery system.

Why is product discovery moving upstream—and what should we change?

Discovery happens before keywords — signals and content shape intent upstream.
Lean into upstream content that AI can cite.
Prioritize structured feeds, rich PDP content, and creator-grade assets so social/AI surfaces can recognize, rank, and recommend your products.

AI tools like ChatGPT now field over 517 million visits per month in the U.S. alone—surpassing Bing’s 376.6 million in December 2024. What used to be a job for search engines—comparison, validation, even recommendation—is now something consumers do with a single prompt. And these results aren’t just scraped from blogs. They’re curated using real-time retailer data, then refined by human editors to improve trust and ensure relevance.

In April 2025, ChatGPT received over 1 billion search visits—outpacing Bing and cementing its role in the discovery landscape.

At the same time, social platforms have become powerful intent indicators. 49% of Gen Z now starts their product discovery on TikTok or Instagram. This is no longer just a behavior trend—it’s a generational baseline. And while it’s easy to dismiss social-first discovery as top-funnel noise, the reality is that people are buying what they see, without ever entering a traditional search query.

Retail media’s role is shifting too. Amazon’s advertising revenue hit $56.2 billion in 2024, which now rivals or surpasses some of the original digital giants. This isn’t just about scale—it’s about trust. When shoppers are ready to act, they go to places that blend content and commerce seamlessly. Retailers have become search engines in their own right.

None of these changes are hypothetical. They’re already happening. And they’re changing where and how products get found. To capture retail demand, connect upstream discovery to retail outcomes with retail media solutions that unify marketplace advertising and the digital shelf.

Has discovery outgrown the keyword—and what replaces it?

Google is shifting to AI-first experiences that reduce your keyword control.
Feed the algorithm, don’t fight it.
High-quality assets, audience signals, and landing pages that map to intents will outperform micro-granular keyword hacks as AI composes experiences.

Google I/O 2025 made one thing clear: the traditional keyword-first model of search marketing is being phased out. In its place, Google is introducing an AI-first architecture that touches nearly every part of the funnel. Shopping campaigns are getting creative upgrades, with video assets now supported in Google Merchant Center. YouTube is becoming a fully shoppable platform, blending upper-funnel engagement with lower-funnel action. And Google’s new AI Mode is reshaping what it means to “search”—from discreet queries to fluid, conversational interactions.

What this means for marketers is a complete redefinition of how intent is captured. Performance Max campaigns don’t use keywords at all, and broad match is encouraged as the default. Instead of building campaigns around exact terms, marketers will be leaning more on contextual signals, inferred intent, and AI-curated journey stages. Messaging and landing pages will be dynamically tailored by Google based on real-time behavioral insights.

Another noteworthy announcement was Google’s introduction of AI Max for search campaigns, which is a clear signal of where things are headed. While it’s still early, the goal is to combine the automation of Performance Max with more search-specific control, giving marketers a way to reach high-intent audiences without relying on traditional keyword strategies. As this evolves, it could become the bridge between what search marketers want (transparency and control) and where Google is clearly pushing the ecosystem (intent-first, creative-led, AI-optimized).

It’s also worth noting that these shifts aren’t just about formats—they’re about outcomes. Google is repositioning its ecosystem to deliver results across the entire journey, not just the moment of conversion. By blending YouTube, search, and Shopping into a more cohesive engine, they’re aiming to break down channel silos and create a programmatic, responsive system that serves the right message at the right time, on the right surface.

For search marketers, this is a moment of reckoning. Keyword management won’t disappear overnight, but its role is shifting. Campaign structure will become less about granular controls and more about preparing high-quality assets and signals that AI can use to assemble effective experiences. The winners in this next chapter won’t be the ones who master match types—they’ll be the ones who feed the algorithm with the best possible inputs and let AI do the orchestration.

What must search marketers change now to stay visible and measurable?

Focus on inputs you control: product data, creative, signals, and cross-surface measurement.
Work your controllables relentlessly.
Standardize feeds and PDPs, map messaging to journey stages, and triangulate impact with experiments and proxy signals when last-click isn’t available.

This new landscape demands new thinking. We can’t just optimize for paid placements anymore. We need to rethink what visibility actually means—and how it gets earned across different discovery surfaces. Retail media is now a core scale channel—Amazon’s ad business reached $56.2B in 2024, underscoring where commerce intent gets monetized, per Adweek 2025. Broader commerce media is projected to exceed $100B in the U.S. by 2027, per McKinsey 2025.

Get the product data right

It sounds basic, but it’s critical. Product content is the input for every discovery experience—whether it’s a search engine, a social carousel, or a generative response in an AI assistant. If your metadata is messy or incomplete, you’re invisible. That includes images, descriptions, attributes, and feeds—every field matters when the platform doesn’t give you a second chance.

Think beyond the ad unit

Most of these discovery channels aren’t pay-to-play. At least not yet. So content quality becomes the currency. Can your product surface organically in a list of recommended items? Can it be recognized and ranked by an AI interface? This isn’t about media dollars—it’s about informational value.

Treat measurement as a work in progress

It’s tempting to try to reconcile everything to last-click or a clean attribution model. But that’s not how people shop anymore. The better approach is to triangulate what’s working, even if it’s messy. Combine behavioral signals, test organic presence, look at influence across platforms—even when it doesn’t show up in the primary dashboard.

Get comfortable with less control

These new surfaces aren’t built for traditional control levers. You won’t always get keywords or placements to optimize against—just signals, prompts, and outcomes.

A lot of these platforms aren’t built for marketers. You don’t get to pick the placement or pay to move up the list. That’s frustrating. But it’s also freeing. It pushes us to invest in better content, clearer value props, and more relevant product data—the kind of stuff that scales across touchpoints.

Marketers are no longer simply bidding on keywords. We’re competing across a broader ecosystem—one that includes AI-generated answers, influencer-led discovery, retailer-native search, and voice-driven shopping moments. The job isn’t just about conversion. It’s about being credible, useful, and findable in more places than ever.

What we’re building toward

At Skai, we’re thinking about what this means holistically—not just as a search problem, but as a broader commerce challenge. The work we’re doing to unify data across retail media, paid search, and social platforms is already helping brands get closer to this new reality. It’s not about chasing every trend. It’s about staying rooted in how real discovery is actually happening—and making sure brands show up early, often, and well. That also means tailoring discovery strategies by audience—what works for Gen Z on TikTok may not apply to Gen X shoppers on Google or YouTube.

Search isn’t a single input anymore. It’s a constellation of prompts, platforms, and content surfaces. Our role is to meet consumers in those moments—before the ad, before the keyword, before the purchase. That’s why we’re developing an AI Discovery Toolkit to help brands benchmark where they stand and strengthen product visibility across every emerging surface.

As Google, Meta, and others embrace AI-first discovery, we’re actively exploring how tools like Celeste and our broader platform can help brands stay ahead—before AI search fully rewrites the playbook.

And if we can do that right, we don’t just follow the shift. We lead it.