AI Is Rewriting Discovery—And Google’s Ad Growth Shows the Old Funnel Isn’t Dead
The internet is moving toward AI-curated discovery, but Google’s latest results show paid search remains remarkably resilient. Marketers now need to manage both realities at once.
The marketing story is no longer “search versus AI”
The most important development for marketers today is not a single campaign or a flashy brand refresh. It is the collision of two trends that once looked contradictory: consumers are increasingly discovering information through AI-curated feeds, video, and conversational interfaces—while Google’s advertising engine is still growing strongly.
Axios describes the shift well: online discovery is moving away from chronological feeds and toward highly personalized, AI-driven video and conversations. Its conclusion is consequential for brands: niche, video-first content gains value in an environment where algorithms—not audiences—do more of the choosing. ([axios.com](https://www.axios.com/media-trends-membership/2026/07/28/internet-video-social-media-personalization?utm_source=openai))
At the same time, Alphabet’s second-quarter results offered a blunt rebuttal to the idea that AI has already dismantled search advertising. Alphabet reported $119.8 billion in quarterly revenue, up 24% year over year, with strong search-ad growth supporting management’s argument that AI is expanding Google’s search opportunity rather than replacing it. ([fortune.com](https://fortune.com/2026/07/23/alphabet-q2-earnings-spacex-gain-ai-investment/?utm_source=openai))
My read: the old funnel is not disappearing. It is being broken into more surfaces, more moments, and far less visible paths to purchase.
Discovery is becoming passive—and that changes creative
For years, digital marketers trained themselves to win an active consumer moment: someone searches, clicks, compares, and buys. That model remains commercially important. But it is no longer sufficient.
When a consumer receives recommendations from an AI assistant, a personalized video feed, or a shopping agent, the brand may never get the clean keyword query or site visit it once relied on. Discovery is becoming more passive and more mediated.
That has one immediate implication: brands need more assets designed to be selected, not merely clicked.
A great 30-second spot can still build memory. A great product page can still close a sale. But the connective tissue now matters more: concise product descriptions, clear category cues, distinctive visual codes, credible creator validation, accessible pricing and availability data, and content that works when excerpted, summarized, or shown beside competitors.
In other words, the brand system is becoming a retrieval system.
This is why generic performance creative is a weak answer. If every asset is optimized only for a short-term click, it becomes interchangeable the moment an AI layer compresses a category into three recommendations. Distinctiveness—what consumers and machines can consistently recognize as you—becomes more valuable, not less.
Google’s numbers are encouraging, but they are not permission to stand still
Alphabet’s results matter because they show that advertiser demand has not collapsed under AI disruption. Search remains a powerful commercial interface, and Google’s scale gives it room to embed new AI behavior into an established ads business. ([fortune.com](https://fortune.com/2026/07/23/alphabet-q2-earnings-spacex-gain-ai-investment/?utm_source=openai))
But marketers should resist the lazy conclusion that nothing has changed.
Alphabet also lifted its 2026 capital-expenditure outlook to $195 billion to $205 billion, and investors reacted negatively despite the company’s huge profit headline and accelerating cloud growth. ([fortune.com](https://fortune.com/2026/07/23/google-alphabet-biggest-quarterly-profit-112-billion-stocks/?utm_source=openai)) That reaction is useful for operators: the market is no longer rewarding AI spending on faith alone. It wants a visible connection between investment, product behavior, and durable revenue.
Marketing leaders should demand the same discipline internally. “We used AI” is not a strategy. “We produced 10 times more variants” is not an outcome. The useful question is whether AI is improving customer relevance, reducing friction, strengthening brand memory, or creating incremental demand.
Measurement is now the bottleneck
The hardest operational problem is attribution. Axios reports that commerce is splintering across channels, making it harder to identify what actually caused a sale and making last-click measurement increasingly inadequate. ([axios.com](https://www.axios.com/media-trends-membership/2026/07/18/ai-shopping-measurement-last-click?utm_source=openai))
That should force a reset in how teams allocate budget.
I would move away from treating a platform dashboard as a source of truth. Instead, operators should build a measurement stack around three layers:
- Incrementality: What sales or behaviors occurred because marketing ran, rather than alongside it? - Brand demand: Is branded search, direct traffic, repeat purchase, or consideration improving over time? - Commerce quality: Are new customers valuable, retained, and profitable—not merely cheap to acquire?
This is less tidy than last-click reporting. It is also closer to reality.
The takeaway
The next era of marketing will reward brands that can operate in two modes at once: capture existing intent through search and retail media, while creating distinctive content that earns selection inside AI-mediated discovery.
For operators, that means treating data hygiene, product information, creative consistency, and measurement as one system—not separate workstreams.
For investors, Google’s results are a reminder that AI has not killed advertising economics. It has raised the bar for proving where attention comes from, who controls it, and whether a brand can remain memorable when the interface between consumer and company is no longer a search box.