The phrase came from Pinterest. But it captured something much larger happening across the entire marketing landscape right now.
“Less URL. More IRL.”
At the exact moment that Google is launching AI agents to manage entire search ad funnels autonomously, that Meta has recovered 90% of the signal Apple’s ATT removed, and that TikTok Shop has become the fourth largest beauty retailer in the UK, some of the most senior marketers in the world are making a counter-bet. Sephora, Heineken, and Hinge are all explicitly prioritizing community and emotional resonance over automation. Cannes rewarded human craft over AI generation. And the data is beginning to suggest they are right.
The most interesting tension in digital marketing right now is not between platforms or formats. It is between the relentless efficiency of AI automation and the irreplaceable value of genuine human connection. This week’s intelligence captures both sides of that tension with unusual clarity.
Here are the seven developments that define where things stand.
1. Google Just Launched an AI That Argues With Your Customers on Your Behalf. Inside Your Ad.
Google’s Business Agent for Leads is a genuinely significant product launch that deserves more attention than it has received. The tool allows consumers to have full-funnel conversations with a Gemini-powered AI agent directly inside a search ad, without ever visiting a brand’s landing page. The AI handles discovery, objection-handling, qualification, and lead capture, all within the ad unit itself.
Alongside it, Google launched Ask Advisor, an AI assistant that reasons across Google Ads and Analytics to provide strategic marketing advice, effectively positioning AI as a co-strategist rather than just an execution tool.
The implication for how marketing work gets organized is direct. If an AI agent is now handling the customer conversation inside the ad and an AI advisor is managing campaign strategy, the human marketer’s function has shifted from managing both of those things to supervising AI systems that do them. That is not a marginal efficiency improvement. It is a structural redefinition of what the job involves.
The brands that figure out how to supervise these systems well, providing the right inputs, catching the outputs that miss, and retaining strategic judgment over brand voice and commercial priorities, will extract significantly more value from the technology than the ones treating it as a hands-off automation and stepping back entirely.
2. AI-Sourced Sessions Last 7 Minutes 35 Seconds. Standard Search Gets 4 Minutes 41 Seconds.
This is the most commercially significant data point in this period’s intelligence, and it builds directly on the 54% higher conversion rate finding from the previous month.
Roth Capital Partners data shows AI-sourced web sessions averaging 7 minutes and 35 seconds, compared to 4 minutes and 41 seconds for standard search traffic. That is a 62% longer engagement window from the same product page, driven by the same underlying mechanism: AI has done the qualification work before the consumer arrives. They are not browsing. They are evaluating a specific recommendation with genuine intent.
The strategic implication compounds. When AI-sourced visitors stay longer, they provide more behavioral signal, interact with more content, and are more likely to convert on a second or third visit even if they do not convert immediately. The quality of the traffic creates a self-reinforcing cycle: better engagement signals teach the AI to surface the brand more confidently in future queries, which generates more high-quality traffic.
This is the clearest argument yet for treating Generative Engine Optimization as a primary acquisition investment rather than a supplementary brand awareness play. The channel is delivering higher-quality buyers and longer engagement than the channel that has been the foundation of digital acquisition for two decades.
3. TikTok Shop Is Now the Fourth Largest Beauty Retailer in the UK. Social Commerce Is Mainstream.
TikTok Shop achieved 60% year-over-year growth in the UK beauty category, reaching a market position that places it ahead of established specialty retailers. This is not a social media trend. It is a retail market share data point.
The distinction matters for how beauty brands should be allocating resources. A brand that treats TikTok Shop as a social promotion platform with some commerce capability attached is investing in a fundamentally different strategy than one that treats it as a primary retail channel with its own inventory management, customer service requirements, and performance metrics.
The brands that made that strategic distinction early are the ones capturing the 60% growth. The brands still thinking about TikTok primarily as an awareness channel are watching their competitors build retail infrastructure in a market that has already moved.
The Thailand data from this same period reinforces the commercial maturity signal. Thai consumers are in a “value-driven spending” phase, conducting significantly more research before buying. This has driven a 1.9x increase in TikTok Messaging Ads, because 80% of buyers now want direct brand contact to confirm product details before transacting. The social commerce journey is not just a browse-and-buy experience. It is a research and conversation experience that requires brand responsiveness at the messaging layer.
4. Meta Recovered 90% of Apple’s ATT Signal. The Cookie Workaround Conversation Is Less Urgent Than It Seemed.
Meta has recovered approximately 90% of the behavioral signal lost to Apple’s App Tracking Transparency framework through Advantage+ automation and content-based inference. The platform is predicting user intent from content signals and behavioral patterns at a scale that has largely compensated for the loss of device-level tracking.
This is a meaningful development for the media planning conversations that have dominated the past three years. The urgency around third-party cookie replacements, clean room infrastructure, and alternative identity solutions was predicated on the assumption that platform signal loss would be permanent and significant. Meta’s recovery rate suggests that, at least for Meta’s owned ecosystem, the algorithmic infrastructure has largely compensated.
There is a nuance worth noting. The same period’s data shows a slight efficiency “wobble” in June that Roth Capital Partners flags as unresolved. The 90% signal recovery is not a clean technical fix. It is a probabilistic approximation built on inference rather than direct measurement. For high-precision targeting use cases, the difference between real behavioral data and inferred behavioral data matters. For broad audience reach and brand objectives, it probably does not.
The practical implication is that the investment case for expensive alternative identity infrastructure becomes harder to justify for brands that operate primarily within Meta’s ecosystem. The case remains strong for brands with significant cross-platform measurement needs or first-party data strategies that extend well beyond social.
5. Automated Creator Ads Delivered 81% Higher ROAS. The Creator Marketing Efficiency Era Has Arrived.
New integrations between Agentio and Meta have produced a specific performance outcome that should change how creator marketing budgets are structured: 81% higher ROAS compared to non-automated creator ads, with a 13% lower cost per action.
The mechanism is worth understanding. Automated creator marketing platforms match brand briefs to creator audiences algorithmically, optimize delivery timing and format, and iterate on performance signals continuously, all without the manual negotiation and campaign management that traditional influencer marketing requires. The result is creator content with the trust and authenticity premium that human creators provide, delivered with the targeting precision and optimization speed that automated systems enable.
This is the combination that has been theoretically promised for years. The data is now showing it works at meaningful scale. For brands that have been treating creator marketing as a high-touch, manually managed channel with unpredictable ROI, the Agentio data is the proof point that the performance infrastructure has matured enough to justify reallocation from traditional paid social to automated creator programs.
“Automated creator marketing is producing 81% higher ROAS with 13% lower cost per action than non-automated equivalents.”
The implication for agency relationships is also direct. The manual labor involved in creator campaign management, the outreach, negotiation, briefing, review, and reporting, is being automated. The value agencies add in this environment is shifting toward creator strategy, brand alignment judgment, and creative quality control, not toward the administrative execution layer.
6. “Less URL. More IRL.” Is a Business Strategy, Not a Brand Positioning Statement.
Pinterest’s “Less URL. More IRL.” framing has crystallized something that Sephora, Heineken, and Hinge are all building toward independently: the most durable competitive moats in a world of AI-generated content saturation are the ones that AI cannot replicate.
Community is one. Emotional resonance is another. Physical experience is a third. The Quadrant x Shopify pop-up during British Grand Prix week is the most concrete execution in this period’s data: a social campaign earning 5.8 million impressions that explicitly converted digital audiences into a premium physical retail experience, moving 10,000 highly engaged fans into a Shopify-powered pop-up store. The digital and physical were integrated as a single commerce journey rather than as separate channels.
The “human moat” framing from Berenberg’s research captures the strategic logic. As AI makes digital content creation cheaper and more uniform, the things that require genuine human presence, taste, craft, community, and physical experience, become differentially scarce. Scarce things in a market of abundant alternatives command premium pricing and premium loyalty.
This is not nostalgia for a pre-digital era. It is a precise read of where competitive advantage is migrating as the digital layer becomes commoditized. The brands investing in IRL experiences, community building, and emotional resonance are not stepping back from digital marketing. They are building the one layer that digital marketing alone cannot manufacture.
7. Addressable Advertising Just Reached Tens of Millions of Homes That Were Previously Invisible.
A collaboration between INVIDI Technologies and Google Ad Manager has unlocked addressable advertising for unconnected set-top boxes globally, reaching households that were previously inaccessible to digital precision targeting. In markets like India, where linear television still reaches enormous audiences through set-top boxes without internet connectivity, this represents access to a scale of reach that programmatic buying has never been able to address.
The implication for global brands with significant emerging market presence is direct. The gap between the precision of digital advertising and the reach of linear television has been one of the persistent inefficiencies in international media planning. Addressable advertising on unconnected screens begins to close that gap, bringing targeting precision to audiences that have historically required broad linear buys with no audience-level optimization.
For brands currently managing separate linear and digital budgets in markets like India, the technology creates a path toward unified audience planning across both environments. The audience that was previously only reachable through linear TV can now receive a message that is informed by the same data signals as the digital campaign running simultaneously.
The Pattern Across All of It
The July 14 data captures the most clearly defined duality in this entire series. On one side: AI agents managing search funnels, automated creator marketing delivering 81% ROAS lifts, Meta recovering its tracking signal algorithmically, and Google positioning AI as a co-strategist. On the other: Pinterest’s IRL pivot, Sephora and Heineken prioritizing emotional moats, Cannes rewarding human craft, and a Lando Norris pop-up converting 10,000 TikTok followers into in-person customers.
These are not contradictory strategies. They are complementary ones. The brands winning in the current environment are using AI to make their digital operations faster, more precise, and more efficient, while simultaneously investing in the human and physical experiences that AI cannot commoditize.
The risk is in choosing one side of that duality over the other. Pure AI efficiency without human resonance produces the “sea of sameness” that Cannes juries are already rejecting. Pure IRL investment without digital precision means reaching people at higher cost with lower measurability. The brands building both simultaneously are the ones with the most durable positioning.
– Manpreet Jassal
Leave a Reply