Marketecture Live III (MLIII), held March 10–11 in New York, delivered exactly what it has become known for: high-signal, content-driven programming paired with rapid-fire conversations designed to keep media, advertising, and marketing leaders current on where the industry is heading.

The event drew a heavily New York–centric audience, and that alone highlighted one of the most important underlying dynamics in the industry today. Compared to San Francisco—where nearly every conversation is now AI-first—New York still feels about a cycle behind. Not unaware, but not yet fully reoriented. AI is clearly top of mind, but it hasn’t yet reshaped the core narrative of how this ecosystem operates.

That said, the gap is closing quickly. The digital advertising ecosystem is uniquely well-positioned in the AI transition. Machine learning, optimization systems, and probabilistic decisioning have long been embedded in programmatic advertising, CTV, and performance marketing. In many ways, this industry is not adopting AI—it is evolving into its next phase.

The coasts will converge sometime in 2026. But for now, the most forward-leaning, AI-native thinking still starts—and often ends—in San Francisco.

A Conference Framed by “Consumers in Control”

The central theme of MLIII—“Consumers in Control”—was evident across nearly every session. From identity and retail media to creative and CTV, the industry is grappling with a simple but profound shift: consumers are no longer being “targeted” in the traditional sense. They are navigating, filtering, and increasingly delegating decisions through AI interfaces.

The setting reinforced the tone. The Glasshouse, with 70-degree weather and sweeping Hudson views, made the sixth-floor networking as valuable as the sessions themselves. Add a DJ, a glass of prosecco, and a highly curated attendee base, and Marketecture has firmly established itself as one of the few must-attend gatherings at the intersection of media, marketing, and advertising. Credit to Ari Paparo and the Marketecture team for building something that consistently delivers both signal and access.

Three Key Themes and Takeaways

1. The Traffic Model Is Breaking—Distribution Is Fragmenting Fast
Publishers are already living in a post-Google world. Referral traffic declines—some cited as much as 50%—are forcing a rapid pivot toward diversified distribution and direct audience strategies.

At the same time, new monetization models are emerging:

• Content licensing deals with LLM platforms (“all-you-can-eat” vs. structured data access)
• Audience-first strategies replacing channel-first thinking
• Early experiments with agent-driven consumption and micropayments

The takeaway: the value chain is shifting from traffic → monetization to content → data → licensing → AI consumption.

2. Retail Media and Transaction Graphs Are Becoming the New Power Centers
Retail media continues to evolve from an ad channel into core infrastructure. Players like PayPal are leveraging transaction graphs (30M+ merchants, massive payment data) to build identity systems with far greater fidelity than traditional advertising graphs.

Two key implications:

• Commerce is becoming embedded everywhere—every interaction is a potential transaction
• Merchants themselves are becoming media owners, monetizing both data and demand

The most important shift: traffic is no longer flowing to merchants—commerce is flowing to the consumer wherever they are.

3. CTV and Premium Inventory Are Reasserting Control
Despite years of programmatic expansion, the CTV market remains heavily bifurcated:

• ~$90B total TV market
• Only ~$15B fully biddable programmatic inventory
• The majority of premium supply still transacted via direct or guaranteed deals

More importantly, a large portion of CTV buying remains structurally inefficient. Upwards of 60% of inventory is effectively “blind,” with limited content-level signal in the bidstream, making much of the inventory look indistinguishable to buyers.
Publishers and platforms are responding by pulling back toward:

• Direct relationships
• Premium environments
• Greater control over data and pricing

The industry narrative of “everything becomes programmatic” is being replaced by a more nuanced reality: premium supply wins, signal matters, and control is being reasserted.

Start-Up Competition: AI Everywhere, but Not Yet Differentiated

My favorite part of the event was the startup showcase, where the next generation of companies pitched their vision for the ecosystem.

AI dominated nearly every presentation—but in a way that revealed something important. Most companies are applying AI within existing frameworks rather than redefining those frameworks entirely.

Standout companies included:

• Gamera — focused on classifying and valuing every ad placement on the open web, shifting away from purely user-ID-based targeting
• OpenGlass — building infrastructure to enable more advanced and non-standard CTV buying experiences
• CrowdAxis — bringing measurement and standardization to the $100B+ experiential marketing category
• Datalinx.ai — positioning as an “AI data refinery,” improving speed and usability of marketing data from weeks to days

The Missed Conversation: Open Web vs. Walled Gardens in an AI World

One thread that wasn’t fully explored—and arguably should have been central—is the coming disruption of the open web versus walled gardens in an AI-driven ecosystem.

There were hints:

• “Zero-click” behavior
• Generative Engine Optimization (GEO)
• Declining direct traffic
• Chat interfaces becoming discovery layers

But the conversation stopped short of the real implication: What happens when “search” is no longer a destination, but a conversation?

Today, the open web captures more consumer attention than its share of ad spend—a gap that AI-driven discovery could either correct or permanently distort.

And underneath this shift is a more fundamental structural change: the next evolution isn’t just AI optimizing campaigns—it’s agents negotiating media on behalf of buyers and sellers, introducing a new protocol layer for how advertising is transacted.

However, that raises a critical question the industry has not yet answered: if this future is powered by agent-driven protocols, but those protocols are fragmented and numerous, do they actually create a new framework—or simply make an already complex ecosystem even more fragmented?

If chat interfaces and agents become the primary layer of discovery and execution:

• The open web risks losing its entry point
• Walled gardens (or “hedged gardens”) may gain even more control through integrated data + distribution
• Attribution, measurement, and monetization models could be fundamentally rewritten

This is not a marginal shift—it is a potential inversion of the entire digital advertising, marketing, and media stack.

Final Take

Marketecture Live III confirmed something important: the industry is leaning into AI—but still thinking about it incrementally.

The real shift is not AI improving advertising. It is AI redefining how consumers discover, decide, and transact.

And when that fully plays out, many of today’s assumptions—traffic, targeting, measurement, even what an “ad” is—will need to be rebuilt from the ground up.

The industry isn’t behind. But it may still be underestimating just how different the future will look.

Best regards,
Ryan Klinefelter
Executive Director
Woodside Capital Partners