Programmatic I/O Day 2 Recap: Signal Loss, Glass Boxes, and the Challenger Class of Social

By Chris Marine | Founder & CEO, Campfire Consulting

Las Vegas, NV — May 2025

The opening keynote said it best: “We’re no longer disrupting—we’re reconstructing.”

David Shing (“Shingy”) kicked off Day 2 with a hard truth that rippled through every session after it: the programmatic ecosystem isn’t just shifting it’s unraveling and being rewired in real time. The smart brands, platforms, and agencies aren’t trying to fix what’s broken. They’re building something new with the tools we all have the same access to.

Here are five signals that stood out from Day 2 of Programmatic I/O:

1. Alt IDs Are Here But Not the Silver Bullet You’re Hoping For

In a session that could’ve been titled “You’re Already Behind,” OMD laid out the stark reality: over half your audience is already invisible. Between Safari, Firefox, and iOS, relying on third-party signals isn’t just outdated — it’s actively costing you money.

Alt IDs — from UID 2.0 to publisher, and brand-owned identifiers offer a way forward, but only if you’re playing the long game. The key is diversification, not overcommitment. "Don’t put all your eggs in one ID," the speaker warned. “There’s no finish line. Just a constant need to adapt.”

And don’t wait for Google. Their cookie deprecation plan has become the industry’s most expensive ghost story. Signal loss is already here and the smarter question is: what’s your test-and-learn roadmap now?

2. Amazon Ads Embraces the Glass Box Era

Amazon’s shift from black box to “glass box” advertising feels like a long-overdue step toward transparency. Performance Plus and Brand Plus were the stars here giving brands clearer reporting and access to AMC audiences (Amazon Marketing Cloud), but with more control and optimization baked in.

As the speaker put it: “If AI makes everything accessible, then first-mover advantage becomes your only advantage.” With Amazon now opening up its tools for deeper attribution and rule-based AI targeting, expect more challenger brands to beat incumbents on speed, not spend.

3. Foresight Mapping and the Coming Identity Whiplash

This was one of my favorite moments. The poor presenters were budded right up against lunch, so when the room emptied out, I stayed back with 8–10 leaders for a deep dive into a practice workshop tactic called “if then, what?” The idea, developed by Projectory, is about applying a “hive mindset” to push beyond individual bias and explore divergent futures.

I loved this chance to go deep rather than wide and to circle around a table where we weren’t just nodding along to trends, but actually projecting implications.

The most provocative scenario? A world where alt IDs never reach cookie-level scale. That would trigger a domino effect: explosion of niche IDs → unsustainable measurement → a swing back toward consolidation. And buried six layers deep? A prediction that AI will need to analyze the chaos, but it will still take human oversight to translate it into strategy.

It was part sci-fi, part strategy retreat but the message was clear: marketers need to practice working backward from potential futures. “The hive mind is the insight engine,” one facilitator said. In a post-cookie world, that might be the only one that still works.

4. From Monetization to Measurement: The NYT’s Tetris Metaphor

Leave it to The New York Times to explain programmatic monetization with a Tetris analogy — and somehow make it work.

Their approach to yield management is designed with intention: direct sales sit at the top of the waterfall, curated PMPs fill the middle, and open auction fills in the gaps. Their message to buyers? Not all impressions are created equal — and if you’re cherry-picking Times inventory, don’t expect a discount.

But the real gem came from a moment of quiet. I asked: “Why doesn't The Times share how our media dollars directly fund journalism — not just clicks?” The silence was brief but powerful. The industry is hungry for that connection. Publishers with purpose should lean in and lead with it.

5. Personalization Is Out. Relevance Is In.

One of the most memorable lines from Day 2 came from Connie Chan Wang, Head of Marketing at Headspace, during the AI Creative session:

"What I would like to hear more about is less about personalization at scale and instead relevance at scale."

That line stuck with me, and frankly, it should guide every creative and media brief moving forward. What she’s referencing is something we at Campfire know and work to practice with our clients investing across channels all too well: that even reaching the same person throughout the day means speaking to different needs and mindsets. The relevance of your message should evolve with them. It really goes beyond the old mantra of “right person, right place, right time.”

Connie and the team at Headspace shared how they used generative AI to create hyper-relevant creative across hundreds of ad variations targeting caregivers, empty-nesters, and stressed-out professionals during the holiday season. But this wasn’t just AI for AI’s sake. It was AI grounded in deep human need — and tested for real-world performance.

The result? A 62% improvement in conversion rates and a 13% decrease in cost per signup, all while honoring the emotional reality of their audience.

It’s a reminder that AI isn’t here to replace creativity. It’s here to amplify relevance. And when used thoughtfully as part of a holistic media strategy, it doesn’t just scale output it strengthens empathy.

6. Challenger Platforms Find Their Voice (And Their AI Tools)

Reddit, LinkedIn, and Spotify took the stage together in what felt like a declaration: we’re not just alternatives we’re building better experiences.

• Reddit leaned into “community intelligence” turning its massive archive of real human conversations into AI-augmented audience insights that marketers can tap (without black boxes).

• LinkedIn focused on B2B precision and professional trust, with AI tools powering dynamic creative, predictive audiences, and clean measurement frameworks.

• Spotify showcased its new generative AI voice ad builder creating high-quality audio in under 20 minutes and solving one of marketers' most annoying friction points.

Each of them acknowledged that how marketers buy matters but so does why. Curation, transparency, and AI aren’t products. They’re philosophies. And the platforms that treat them as such are the ones earning long-term loyalty.

Final Thought

“If we want to change the future, we have to change the future we see.” That line from the keynote wasn’t about blind optimism. It was a call to action for everyone in media to wake up to the moment we’re in.

The scaffolding of yesterday’s ad ecosystem: cookies, default platforms, one-size-fits-all creative, and predictable audience targeting isn’t just evolving. It’s crumbling under the weight of new privacy laws, platform shifts, and generative AI that’s reshaping what’s possible and who controls it.

But this isn’t the end of human influence. It’s an inflection point. AI might take over the infrastructure, but it’s still up to us to design the blueprint. The brands and marketers who rise now will be the ones who use these tools not just to automate but to create better, faster, and more relevant stories. Stories that still sound, feel, and move like they came from someone real.

That’s the paradox of this next chapter in media: the more intelligent our systems become, the more essential it is that we stay rooted in what makes a message land and what makes people care. Emotion. Purpose.

Authentic connection isn't something AI can generate. But it might just be what helps us generate the future we actually want.