Every year at retail's biggest show, the three halls and dimly-lit basement of the Javits Center flood with vendors hawking the next big thing. But 2025's NRF has a different energy—we're not just talking about new technology, we're talking about new shoppers. Not the Gen Alpha demographic analysts have been breathlessly tracking, but an entirely new species of commerce participant: the agent.

Platform shifts don't just change how we shop—they change who's shopping. The desktop-to-mobile transition wasn't merely technical; it fundamentally altered how humans browsed, compared, and purchased.

We built responsive design to accommodate different screen sizes.

Now we need responsive intelligence to accommodate different species of shoppers.

This marks the third great platform shift in digital commerce:

• First came the information revolution (desktop web)
• Then the attention revolution (mobile/social)
• Now, we're entering the intelligence revolution (human-agent commerce)

Each shift required new design patterns, new language, and new ways of thinking about commerce itself. Just as we once had to rebuild our interfaces to be "mobile-responsive," we must now make them "agent-responsive." But this shift goes deeper than any before it: it represents the first time humans will truly share commerce channels with non-human participants.

The stakes couldn't be higher. We risk creating what I call "Agentic Ghettos"—segregated commerce experiences that echo the mistakes of "m-dot" mobile sites. Instead, we need a unified theory of commerce that serves both human and artificial intelligence gracefully. The future of commerce isn't human or agent—it's both species shopping side by side.

Just as humans and Neanderthals once coexisted, we now coexist with another form of intelligence using our tools, our territories, our technologies.

Just as the responsive web required us to design for multiple screen sizes, agentic commerce will require us to design for multiple species of shoppers.

From Linear to Omnimodal: a New Commerce Pattern

The attention economy has been built on a fundamental scarcity: humans have limited cognitive bandwidth and finite attention spans. Every brand, every platform, every touchpoint competes for a slice of this finite resource. But what happens when attention becomes infinite?

On New Year's Eve 2024, I booked a dinner reservation through Google Assistant with a single tap. The agent called the restaurant, negotiated the time, and texted me confirmation. When plans changed, it handled the cancellation just as smoothly. This seemingly simple interaction reveals a profound shift: my agent's attention was unlimited, operating independently of my own finite bandwidth.

This points to a future where infinite brands can compete for infinite attention through infinite agents. Our current checkout funnels are necessarily linear because human decision-making requires focus. But agents don't work this way—they can process multiple decision trees simultaneously, evaluate countless options in parallel, and operate across multiple contexts at once.

Defining Omnimodal Commerce

We define omnimodal commerce as a species-fluid commerce experience. To be truly "omnimodal," commerce platforms must embrace three core principles of interoperability:

1. Contextual Fluidity: Systems must seamlessly adapt between human and agent interactions, recognizing and responding to each appropriately. Just as the Google Assistant knows when to make a voice call versus send a text, commerce platforms must flex between attention-guided and parallel-processing modes.
2. Intent Translation: Platforms must be able to translate between human intent (often emotional, contextual, and imprecise) and agent intent (structured, specific, and programmatic). This bidirectional translation layer is crucial for hybrid interactions.
3. State Persistence: Unlike linear funnels that reset when attention breaks, omnimodal systems must maintain state across multiple parallel interactions, allowing both humans and agents to weave in and out of transactions fluidly.

Omnimodal is fundamentally different from multimodal capabilities that vendors like ChatGPT and Nvidia speak about. ‘Multimodal’ capabilities within AI are sensorial—sight, sound, and textual input. Omnimodal commerce, by contrast, is the ability to fluidly serve a commerce experience through a single interface to an agent or a human, with or without the need for a linear funnel.

Commerce has been evolving beyond its linearity since 2020. This explains why we're seeing commerce design evolve beyond traditional patterns. Consider the new YZY site, with its 2.5D dimensionality that feels like shopping in StarFox 64, or sites like Drake Related and Narcos employing isometric designs. These aren't just aesthetic choices—they're early attempts to break free from the linear constraints of human-only commerce. The dimensionality of commerce is already here, agents accelerate it.

The stark contrast between Twitter's bot ecosystem and Meta's recent agent controversy illustrates this evolution perfectly. Twitter's bots, commissioned and shaped by the crowd, evolved organically despite (or perhaps because of) their unofficial nature. Meanwhile, Meta's attempt to legitimize AI agents through corporate channels faced immediate backlash. As William Gibson famously noted, "the future is already here – it's just not evenly distributed." The crowd seems to be leading where corporations hope to follow.

From Resistance to Acceptance

"I think that Facebook and Instagram have learned that even automated engagement is engagement and makes people feel good," I noted in a recent Future Commerce episode exploring AI agents. "Some engagement makes people feel less lonely." My co-host Brian Lange agreed, suggesting that this basic human need for response and recognition might be driving the acceptance of non-human engagement. As we explored in that conversation, perhaps what appears as manipulation or automation on the surface actually serves a deeper human need in our increasingly isolated digital world.

This tension between human connection and automated interaction has played out most visibly in the sneaker industry, where the battle between humans and bots has been raging for years. In 2019, retailers like KITH and Berrics made headlines by deliberately tricking bots into purchasing thousands of dollars worth of unwanted inventory. It was commerce's equivalent of building defensive walls—an attempt to keep the "other species" out.

But commerce, like nature, finds a way. What's fascinating about the sneaker bot era of 2018-2022 wasn't just the resale market it enabled—it was the entire economic ecosystem that emerged around the bots themselves. Popular services like CyberSole, Kodai, and Wrath operated on subscription models, charging monthly or annual fees for their software. This created a parallel economy: while some traders profited from reselling sneakers, others built profitable businesses providing the tools and infrastructure for that trade.

This infrastructure went beyond just the bots. Cloud servers, proxy services, and cook groups all emerged as vital parts of the ecosystem. Even as retailers developed new bot detection methods, these services evolved, creating a perpetual cycle of innovation. The language of commerce expressed itself on both sides of the transaction—not just in the resale margins, but in the sophisticated software-as-a-service business models that emerged to support it.

Just as ancient human settlements eventually moved from conflict to trade with neighboring groups, the sneaker industry has gradually shifted from resistance to acceptance. Today, platforms design for both human collectors and automated purchasing agents, recognizing that both are vital parts of the ecosystem. The industry's journey from antagonism to accommodation offers a preview of how broader commerce will evolve.

The Language of Coexistence

Our digital interfaces currently presume malice in automation. Since the earliest days of the commercial internet, we've built walls to keep "bots" out—from CAPTCHA to Cloudflare, from rate limiting to IP blocking. This antagonistic relationship reflects an era when bandwidth was precious and DDoS attacks were common. But we're already coexisting with automated participants, albeit painfully.

The Power of Words

Language shapes reality, and our current lexicon reveals deep-seated biases:

Antagonistic Terms:

• Bot: Implies mindless automation, spam
• NPC: Suggests lack of agency or intelligence
• Crawler: Connotes unwanted intrusion

Collaborative Terms:

• Agent: Suggests authority and purposeful action
• Copilot: Implies partnership and assistance
• Assistant: Focuses on helpful augmentation
• Extension: Represents amplification of human capability

As discussed in a recent Future Commerce episode, "Some engagement makes people feel less lonely... Perhaps what appears as manipulation or automation on the surface serves a deeper human need in our increasingly isolated digital world."

‍Design as Language

The lingua franca of eCommerce design has remained remarkably static—header navigation, product grid, cart icon in the top right. This standardization served us well in the era of human-only commerce, but omnimodal design requires a fundamental rethinking. We need interfaces that are equally accessible to both human shoppers and their agents, without speed bumps and hurdles that prove humanity.

The danger lies in creating what we might call "Agentic Ghettos"—separate, invisible API-only paths for automated participants. This would mirror the mistakes of the mobile web era, where "m-dot" websites created segregated, often inferior experiences. Just as responsive web design unified the mobile and desktop experience, we need a new responsive paradigm that gracefully serves both human and agent commerce.

This transformation is already visible in how major platforms are evolving. Salesforce's ambitious "Agentforce" initiative, launched amid industry-wide contraction and competition from newer players, signals recognition that the future of commerce requires new interaction models. While traditional eCommerce interfaces may be losing ground to Shopify's human-centric approach, the next platform shift will demand interfaces that can seamlessly accommodate both species of commerce participants.

The Great Migration

We're witnessing what anthropologists might recognize as a territorial shift. Just as ancient humans and Neanderthals established shared hunting grounds while maintaining separate domains, commerce channels are becoming increasingly mixed-use territories. Some "regions"—like high-frequency trading or commodity purchasing—will become primarily agent domains. Others—like luxury experiences or bespoke services—may remain primarily human territories. Most interesting will be the vast middle ground where both species coexist, each leveraging their unique advantages.

This migration is already visible in how certain commerce tasks are being ceded to agents: price comparison, inventory monitoring, reorder management. Meanwhile, humans are carving out new niches around experience, curation, and emotional resonance. It's not a simple replacement narrative—it's a complex story of territorial overlap and specialization.

Designing for Our Shared Future

The implications for Commerce platforms are profound. Just as early humans and Neanderthals shared tools—with evidence suggesting they learned and borrowed from each other's technological innovations—we're entering an era where interfaces must serve both human and agent users effectively. This means moving beyond the linear, attention-guided patterns that have dominated human-centric design.

Future commerce platforms will need to support both sequential and parallel processing, both emotional and logical decision-making, both singular and multiple simultaneous transactions. They'll need to authenticate both human and agent users, manage hybrid interactions, and maintain trust across species boundaries.

A New Chapter in Commerce Evolution

Just as homo-sapiens and Neanderthals once shared tools, hunting grounds, and even DNA, we're entering an era where humans and agents will share interfaces, channels, and commerce patterns. This isn't just a new technology—it's a new chapter in how intelligence, both human and artificial, engages with commerce. The question isn't whether we'll coexist, but how we'll adapt to shape this shared future.

For brands and platforms, the challenge ahead isn't just technical—it's evolutionary. Success will come not from resisting this change, but from embracing the opportunity to design for both species of commerce participants. The next platform shift isn't just about new interfaces—it's about new intelligences, new behaviors, and new patterns of coexistence.

The future of commerce isn't human or agent—it's both.