The Second Audience

I was watching a panel the other week. Peter Diamandis, Alexander Wissner-Gross, a handful of researchers swinging between "this changes everything" and "we're all doomed." The topic was Google's new Universal Commerce Protocol, and the consensus seemed to be forming around extinction. The web as we know it, finished.

Then Wissner-Gross said something that stopped the pendulum.

"This is not the end times," he said. "It is a way to start to standardize e-commerce from within Gemini and other chat agents. That's all it is."¹

Those four words stuck with me. Not because they're dismissive, but because they're the most honest thing I've heard anyone say about what's happening to the web this year.

The Protocol

This week, Google made items from Etsy and Wayfair purchasable without anyone visiting either site.² The mechanism is UCP, an open-source protocol that lets AI agents browse a store's inventory, check prices, and complete a purchase inside a conversation. Shopify, Target, and Walmart are next.

If you've been tracking the "death of the website" headlines, this feels like confirmation. Why visit a store when the agent already handled it?

But Wissner-Gross is right. This isn't an ending. It's a standardization.

The Second Audience

Automated traffic now accounts for 51% of all web activity.³ Your website already has more non-human visitors than human ones. Most businesses haven't noticed because everything about their site still faces one direction: toward a person with a screen and a scroll wheel.

Joe Toscano, writing in Forbes, has a name for the other direction: "Agent Experience," or AX.⁴ Think of it as UX for algorithms. Your site was built for people who browse, feel your brand, respond to your photography and layout. Now a second audience shows up that ignores all of that. It reads what's structured. Specifications, inventory, pricing. Clean answers it can hand back to whoever sent it.

Same storefront. Two very different visitors.

The two audiences don't compete. Humans want warmth and story. Agents want parseable data. They just read differently.

Translation, Not Extinction

McKinsey estimates that AI agents could mediate $3 to $5 trillion in global commerce by 2030.⁵ That's not a number you wave away. But people still browse, still window-shop, still buy things because the packaging caught their eye or the display made them pause. Agents won't replace that. They'll handle the transactions people would rather skip: reorders, price comparisons, the "find me the same thing but cheaper" errands.

Learning the second language.

The businesses figuring this out aren't rebuilding from scratch. They're adding a second language to the site they already have. The human side stays warm. The agent side stays structured. Same address, two conversations happening at once.

Wissner-Gross said it well. Not the end times. A standardization. Your website isn't dying. It just learned it has a second audience, and that audience reads with different eyes.

---

References

Footnotes

1. Diamandis, P. & Wissner-Gross, A. (2026). "Abundance Summit Panel Discussion." YouTube ↩

2. Srinivasan, V. (2026). "What to expect in digital advertising and commerce in 2026." Google Blog ↩

3. WP Engine. (2026). "The Shape of the Web in 2026." WP Engine Blog ↩

4. Toscano, J. (2026). "In The B2AI Era Agent Experience Is Rewriting The Marketing Playbook." Forbes ↩

5. Mahajan, D., Mayer, H., Schumacher, K., & Roberts, R. (2026). "The automation curve in agentic commerce." McKinsey ↩