eCommerce

Get Ready for the First Cyborg Monday

In my recent Fast Company article, I introduced the idea that this holiday season won’t just be another Cyber Monday. It will be the first Cyborg Monday—when not all shoppers are human. AI agents are now researching, comparing, and even transacting on our behalf.

If the early years of SEO taught us anything, it’s that the leaders who optimized first built an enduring advantage. This is that moment again. The funnel has shifted from blue links to AI answers, and from journeys navigated only by people to journeys co-navigated by people + their agents.

Ghost shoppers, not just ghost traffic

I’ve described these agents before as ghosts. They skim and summarize product pages invisibly, long before a human ever decides to click through. And they don’t all behave the same.

  • Crawler-based bots like GPTBot, PerplexityBot, Googlebot, and Bingbot fetch and index at scale. They usually respect robots.txt, though execution of JavaScript varies (Google and Bing can render it, but only asynchronously and with limits).
  • Session-based “delegated” agents (surfacing as ChatGPT-User, Claude-User, or Perplexity-User in logs) fetch content on behalf of a specific human query. They don’t follow robots.txt because they’re not indexing or training their model, they’r representing the human. They are fetching directly to answer a person in real time.

This distinction matters. The second class is where Cyborg Monday comes alive. It’s when a human delegates part of the shopping journey to their AI, which then appears in your server logs as a ghost in between you and your customer.

Make your catalog readable to AI (and to people)

If bots can’t parse your PDPs, you’re invisible. That invisibility is why so many sites feel “ghosted” by AI shopping agents. The fix starts with structured, machine-friendly product data.

We don’t actually know yet if today’s AI agents are parsing Product schema as a primary signal—but that’s not the point. The point is that schema.org gives us a structured, standards-based way to describe products, and it’s the perfect baseline to build from once we learn which methods the agents prefer. Think of it as laying the pipes before you know which utility company will connect. And yes, if you’ve been around long enough to remember the early SEO hacks, you’ll appreciate the irony: unlike hiding keywords in white text on a page, this is a transparent, best-practice way to make sure both humans and machines understand what you’re selling.

Use schema.org properly. Your official product name must stay in Product.name—no brand wants AI agents confusing a trademark with a generic label. But schema.org gives you safe ways to layer in clarity for the LLM:

  • alternateName: a plain description, e.g. “Men’s lightweight running shoe.”
  • disambiguatingDescription: a one-sentence factual clarifier, e.g. “Neutral road runner designed for daily training.”
  • model: a field for the marketing moniker, e.g. “Cloud Walker 2.0.”

Together, these fields let you preserve your official naming and give LLMs the context they need. Reinforce with category, brand.legalName, sku, and gtin so the item resolves cleanly across systems.

Enrich your catalog. Adding attributes isn’t just SEO hygiene anymore—it’s what lets agents differentiate your product from dozens of lookalikes. Platforms like Stylitics can enrich apparel attributes, while newer tools like Refibuy audit your PDPs from the perspective of AI shopping engines and flag “GEO blockers.”

How to check if you’re ready:
Open one PDP, right-click “View Page Source,” and search for @type": "Product". If you don’t see JSON-LD markup, you’re missing a quick win. To validate further, drop the URL into Google’s Rich Results Test. One PDP is enough—if your templates emit schema, it should be present on all.

Tune for GEO, not just SEO

Generative Engine Optimization (GEO) isn’t a spec yet. It’s a shift in mindset: optimize not just for how pages rank, but for how answers are generated. Google’s AI Overviews, Bing’s AI Mode, and shopping-oriented LLMs increasingly privilege content that is clear, structured, and unambiguous.

That means:

  • Lead PDPs with a plain first paragraph. Agents heavily weight concise, factual context.
  • Use canonical URLs and consistent titles so each product is a single, unambiguous entity.
  • Where possible, add short comparisons directly on PDPs (“Compared to X, this model is lighter/longer-lasting…”). Summarizers love relative framing.

And yes—speed still matters. People bounce from slow pages. Performance influences search ranking. But for bots, it’s clarity, not cleverness, that determines whether you appear in the answer at all.

Cleverness without losing clarity

Some brands fear that being “plain” will dilute their personality. They shouldn’t. Keep the playful copy for people—the witty headlines, the lifestyle-rich descriptions. But pair it with structured clarity in the markup.

Think of it as a bilingual strategy: human-facing copy that delights, and machine-readable fields that define. The former builds brand, the latter builds discoverability. You don’t have to sacrifice cleverness; you just have to mirror it with clarity where it counts.

Prepare your accounts for agents

Cyborg Monday isn’t just about discovery. Increasingly, shoppers will want their AI to log in, check orders, retrieve sizing history, or apply loyalty points. That requires a safe way to delegate access.

The Agentic Commerce Protocol (ACP) is one emerging standard. It defines how agents can coordinate checkout and share tokenized credentials while still keeping the merchant of record in control. Stripe and OpenAI have already partnered around delegated payments, and early implementations are live. I shared some ideas on this previously where we might need an interim solution for the bots to consume their own HTML and minimally need to rethink account management in an era of shopping agents.

Your action: audit where account data could be safely exposed to an authorized agent—order history reads, loyalty balance, subscription info. Even if you’re not agent-ready today, understanding ACP now keeps you from being blindsided when customer expectations shift.

Measure your agent visibility

The hardest part of this shift is that AI shopping bots are invisible to your current dashboards. As I wrote recently in Yottaa’s blog, traditional analytics platforms don’t register these agents.

Here’s what you can do:

  • Inspect server logs for GPTBot, PerplexityBot, or session-based ChatGPT-User fetches.
  • View your PDPs with JavaScript disabled—if critical content disappears, many agents will never see it.
  • Pilot an external auditing tool like Refibuy to evaluate “AI-readiness” of your catalog.

It’s the same playbook we ran in the early SEO era: watch the referrers, measure visibility, and fix the gaps before your competitors do.

A quick personal note

The reality of this shift hit me on a flight with my 10-year-old daughter. She loves music, but I don’t want her to have a phone yet. I found myself wondering: What’s the modern Walkman that plays Spotify? My instinct wasn’t to Google it. I asked Perplexity. In seconds, it summarized options, reviews, and where to buy.

In that moment, I wasn’t the shopper. My agent was. Multiply that across households this holiday season, and you see why “Cyborg Monday” isn’t just a clever phrase—it’s the new funnel.

Where to start this week

Make sure your PDP templates emit Product schema with both official and alternate descriptors. Validate with Google’s Rich Results Test. Enrich attributes where gaps are obvious. Decide your bot policy and put ACP discovery on your roadmap.

The early movers captured SEO. The early movers will capture GEO. On Cyborg Monday, milliseconds and markup will decide whether you win the shopper—or the shopper’s agent.