ChatGPT code red and why we’re switching to Gemini

Transcript

Is Your Site AI-Ready? What Black Friday, Gemini and ChatGPT Just Revealed About E-commerce

If you’re still wrapping your head around social media bans, age verification, AI tools and another Black Friday in the books, you’re not alone. Across e-commerce, there’s a common question emerging:

“How hard is it to shop from you – and how easy are you making it for both humans and AI?”

This article unpacks three big shifts happening right now:

• New age-verification rules and what they mean operationally
• Early signals from Black Friday demand
• How AI tools like Gemini, ChatGPT and Grok are quietly becoming the new “pre-search” layer before Google

And most importantly: what to do about it.

1. Social media bans, age checks and avoiding the “upload staff IDs” nightmare

Many businesses are still trying to work out how the social media age-verification and “safety” rules affect day-to-day operations – especially around ad accounts and staff access.

Questions like:

• Which staff IDs do we need to upload?
• Do we really have to upload IDs at all?
• How do we stay compliant without creating a privacy headache?

One practical workaround, if you pay for Google Workspace, is simply to use Google’s built-in age verification: you can designate that everyone under your workspace domain is over 18 via a single checkbox. That can avoid the operational and privacy friction of uploading individual staff IDs to social platforms.

At the same time, OpenAI has announced it will introduce age verification, officially framed around controlling access to adult content creation. Most agencies won’t be creating adult content, but some will need to navigate this for ad banners, creative production, and similar work.

The bigger picture: age gates and verification flows are only going in one direction – stricter and more pervasive. If you can centralise this verification inside trusted systems like your workspace identity, you reduce the admin load on your ad operations team.

2. Black Friday: demand looks softer, but results are spikier

Looking at Google Trends data for “Black Friday” searches over the last five years, Google is currently predicting a dip in search interest compared to previous years (with the usual caveat that not all data is in yet).

Interestingly, among the subset of brands that did run Black Friday this year, many saw their best results in years – significantly better than last year and the year before.

So, what’s going on?

• Some brands, especially in “grudge purchase” categories (things people have to buy, not want to buy), either sat Black Friday out or went light. For them, discounting isn’t essential.
• Shoppers who are in the market are often more intentional and more researched. They know what they want; they’re hunting for the right model and the right deal.

If your brand skipped Black Friday, it’s worth asking:

• Was it a strategic choice based on your category and margins?
• Or was it simply too hard from a technical or operational perspective to spin up offers and landing pages?

Either way, the bigger shift isn’t just “Black Friday up or down,” it is where shoppers are doing their pre-purchase research.

3. AI is becoming the new “pre-search” – and that changes everything

There’s a lot of noise about AI “bubbles” and rumours that OpenAI is planning to introduce advertising inside ChatGPT. At the same time, OpenAI has reportedly called an internal “Code Red” because of mounting competition.

Behind the headlines, what matters to e-commerce is how buyers are actually behaving.

Many users are now:

1. Doing their research in AI tools – Google Gemini, ChatGPT, Grok, etc.
2. Narrowing down to a specific brand or model.
3. Only then hopping into Google Search to do a brand search and click a paid ad or organic result.

For some brands, traffic from the ChatGPT channel in analytics is already generating more revenue than their Facebook paid campaigns. That’s a huge signal. And currently, no one is even paying to promote inside ChatGPT.

If and when advertising launches inside these AI tools, some e-commerce brands will be miles ahead because they already understand how users research there.

4. A real-world example: buying a 3D printer the “AI way”

Consider a recent real-world buying journey for a 3D printer.

The experience on typical retailer sites looked like this:

• Multiple printer models with near-identical product descriptions, despite important differences in features.
• Confusing pricing, including one site that showed GST-exclusive prices in ads, making deals look better than they really were.
• No FAQs explaining use cases, differences between models, or common pre-purchase questions.
• Poor clarity on delivery costs and shipping options.

In other words: friction everywhere.

Instead of trawling through these sites, the buyer opened Google Gemini and asked:

• What are the best Black Friday deals on these printers?
• What are the actual differences between specific models?

Gemini did the hard comparison work and summarised it all. That saved a huge amount of time compared to traditional browsing.

However, Gemini didn’t make it easy to click straight through to specific products, so the final step still happened where Google is strongest: brand search. With the model chosen, the buyer went to Google Search, looked up that exact printer, and clicked paid ads.

Key takeaway:

• Research is increasingly happening in AI tools.
• Transactions may still be attributed to Google Ads or direct traffic.
• If your product pages are confusing, missing FAQs or hiding shipping costs, AI has very little to work with – and humans get frustrated too.

5. Ignore “rank #1 in ChatGPT” hacks – focus on being AI-ready

You’ve probably seen the hype: “Here’s how to be number one in ChatGPT!”

In practice, that’s not how these tools are used.

When a shopper is researching a product, they’re having a conversation with an AI:

• Asking multiple questions from different angles
• Comparing brands, models, specs and reviews
• Clarifying use cases (“Is this good for beginners?” “Can it do X?”)

There’s no single “results page” to rank on. Large language models (LLMs) are only about three years into mainstream use – this is like 2003–2005 in Google terms. The trick-based SEO era for AI hasn’t even properly started, and chasing gimmicks now is a distraction.

What does work – both for humans and AI – is exactly what’s always worked:

• Clear, detailed product information
• Visible social proof (reviews, ratings, testimonials)
• Comprehensive, honest FAQs
• Transparent shipping, returns and warranty info
• A strong, visible value proposition (“Why buy from us?”)

If you’re pouring budget into backlinks while your product pages are thin and confusing, you’re almost certainly misallocating spend. Move some of that budget into making your site AI-ready and customer-ready instead.

6. Make it easier to buy from you (and easier for AI to recommend you)

Here are practical questions to ask of every key product or category page:

1. Can a shopper easily see why they should buy from you instead of a competitor?
   • Is it Australian-made?
   • Do you have local distribution and fast shipping?
   • Do you offer better warranties or support?
   • Are those points clearly visible above the fold?
2. Are FAQs embedded at the product level?
   • Common questions, objections and comparison points should live right on the product page, not buried in a generic help centre.
3. Are delivery costs and options clearly signposted?
   • If customers (and AI tools) have to dig or guess, you’ll lose trust and conversions.
4. Is your content structured so AI can understand it?
   • Clear headings, bullet points, descriptive labels and well-written copy all help LLMs summarise your offering accurately.
5. Are you using AI to create this content efficiently?
   • Tools like Google Gemini, NotebookLM and Google AI Studio make it easy to:
     • Generate and refine FAQs
     • Draft ad copy and resized ad variants
     • Run competitor analysis
     • Turn raw product info into polished, structured copy

Many teams are now building their own lightweight internal tools or apps to generate FAQs on the fly, perform competitive comparisons and standardise product content. That’s where AI time is best spent: creating assets that help customers make decisions faster.

7. The question that matters now

One prospective client, confident they’d done a lot of UX work, was asked a simple question:

“Why should I buy from you rather than your competitor?”

The answer was a long string of ums and ahs.

If you can’t answer that clearly and quickly, your website can’t either. And if your website can’t, AI tools certainly won’t.

So bring it back to the core:

• How difficult is it to shop from you?
• How easy are you making it – for both humans and AI – to choose you?

If you get that right, you’ll not only lift your existing conversion rate from current traffic, you’ll also position your brand to be surfaced and recommended more often as AI-driven commerce matures.

Call to action

If you’re serious about preparing for an AI-driven shopping journey, start by auditing your key product pages:

• Are your value propositions clear?
• Are your FAQs robust and product-specific?
• Are delivery, returns and pricing fully transparent?
• Can AI tools easily understand and summarise your offer?

From there, explore how AI tools like Gemini, NotebookLM and ChatGPT can help you scale out that content quickly and consistently.

Want to benchmark how “AI-ready” your site really is and where you’re losing conversions? Now is the time to do it—before your competitors become the default answer when customers ask an AI, “Which store should I buy from?” If you have any questions or would like a site review, please feel free to reach out to Jim Stewart at [email protected]

Jim

Jim’s been here for a while, you know who he is.