I watched a familiar scene play out. A well-crafted proposal — clear product content, ingredient explanations in plain language, structured data for Google’s product feeds — went up the chain at a heritage beauty brand.

The response from head office? We agree this could add value. But the way consumers search is changing. AI-driven and agentic search models are becoming increasingly important. Before investing in more content, we should first understand how these new models work.

Translation: nothing happens.

No owner. No deadline. Just a topic “kept on the radar” — which, if you’ve ever worked with a large organisation, you’ll know is where good ideas go to be quietly forgotten.

I see this everywhere right now. Marketing departments at big brands are frightened of AI search — and their fear response is to freeze. It’s the most expensive mistake they could make.

The logic is exactly backwards

The argument goes: AI search is changing everything, so let’s pause our content work until we understand it.

But ask yourself — what do ChatGPT, Perplexity, Google’s AI Overviews and the new wave of AI shopping agents actually read when they recommend a product?

Written, structured content. That’s it. That’s the whole answer.

  • AI search engines can’t watch your beautiful brand videos.
  • They can’t absorb the benefits explained by your lovely presenter at minute 2:14.
  • They can’t infer what your hero ingredient does if it only appears as an unpronounceable INCI name.

They read text. They read structured data. They read product feeds, FAQs, specifications, and clear benefit statements. If your product pages are thin on written content, you are — quite literally — invisible to the fastest-growing discovery channel in existence.

So “waiting until we understand AI search” before adding written content is like refusing to learn to swim until you understand water. The water doesn’t wait.

Why big brands freeze (and small ones win)

Having worked both sides of this fence for years, I have some sympathy. Inside a large organisation, the incentives are upside down:

  1. The worst outcome politically isn’t losing traffic — it’s losing control. A local market or distributor succeeding with something head office didn’t lead is more frightening than a slow, shared decline.
  2. Process becomes armour. Working groups, global frameworks, “strategic reviews” — they all feel like action while guaranteeing none is taken.
  3. AI is feared twice over. It’s the reason to distrust new content (“was this written with AI?”) and the reason to delay all content (“we must understand AI first”). You can’t win against that logic. You can only watch competitors route around it.

Meanwhile, smaller and nimbler businesses simply… do the work. They publish clear benefit-led copy. They mark up their pages with structured data. They optimise their product feeds. And they’re the ones being cited when someone asks an AI assistant, “What’s the best serum for a damaged skin barrier?”

The gap is widening every month. AI search doesn’t reward brand heritage. It rewards clarity, structure and substance.

What “AI-ready content” actually means

The good news: preparing for AI search isn’t mysterious, and it isn’t separate from good SEO. It’s largely the same work, done properly. Here’s where I focus with clients:

1. Say what the product does — in plain language

Key ingredient’s scientific name — nobody is searching for that. They’re searching for “repairs the skin barrier” or “long-lasting hydration.” Every product page should explain its benefits the way a customer would describe them to a friend. AI models match meaning, not jargon.

2. Get the words out of the videos

Video is brilliant for humans and nearly useless for machines. Transcribe your videos. Turn the key points into on-page text, FAQs and bullet points. You’ve already paid for that messaging — make it readable.

3. Structured data is non-negotiable

Product schema, FAQ schema, review markup, ingredient and specification data. This is how search engines — traditional and AI alike — understand what your page actually contains. It’s also what powers rich results and feeds AI shopping agents.

4. Treat your product feed as a shop window

Google Merchant Centre feeds increasingly drive both Shopping ads and free listings — and they feed AI-powered shopping experiences too. Optimised titles, complete attributes, clear descriptions and product highlights aren’t admin. They’re sales infrastructure.

5. Answer real questions on the page

AI Overviews and chat assistants love content that directly answers questions: Is this suitable for sensitive skin? How long until I see results? Can I use it with retinol? If your page answers them, you can be the cited source. If it doesn’t, your competitor will be.

The uncomfortable truth

Traditional search is evolving — that part of the corporate hand-wringing is true. Click-through rates from classic blue links are falling. More questions get answered without a website visit at all.

But the response to that isn’t less content. It’s better-structured, more substantive content — because in a world where an AI intermediary decides which brands to mention, the brands with clear, machine-readable, genuinely helpful information win the recommendation.

There is no future version of search — agentic, conversational, or otherwise — where having well-structured written content about your products is a mistake.

Don’t wait for the strategy committee

If you’re a business owner or a marketing lead in an organisation that can move quickly, this is your moment. The big players are paralysed in meetings about “understanding the new landscape.” Every month they spend understanding it, you can spend ranking in it.

Start with the basics:

  • Audit your product pages: would an AI reading only the text understand what you sell and why it’s good?
  • Add structured data everywhere it’s relevant.
  • Optimise your product feeds properly.
  • Turn your video and packaging messaging into on-page text.
  • Answer your customers’ real questions, in writing, on the page.

None of this requires waiting for anyone’s permission — and all of it works for traditional SEO today while positioning you for whatever search becomes tomorrow.