Oreo Spent $40 Million on Marketing — Yet Went Silent in AI Search

Hi everyone, this is Neo.

Today I want to talk about a business paradox that sounds absurd but happens every single day:

Brands spend big money locking up their own content to “protect” it — then spend even more money pushing that same content to the same people through middlemen.

The author of the original piece gave it a very precise name — the “Protection Paradox.”

This paradox has existed in B2B content marketing for a long time, but in the AI search era, its damage is amplified tenfold. Because what you’re locking away isn’t just a PDF — it’s your right to be seen in AI engines.

1. B2B’s “self-tax”: locking the whitepaper in a vault, then paying to buy the key back

The classic B2B lead gen workflow goes like this:

  1. The content team spends two months producing an industry trends report and calls it “thought leadership”
  2. The marketing team saves it as a PDF and wraps it in a 10–15 field form: name, job title, industry, budget, tech stack, purchase timeline…
  3. Sales says “we only want serious buyers,” so the form gets longer
  4. The content is finally “published” and the team pops the champagne

On the surface, everyone is doing the right thing:

  • Content team: we produced a high-quality asset
  • Marketing team: we generated X MQLs
  • Sales: our lead quality is excellent

But what’s actually happening?

The PDF is trapped behind a form. Search engines struggle to parse it, AI systems can barely extract anything from it, and the people who actually want to read it get scared off by 15 fields. The most ironic part — your prospects, partners, and industry KOLs give up at the form step.

Then the story enters phase two.

To “expand reach,” you license that whitepaper to content aggregators like TechTarget. Their playbook is simple:

  • Break your report into multiple SEO-friendly articles
  • Collect demand with minimal forms
  • Sell the leads back to you at $15–30 each

You used your own content to tax yourself.

Neo’s take: When a lot of B2B independent site owners do content marketing, they treat gating as the default setting. But the cost of gating isn’t just lost readers — in the AI search era, gated content effectively disappears from AI training data and citation sources. While you’re protecting leads, you’re also protecting AI from seeing you.

2. Your content circulates everywhere on the internet — except on your own website

Gating has another, more subtle side effect.

Once that whitepaper is locked behind a form, your organization starts “leaking” those ideas organically — just not through your website:

  • The PR team pulls out the most striking data and charts and packages them for journalists and analysts. The articles journalists write read a hundred times better than your original PDF — and rank better too
  • Customer success sends the report to key clients, who forward it to internal groups, where it gets screenshotted and posted on LinkedIn
  • Partners take your core arguments, add their own spin, and turn them into versions tailored to their audiences — usually clearer, more focused, and easier to find

The result: your ideas are all over the internet, but the source is the hardest place to reach.

Prospects walk into meetings able to recite your framework and data from memory. But when they want to trace it back to the source, they don’t open your website — they open some industry media take, a partner’s landing page, or a third-party content library.

You’re training the entire market to treat someone else’s interpretation as your official version.

Neo’s take: Anyone running an independent site knows backlinks matter. But have you ever considered that your partners, the media, and your customers are essentially building backlinks for you? The problem is, if the source is locked down, those backlinks all point somewhere else instead of you. Your content becomes the industry’s public reference material, while traffic and trust flow into everyone else’s pools.

3. When a “research report” becomes just another funnel

The author of the original piece gives a very typical example.

While writing an article, he needed a statistic on AI adoption rates and found a study that looked authoritative. Clicking through, the page showed only 3–4 headline numbers, one big image, and a “download the full report” button.

Click it, and a multi-layer funnel pops up: email capture, phone confirmation, product pitch, “personalized outreach.” You never get a clean, downloadable PDF with methodology.

That “study” gets cited widely, but it’s fundamentally unverifiable. You’re citing a marketing asset, not research.

That’s the price of the protection paradox at the “knowledge level”: we bury reports to “protect their value,” and the whole market ends up making decisions from unverifiable fragments of data.

Neo’s take: When independent site owners create industry reports, whitepapers, and data studies, you have to be clear about the goal. If it’s about building authority and trust, then it needs to be freely citable, indexed by AI, and spread across the web. If it’s about capturing leads, then design it as a landing page — not as a “research output.” Both goals are valid, but you can’t mix them. Once a research report becomes a pure funnel tool, its credibility drops to zero.

4. The Oreo lesson: blocking AI crawlers is paying to make yourself invisible

B2C brands haven’t escaped this paradox either.

Mondelez (Oreo’s parent company) recently revealed a shocking number in an interview:

Oreo — one of the most famous cookie brands in the world — doesn’t show up 90% of the time when AI recommends cookie-related content.

Why? Because Mondelez, out of a desire to “protect IP and control content,” treated AI crawlers like suspicious bots and blocked them from indexing brand content.

Every department did what it thought was right:

  • Legal: minimize unauthorized use
  • IT: restrict unknown automated traffic
  • Marketing: “we’re Oreo, of course we’ll be mentioned”

The result? Oreo went silent in AI search.

The irony is that Oreo’s marketing spend is enormous:

  • Large-scale social media campaigns and influencer collaborations engineered for virality
  • A partnership with Accenture and Publicis investing over $40 million to build a custom generative AI content platform for social ads, e-commerce images, even TV commercials
  • Expected to cut content production costs by 30–50% using AI tools

On one hand, $40 million spent getting algorithms to talk about Oreo. On the other, robots.txt telling the most important new algorithms “you’re not allowed to look at me.”

That’s the most absurd version of the protection paradox: brands paying to manufacture their own invisibility.

Neo’s take: B2C independent site owners — if you’re weighing whether to block AI crawlers from your site, Oreo is your warning sign. Your competitors won’t block them. AI search and AI recommendations are reshaping how consumers discover brands. If you choose “protection,” all you’re protecting is your chance of being seen. IP absolutely matters, but the risk of being ignored is far greater than the risk of being copied.

5. Why do smart teams keep making this mistake?

The protection paradox keeps happening not because people are stupid, but because systems are stupid.

  • Content teams are measured on engagement and lead counts → naturally biased toward gating
  • Legal is measured on risk reduction → naturally biased toward restricting crawlers
  • IT is measured on controlling access and cost → naturally biased toward blocking unknown traffic
  • Marketing is measured on MQLs → naturally biased toward longer forms

Everyone is optimizing locally, and the sum of it is a globally worst outcome.

The deeper problem: in most companies, nobody is accountable for discoverability.

SEO has an owner. Paid ads have an owner. Social has an owner. Email has an owner. Security has an owner. But “making sure content can be found and used in search and AI-driven discovery environments” is usually implicit rather than explicitly assigned.

So every team hits its numbers and the dashboards all look healthy. But step outside the system and look at the result — the content exists, but it isn’t showing up where it should.

Neo’s take: If you run an independent site, don’t wait for a crisis to coordinate across departments. Assign one person (or one function) explicit ownership of discoverability, spanning content, tech, marketing, and legal. That person doesn’t need to do everything — just make sure departmental decisions don’t cancel each other out. In the AI era, being found is a strategic function, not a side effect.

6. How to design a protection strategy that doesn’t erase you

The author of the original piece is clear about his stance: the answer is not swinging to the other extreme and opening up all your content. Nor is it locking everything down. The key is knowing what you’re protecting, and whether your protection method is quietly working against you.

The core principle: let your core ideas circulate freely, and put the capture gate later in the journey.

How, concretely?

1. Separate “ideas” from “assets”

  • A report’s core arguments, data insights, and frameworks — fully open. HTML format, structured data markup, easy for search engines and AI to crawl
  • Only gate when you need deep interaction (a customized ROI calculator, one-on-one consultations, full data downloads)

2. Put the gate in the right place

  • Don’t gate at the awareness stage — the user hasn’t decided to trust you yet
  • Gate at the evaluation or decision stage — intent is clear by then, and exchanging information feels natural

3. Rethink “protecting IP”

  • Old way: protecting IP = controlling distribution channels
  • New way: protecting IP = making sure your version is the canonical version that gets cited, shared, and built upon
  • That requires content that’s clearly structured, indexable, and easy for AI to parse — so when others cite you, the source points back to you

4. Fix the incentive system

  • If teams are rewarded only for gated conversions, they’ll gate everything
  • If rewards are tied to revenue, long-term value, and presence in search and AI-driven discovery, the trade-offs change completely

Neo’s take: Content strategy for independent sites is essentially a balancing act between “being found” and “being converted.” My advice: spend 80% of your effort on discoverable content (open blog posts, research reports, industry data), and 20% on gated content (demos, tools, deep resources). The former wins awareness and builds authority; the latter filters for high-intent users. Do it the other way around, and you get the Oreo outcome.

Summary

The core lesson of the protection paradox is simple:

The real risk isn’t being copied. It’s being ignored.

When your content can’t be found, people don’t stop asking questions — they just get answers from someone else. That “someone else” might be an aggregator, a partner, a competitor, or a general-purpose AI model that never really “knew” you.

If you spend more money amplifying content through middlemen than making it discoverable at the source, you’re not protecting value — you’re paying a premium for being seen.

The brands that win the next decade will still protect what truly matters. But they’ll honestly distinguish between “guarding assets” and “burying assets” — and make sure they’re not taxing themselves just to be seen.

Hope this post gives you something to think about. If you’re also wrestling with “should I gate my content” or “should I block AI crawlers,” drop a comment — let’s talk.