Show up in the sources the AI cites, not in the answer — the layer-zero strategy
67% of the sources ChatGPT cites most are off-limits to brands directly. But you can get into those sources. Here's how the layer-zero strategy works.
There's a way of thinking about AI visibility that most companies ignore — and that, once understood, opens a completely different strategic path.
The conventional logic is: "I want my company to show up in ChatGPT's answer." The layer-zero logic is: "I want my company to show up in the sources ChatGPT will cite — and those sources will carry me into the answer."
The distinction seems subtle but it completely changes what you do in practice. Competing for the direct answer means trying to appear in the text the AI generates. Competing on layer zero means getting into the documents, pages, and platforms the AI uses as the basis for building that text.
Why 67% of ChatGPT's top sources are off-limits to brands directly
Analysis of the sources ChatGPT cites most shows that roughly 67% of the most frequently referenced URLs are domains where brands can't create content directly: Wikipedia, .gov, .edu, major press outlets (NYT, BBC, and their equivalents), the App Store, the Play Store, and other institutional repositories.
That means two-thirds of the highest-weight sources for ChatGPT citations are, in principle, outside any company's direct control. You can't create a Wikipedia page for your company without meeting strict editorial criteria. You won't appear on a .gov or .edu on your own initiative. You don't control what major press outlets write.
But "off-limits to direct control" doesn't mean "off-limits entirely." There are legitimate paths into those sources — and the company that masters those paths gets access to a citation layer far more defensible than any content on its own site.
The four layer-zero sources and how to get into each
Layer zero 1 — Wikipedia. The most familiar and most regulated. To create or expand an article about your company or about a topic you're a reference on, you need:
- Proven editorial notability (significant coverage in independent sources)
- Neutral content, with no promotional tone
- Verifiable references from reliable external sources
- Compliance with Wikipedia's editorial policies
For companies without enough editorial notability, the alternative is to create or expand articles about the sector topics where you have expertise — not about the company, but about the field it operates in. A well-developed article on "fleet management at logistics companies" created by a company in the sector is a layer-zero asset for queries related to that topic.
Layer zero 2 — Wikidata. More accessible than Wikipedia, Wikidata accepts any verifiable organization. A Wikidata item with your company's correct attributes (type, sector, location, founding, site, social profiles) puts your brand in the entity database that feeds both Wikipedia and multiple AI systems.
Layer zero 3 — Press coverage in authoritative outlets. Being mentioned in a story in a major newspaper, a top business publication, or a high-authority sector outlet is the most powerful way to get into the layer zero of AI citation. That coverage isn't bought — it's earned through a relevant angle, exclusive data, cases of genuine editorial interest, or positioning as a go-to expert.
The most efficient path for most companies: have a relevant proprietary data point and present it to journalists as an exclusive. Research with a novel finding about the sector has a real chance of coverage — and that coverage becomes a layer-zero source.
Layer zero 4 — Institutional directories and industry associations. For every sector, there are directories and associations AIs treat as authoritative sources: business registries (with verifiable company data), securities regulators (for public companies), professional boards (for medicine, law, dentistry, and other licensed professions), industry associations, chambers of commerce. Being registered and up to date in these institutional sources doesn't generate a direct citation in the answer, but it creates the verification layer AIs use to confirm that a company exists and operates in its declared domain.
The second-order effect: layer zero generates layer one
The layer-zero logic has a compounding effect: being in the highest-authority sources increases the probability of appearing in medium-authority sources, which increases the probability of appearing in the AI's final answer.
When your company is on Wikidata, a press story about the sector is more likely to mention it as an example. When it's in press stories, specialized blogs and publications are more likely to cite it. When specialized blogs and publications cite it, the AI is more likely to use it as a reference in the final answer.
Building layer zero doesn't replace producing your own content — it complements it. Without your own content, AIs have no material to use in the answer even if they know the company exists. Without layer zero, your own content lacks the external corroboration that increases the model's confidence in the citation.
How Crowly can help you measure the layer-zero effect
The impact of layer-zero actions (Wikipedia, Wikidata, press coverage, institutional directories) on AI visibility is slow — slower than publishing a blog post. Crowly, with weekly monitoring, lets you see the trend over months and correlate visibility spikes with specific layer-zero-building actions.
See what your current baseline is before you start working on layer zero. Free diagnostic, immediate result. Analyze my brand →
Sources:
