How Troviun measures structural overlap.
The reading is always relational: we measure the structural distance between your product and the products verified at the source. The result is a structural overlap index from 0 to 100: the higher it is, the greater the convergence with the market.
1. The five dimensions
How the user moves through the product: the order in which decisions appear and the path to the core action.
What the product claims to solve, for whom, and the narrative that supports that promise.
The format of what the AI returns to the user and the kind of decision that output drives.
Which pieces the product exposes and how they fit together.
The combined choices that identify a product even after all the copy is replaced, the hardest dimension to disguise.
The dimensions do not weigh the same: the structural signature carries more than the others, because it depends least on language. The exact weights and the aggregation formula are not published here.
2. Category convention versus execution choice
Every shared signal found at the source is classified into two types. Convention is what any product in the category adopts, it distinguishes no one and weighs little. Execution is the authored choice that could have been made differently and was not: that is what truly brings two products together.
A report with many convention signals and no execution signals describes a crowded market. The opposite describes occupied ground.
3. Verdict bands
direct comparison: Similaridade estrutural muito alta. Risco reputacional relevante.
market: Um produto no mercado replica sua estrutura de perto.
direct comparison: Vários elementos coincidem. Vale diferenciar antes de escalar.
market: Existem produtos parecidos, mas ainda há espaço para posicionamento.
direct comparison: Semelhanças pontuais, dentro do esperado para a categoria.
market: Nenhum produto similar estrutural relevante encontrado.
4. Chain of evidence
No product enters the report without being fetched at the source. Every scan stores the content and metadata collected from each URL, with a date. On paid plans, each snapshot receives a SHA-256 hash and a UTC timestamp: a third party can recompute the hash and confirm the content is the same one that produced the reading.
The report references the engine version that produced it, which allows re execution and comparison over time.
5. What Troviun measures well, and what it does not
The instrument was built to detect recent structural copying between a product and whoever imitated it, and to measure convergence across products in the same portfolio. That is where the reading has practical value. Comparing mature giants of a category, which arrived at similar formats independently over years, sits outside what the reading answers well: in those cases the resemblance is usually category convention, not copying.
The full methodology, with weights and formula, is versioned with the engine and referenced in every report.
The Troviun index is a statistical indicator of structural similarity. It is not an expert report or legal opinion. The reading describes structural similarity observed on public pages, with dated evidence. Any legal conclusion is up to your lawyer.