What the engine measures, and how.

Francesco Massara · The author

Foundations

The full reference list of each chapter is in the book. This page tracks only the author’s own works and the philosophical foundations the framework depends on; when a status changes, it changes here first.

Every brand on this site is described by four capacities and by their combination, the Eleatic Power (EP). The four capacities come from Part V of the book: Attractiveness (drawing objects, people and occasions into the brand’s orbit), Influence (propagating practices and change beyond its perimeter), Homogeneity (preserving the recognizability of the signature) and Connections (relating the objects of the assemblage to each other).

The engine was born as the answer to a reviewers’ question: what would an empirical researcher do differently with this theory? This is what.

EP combines the four capacities so that imbalances cost: a brand cannot compensate a collapsed capacity with an inflated one, which is the configurational thesis of Chapters 11–12. The profile (Balance, Drifty, Continuity, Strong Connections) is the shape of the deviation of the four capacities from their own mean, compared with the three archetypes of Figure 12.2; when the deviation is small, or resembles none of them, the profile is Balance.

Evidence discipline. The engine scores the evidence, not the fame: every level must be justified by a dated, sourced evidence line, interpreted within the brand’s category, and absence of evidence is stated as such rather than inferred from silence.

Dated, never overwritten. Every assessment is a snapshot with a date. Snapshots are archived, never edited: a new one is added beside the old. Brands in the living dataset are re-assessed every 30 days; the re-assessment starts from the previous snapshot and may change a level only on new or contradicting evidence.

Readers can report, editors verify. Under every indicator a reader can report missing evidence. Reports enter a verification queue; verified reports are handed to the engine at the next assessment as evidence to check. A report never changes a score by itself.

Comparison sets. Benchmarks compare a brand with its category when the category holds at least five brands, otherwise with its arena (nine groups of related categories), and say so. Medians and dispersion describe the set; they are not targets.

Attractiveness · A1 Context breadth · A2 Relational in-degree · A3 Demand pull
Influence · I1 Practice diffusion · I2 Norm & standard setting · I3 Linguistic penetration
Homogeneity · H1 Signifier continuity · H2 Drift control · H3 Cross-context consistency
Connections · K1 Internal coupling · K2 Network density · K3 Data & identity continuity

The indicators are discussed in chapters 11 and 12 of the book.

The full scoring protocol – anchors, search plan and rubric – is documented in the Methodology Note v0.3, available on request to researchers and institutions: [email protected].

The engine reads public web evidence only: company sites, press, trade publications, rankings, public registries. It never uses private data, surveys or the site’s own users’ data. Each evidence line names its source and its date, so that a reader can check it.

Two layers of data coexist and are labelled everywhere. The seed cache (methodology v0.1–0.2, September 2026) is a desk-based synthesis of one hundred global brands, scored directly on the four capacities with four evidence lines each; it exists so that the Observatory is readable from day one. The living dataset is made of model-run assessments with the twelve indicators and dated, sourced evidence; a brand moves from the seed to the living dataset the first time it is assessed live, and from then on it is re-assessed every 30 days.

Brands are classified in 43 categories grouped into 9 arenas (Fashion & Luxury · Food & Beverage · Technology & Software · Mobility & Energy · Finance · Media & Telecom · Retail & Commerce · Industry & Services · Public, Education & Sports). The engine must choose one of the 43 labels; it may not invent categories.

It measures what leaves public traces. A brand with a strong, local or scarcely digitized materiality may be under-represented, because the engine can only find what is written somewhere. The book says it in Chapter 12: the model does not replace observing the network; it suggests where to look.

It is a model reading evidence, not a survey and not a balance sheet. Scores are model-run: two assessments of the same brand on different days can differ because the evidence found differs. That is why every score carries a date and a confidence level, and why history is shown rather than hidden.

Descriptive, not causal. A change between two snapshots is a change in the evidence the engine could find on those dates. The site explains movements from the evidence; it does not attribute them to business results.

Early dataset. Categories with fewer than five brands fall back to their arena for benchmarks, and the page says so. Movers appear only when a brand has at least two dated snapshots.

A change of version is announced here before it is applied, with what it changes and why. Scores computed with different versions are never mixed in a trend without saying so.

The book. Massara, F. (2026). From the Mind to the World: Discovering What Brands Really Are. Tortona: Francesco Massara (Amazon KDP). ISBN 979-12-986667-0-2. Italian edition: Dalla mente al mondo. Alla scoperta di che cosa sono davvero le marche. ISBN 979-12-986667-1-9. Both editions are available in paperback and Kindle on Amazon worldwide.

A score from this site. Massara, F. (2026). Eleatic Power assessment of [brand], snapshot of [date], methodology v0.3. fromthemindtotheworld.com/brand.html?name=[brand]. Retrieved [date].

A research idea from this site. [Author] (2026). [Title of the idea]. Research Ideas, From the Mind to the World. fromthemindtotheworld.com/ideas.html#[id].

Results deposited on this site. [Authors] (year). [Title of the work]. Results deposited under “[Title of the idea]”, Research Ideas, From the Mind to the World. fromthemindtotheworld.com/ideas.html#[id].

Every brand page carries the snapshot date and the method version you need for the citation.

The hypotheses, counter-intuitions and open questions the model makes visible are collected on the Research Ideas page, each stated as a research question with a possible test. They live in the Academic area, open to theses and empirical work. Readers and researchers can propose their own; accepted ideas are published with the proposer’s name. If you are interested in using the database for research, write to [email protected].

Saving your work in My Space is private by default. Separately, and only if you choose so, you will be able to contribute your observations, anonymized at the source, to a research archive of readings, measurements and comparisons. The two consents are distinct and never confused: private teaching data versus voluntary scientific contribution. The option is being prepared and will appear in My Space.

Reports of missing evidence are open to every registered reader today, on every brand page.