Evidence over opinion
Every assessment in RBSI is answerable from evidence that can be named, dated and checked. Qualitative judgement is never excluded — it is anchored to something observable. Version 1.1.
Assessment points and benchmark rules
An assessment point is a single fact or measurement about a brand. Every scored point is measured against a governed benchmark rule, and the result is a fixed evidence rank, so two brands with the same value always receive the same rank.
| Evidence rank | Meaning |
|---|---|
| Critical | Severe enough to be escalated to the risk assessment, from any topic. |
| Weak | Clearly below the benchmark. |
| Neutral | In line with the benchmark. A genuine result, never a default for missing evidence. |
| Positive | Above the benchmark. |
| Strong | Well above the benchmark. |
| Very Strong | Exceptional against the benchmark. Requires broad, independent corroboration at topic level. |
A benchmark rule is not one universal formula. It is matched to the nature of each measurement: absolute levels, comparison with a defined peer group, change over time, ratios, desirable ranges, severity of an adverse condition, and qualitative evidence states. Rules are set centrally and are never set by the brand being measured.
Get the full document
How RBSI works, its topics and assessment points, and its governance, in a single reference document (Version 1.1, Creative Commons edition).
How a topic is assessed
Evidence coverage and stability
Every score is published with an evidence-coverage indicator and a confidence level, which show how much of the model could be assessed. Coverage never changes the score. The published score also moves only gradually between scoring events, so a change reflects the evidence rather than noise. The calculated score is always retained for audit.
What the framework does not do
Leasing systems that consume RBSI outputs apply their own site-fit logic on top: footfall counts, void analysis, lease economics. The framework evaluates brands; site decisions are downstream.
RBSI outputs are relative index readings, not a DCF model, a transaction multiple, or an audited valuation. They inform valuation thinking — they do not replace it.
It reads evidence of consumer activity and translates it into intelligence about strength, momentum and affinity. It does not model individual purchase decisions.
It structures and evidences analytical judgement. The framework does not produce conclusions on its own.
Versioning