Brand consistency calculator
The brand consistency calculator turns ten signals from a brand's pages into a single 0 to 100 score. This page shows exactly how that calculation works, and lets you run it free on any website.
The problem: a score nobody can reproduce is just an opinion
A consistency number is only useful if you can see how it was reached. A black-box score invites the obvious objection: says who, and on what basis. To be trusted and acted on, the calculation has to be explicit and repeatable.
That is the bar this calculator is built to meet: the same inputs produce the same score, and every point is traceable to a signal on the brand's own pages.
Why a single gut-feel rating is not enough
A reviewer's one-to-ten gut rating cannot be reproduced, compared, or defended. It also tends to anchor on whatever the reviewer noticed first. A calculation built from fixed components avoids that: it forces every dimension to be considered and weighted the same way each time.
The point of a calculator is not false precision. It is consistency of method, so the number means the same thing across brands and across time.
The calculation, worked through
Each of the ten components is scored from 0 to 20 from public website signal. The five positive components (message clarity, trust signals, customer-promise consistency, reputation evidence, voice coherence) add coherence. The five negative components (identity confusion, trust drift, audience confusion, risk signals, operational dissonance) capture friction.
The Net Entropy Score is the positive signals minus the negative ones. For the website layer, that net is mapped onto a 0 to 100 brand consistency score. A brand strong on message and voice but weak on trust signals and carrying identity confusion lands in the 60s; closing those gaps moves it toward the low 70s and above.
The full formula is defined in the NES working paper (Appendix A).
Run the calculator
You do not have to score the ten components by hand. The free scanner runs the full calculation on any public website and returns the 0 to 100 score in about two minutes, with every component's score, reasoning, and quoted evidence so you can see how the number was built.
For a deeper calculation, the Review-Inferred tier runs the same model over public customer language, and the Measured tier over a direct cohort survey.
The ten components in the calculation
These are the inputs to the score. Positives build it up; negatives pull it down; the score is the net, expressed 0 to 100 for the website layer.
Frequently asked
How is the brand consistency score calculated?
Ten components are each scored 0 to 20 from public website signal. Five positive components are summed, five negative ones are subtracted, and the net is mapped to a 0 to 100 brand consistency score. The full formula is in the NES working paper.
Can I calculate it myself?
You can approximate it by reading the key pages and scoring each component, but the free scanner runs the full calculation automatically and shows the evidence behind each component.
What score should I aim for?
Across 119 DTC websites in June 2026, the typical score was 71 and the highest 80, so the low 70s is roughly the current bar. The component breakdown shows which inputs to improve to get there.
Is the calculation the same for every brand?
Yes. The same ten components and the same formula are applied to every brand, which is what makes the scores comparable.