01
the total belongs to code.
the rubric weights, level mapping, assessed score and coverage are deterministic. no model or user-supplied total can overwrite them.
BRAND NAME SCORE // LOCAL DIAGNOSTIC
score your brand name before you build around it.
put one candidate through a fixed 100-point rubric. the machine scores 75 inspectable form points, leaves audience meaning and search collision unassessed, and keeps legal certainty out of the theater.
free · no email · no model call · no search or legal check
01 // BRAND NAME SCORE
put the name on the instrument.
the name is enough to start. add context to keep the brief beside the result, not to manufacture semantic evidence.
result parked. the inputs changed. run the score again before building.
ASSESSED RESULT // 75% COVERAGE
/ 100 assessed
- coverage
- confidence
- strongest 2
- primary weakness
02 // DIMENSION LEDGER
every point leaves a receipt.
03 // WHAT WOULD RAISE IT
fix the weak joint, then run it again.
HOW THE INSTRUMENT THINKS
a fixed rubric, with the trapdoors labeled.
02
context is a brief, not proof.
the tool records business and audience context without scoring it as semantic fit. that question stays open until the intended audience supplies evidence.
03
form evidence has a boundary.
letter shape, length and pronunciation cues are inspectable. market response and cultural meaning still need people from the intended audience.
THE 100-POINT RUBRIC
7 dimensions. 2 stay empty on purpose.
distinctiveness
whether the name has a recognizable shape rather than generic category language
20 points
memorability
recall potential after brief exposure
15 points
pronunciation
whether people are likely to say and repeat the name consistently
15 points
audience and positioning fit
semantic fit with the intended audience and position, which is not inferred from supplied context
15 points
visual identity potential
whether the name can support a useful wordmark, mark, and wider visual system
15 points
product-range extensibility
whether the name can survive expansion beyond the first product or service
10 points
surface-level search collision
obvious visible conflicts found through an implemented public search check
10 points
WHERE IT STOPS
the machine knows the edge of the map.
search collision is not assessed.
this version runs no public search request. the missing 10 points reduce coverage instead of receiving an inferred score.
a score is not legal clearance.
the result establishes no trademark, company-name, domain or social-handle availability. check the appropriate records before committing.
audience fit is not assessed.
the missing 15 points reduce coverage. local code can record your context and test explicit constraints, but it cannot prove how the intended audience will interpret a word.
WORKED METHOD FIXTURES // COMMITTED EVIDENCE
3 runs the code has to repeat exactly.
these are deterministic method fixtures, not name suggestions or availability claims. audience fit and search collision remain not assessed in every run.
v02-score-fixture-001
tenon
a compact name for design-aware clothing and daily accessories
- assessed score
- 92.1 / 100
- coverage
- 75%
- confidence
- medium
- weakest joint
- product-range extensibility
v02-score-fixture-002
global creative solutions
a descriptive studio name tested for generic language and recall load
- assessed score
- 65.2 / 100
- coverage
- 75%
- confidence
- medium
- weakest joint
- distinctiveness
v02-score-fixture-003
field skin
a category-adjacent skincare name tested for room to expand
- assessed score
- 82.5 / 100
- coverage
- 75%
- confidence
- medium
- weakest joint
- product-range extensibility
FAQ
the part numbers cannot decide.
what makes a business name good?
a useful name is easy enough to say and remember, distinctive enough to recognize, suited to the intended customer and flexible enough to survive the next product. visual potential matters too, because the word must work on real surfaces. none of those qualities establishes legal availability, so the commercial checks sit outside this score.
how is the brand-name score calculated?
7 dimensions carry fixed weights totaling 100. deterministic code maps each assessed level to points, adds those points and renormalizes the result across assessed dimensions. audience and positioning fit contributes 15 unassessed points, and search collision contributes 10 unassessed points. coverage is therefore 75%, with neither gap receiving an invented estimate.
can AI determine whether a name is trademarked?
not from a naming score. a proper clearance process depends on jurisdiction, classes, current records, confusing similarity and legal judgment. this tool does not run that process. use the appropriate official databases and professional advice before treating a candidate as clear.
what score should a brand name achieve?
there is no universal pass mark. use the score to locate avoidable weaknesses and compare candidates under the same context. a lower-scoring name with real customer resonance may still be the right decision. small differences should never be treated as mathematical destiny.
is a short name always better?
no. shorter names often reduce recall load and give a wordmark more room, but brevity can also produce generic language, ambiguous pronunciation or heavy search collision. the right length is the shortest form that still carries a recognizable idea and works when spoken.
how can i improve a weak brand name?
start with the weakest assessed dimension and change one property at a time. remove generic category filler, simplify competing pronunciations or test a less restrictive product word. test audience meaning separately, then score the revision under the same inputs so the structural comparison stays fair.
can i compare two names directly?
yes. the comparison tool runs 2 to 5 candidates through this same rubric with one shared context. it blinds internal evaluation order, exposes the full matrix and calls results essentially tied when the leading assessed scores sit within 3 points.