01
one context keeps the bout fair.
every candidate receives the same business, audience, desired perception and geography. changing the brief between names invalidates the comparison.
NAME COMPARISON // 2 TO 5 CANDIDATES
compare your business-name shortlist side by side.
one context. one rubric. every candidate enters in a blinded order, and a narrow lead stays a tie instead of becoming fake certainty.
free · no email · no model call · no search or legal check
01 // SHORTLIST INPUT
bring the whole shortlist.
2 names are enough. add shared context and the dimensions that matter most, then inspect the matrix before choosing.
result parked. the shortlist changed. compare it again before choosing.
ASSESSED RESULT // SHARED RUBRIC
02 // SIDE-BY-SIDE LEDGER
same weights. same context. visible gaps.
03 // THE CASES
why the leader leads, and why the runner-up stays alive.
HOW THE INSTRUMENT THINKS
a fixed rubric, with the trapdoors labeled.
02
input order does not decide the ranking.
candidate names are placed into a stable blinded order before scoring. permutation tests cover every order of a 5-name shortlist.
03
close means close.
a gap of 3 assessed points or fewer enters the tie-break. selected assessed priorities come first, confidence comes next and equal evidence stays tied.
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.
the matrix compares form, not ownership.
search collision is visibly not assessed. the table does not establish trademark, company, domain or handle availability.
the highest number is not an instruction.
the strongest local diagnostic can still lose once real audience response, strategic meaning and legal review enter the room.
priorities cannot fill empty cells.
priorities can separate equal or close assessed scores only when every candidate has evidence for that dimension. audience fit and search collision never enter this version.
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-compare-fixture-001
clear local leader
a compact name against a long descriptive studio construction
- 1tenon92.1 / 100
- 2global creative solutions65.2 / 100
tenon leads the next candidate by 26.9 assessed points on the shared rubric
v02-compare-fixture-002
essentially tied
two equally compact verbal shapes used to prove the tie boundary
- joint 1luma92.1 / 100
- joint 1noma92.1 / 100
the leading assessed scores are equal; no assessed priority was selected, and confidence is equal, so the result remains tied
v02-compare-fixture-003
clear local leader
a 3-name shortlist with distinct range and wordmark trade-offs
- 1orlo92.1 / 100
- 2morrow house85.2 / 100
- 3field skin82.5 / 100
orlo leads the next candidate by 6.9 assessed points on the shared rubric
FAQ
the part numbers cannot decide.
how should i choose between two business names?
judge both names against the same customer, positioning and commercial plan. compare the dimension evidence, say each name aloud, test it on likely products and ask the intended audience what they remember without prompting. then complete the relevant legal and availability checks before committing.
what criteria matter most when comparing names?
distinctiveness, memorability, pronunciation, visual potential and product-range extensibility expose structural differences. audience fit matters to the decision, but this tool does not score it from a written brief. declare the assessed structural dimensions that matter most, then test meaning with the intended audience.
is the highest-scoring name always the best choice?
no. the score is a transparent local diagnostic, not a command. real customer memory, cultural meaning, founder conviction, legal review and commercial constraints can change the decision. use the ranking to ask better questions, then keep the evidence beside the judgment.
how close can two name scores be before they are effectively tied?
a difference of 3 assessed points or fewer enters the close-result group. selected dimensions assessed for every candidate break the tie first, followed by confidence in the structural evidence. when neither separates the names, the result remains essentially tied.
should i ask customers to vote on a business name?
a simple popularity vote often rewards the most familiar option. a better test asks people to recall each name later, pronounce it from writing and explain what kind of business they expect. keep the sample close to the intended audience and do not present the vote as legal or strategic proof.
how do i compare logo potential before designing the logo?
set every candidate in a plain wordmark at small, wide and square sizes. look for awkward length, repeated shapes and letters that can support a useful mark. the matrix includes a form-based visual-potential reading, but a real identity still needs to be designed and tested on actual applications.