Name generation
See also: candidate generation
Name generation is the stage where candidates are produced against the chosen creative directions. Volume matters because it makes real elimination possible: a shortlist drawn from twenty candidates is a shortlist of the least bad.
In practice
Generating several hundred candidates is not showmanship. Elimination needs something to eliminate: if only twenty exist, the survivors are whatever was least objectionable rather than what was genuinely strong.
Fortune's profile of the Accenture naming describes the namer reviewing thousands of original and submitted names, with the winner first submitted by an employee, and beating the next 50 candidates in a senior-partner vote by roughly two to one. The same article describes the namer's usual funnel: over 1,000 names developed, 100 to 150 taken to trademark screening, at least 50 low-risk names left, three to five finalists chosen by the client. The numbers are one practitioner's account, not a standard, but the shape is the point: wide first, then hard cuts.
In a naming project the failure is a generation stage that is too narrow or too similar. Twenty candidates built from one idea survive screening in one lump or die in one lump, and the rest is taste. Generate across several creative recipes and directions, keep the output as a longlist that is never shown as a result, and let the cut produce the shortlist. See Business name generation is more than just a name.
Where this sits in the process
The studio treats this as part of a single run rather than a separate service: the category is read, the creative direction is chosen from what the research found, candidates are generated in volume and then cut against criteria that were written down first. See how the studio works for what that produces, or the naming module for the stages themselves.
Related concepts
- Longlist: The working document nobody should ever be handed: what a longlist is for and why receiving one means the work is unfinished.
- Shortlist: What separates a shortlist from a longlist: not length, but the evidence attached to every survivor.
- Creative recipe: The constructed techniques behind name generation: what a recipe is and why choosing them is most of the craft.
Further reading
- Business name generation is more than just a name: A business name has to clear cultural, linguistic and competitive checks before taste gets a vote. What those checks are, and what each one rules out.
Frequently asked questions
What is name generation?
Name generation is the stage of a naming project where candidates are produced against the chosen creative directions. Volume matters because it makes real elimination possible.
How many names should a naming project generate?
Hundreds, usually more. A professional namer quoted by Fortune describes developing over 1,000 names and cutting them to 100 to 150 for trademark screening. A shortlist drawn from twenty candidates is a shortlist of the least bad.
Can AI generate business names?
Yes, and any general AI tool can produce candidates. What decides the result is what comes before and after: the brief, the chosen creative direction and the screening that removes names against written criteria.
What happens after names are generated?
The candidates are cut: first against hard criteria, then through linguistic, cultural and competitive checks, then to a shortlist with the evidence behind each name. Checking a shortlist for existing trademark rights is a separate, manual step that you own.