A brand naming process is the structured sequence a team follows to get from brand strategy to a name that is legally clear, easy to say, and still usable when the business grows. To run a brand naming process properly you align on criteria, generate widely, screen continuously and document the decision.
Naming is not a brainstorm with a lucky ending. Practitioners describe it as a craft with a process, and the difference shows up in the outcomes: teams that run the steps end up with a name they can trademark, explain and extend, while teams that skip straight to favourites end up filing for a mark that somebody else already owns.
This guide walks through the whole sequence — brief, research, territories, generation, screening, consumer testing and legal clearance — with the stage volumes, timings and facilitation detail that usually get left out.
Table of Contents
- What You Need
- How to Run the Brand Naming Process Step by Step
- Write a Clear Naming Brief Before You Run a Brand Naming Process
- Research Existing Names and Market Conventions
- Generate Diverse Name Ideas
- Screen Names for Meaning and Usability
- Test the Shortlist With the Audience
- Check Legal and Practical Availability
- Present and Approve the Final Recommendation
- Common Mistakes
- Frequently Asked Questions
- How long should a brand naming process take?
- Should a brand name be tested with customers?
- What is the difference between a brand name and a product name?
- When should trademark checks happen in the naming process?
- How many brand name ideas should a team generate?
- Can a brand name be changed later if the business evolves?
- Conclusion
What You Need
You can run this with a small in-house team or hand it to an agency, but six inputs have to exist before anyone generates a single name. Gathering them first is what stops the process from stalling in round three.
- The brand brief: purpose, values, positioning, personality and the unique selling proposition, written down rather than remembered.
- Customer insight: the words customers actually use for the problem, gathered from interviews, support tickets, reviews or a short survey.
- Competitive name audit: every name already working in your category, plus the two or three adjacent categories you might move into.
- Category expectations: whether buyers in your space expect plain-English descriptive names or something abstract.
- Linguistic constraints: markets you sell in, languages your team speaks, and any cultural taboos already known.
- Trademark position: a current search on the obvious candidates before you invest effort, plus a named attorney or trademark agent if you intend to file.
Decide your stage volumes before you start. The numbers below are the working default most teams land on after a couple of rounds, and they stop the two failure modes that waste a budget: generating too few names to have any choice, and generating so many that nobody can hold them.
| Stage | How many names | Who is in the room |
|---|---|---|
| Generated | 100 to 300 | Facilitator plus 5 to 8 stakeholders |
| Longlist | 20 to 30 | Naming lead |
| Shortlist | 5 to 8 | Naming lead plus legal |
| Finalists | 2 to 3 | Decision group |
Budget and timeline scale with who runs it. A solo founder working from a written brief can compress the whole sequence into two to four weeks. A team running a facilitated workshop with consumer testing usually needs six to ten weeks. An agency-led project adds the weeks spent on legal clearance, which depend on how much conflict the search turns up.
How to Run the Brand Naming Process Step by Step
The process moves in one direction: align on strategy, define the criteria, generate widely, narrow on evidence, then clear it legally. Each phase produces an output the next phase consumes, and a phase with no output is a phase that quietly gets skipped.
Write a Clear Naming Brief Before You Run a Brand Naming Process
The brief turns brand strategy into criteria a stranger could score a name against. Without it, the discussion collapses into taste, and taste is not a decision method.
Fill in these fields: positioning statement, personality traits with a worked example for each, primary audience and secondary audience, the category the business competes in today and the two it might enter, the words customers use, the geographies that matter, and a hard length and pronunciation rule. Two to three syllables works for almost everything; a name under six letters is hard to protect and hard to trademark distinctly.
Add kill criteria while you are still calm. Write down what would disqualify a name outright: no trademark registration possible in the relevant class, an unfortunate reading in a target language, spelling that cannot be guessed from the sound. Agreeing these before anyone falls for a favourite is the cheapest insurance in the whole process.
Research Existing Names and Market Conventions
Research tells you which direction is already crowded. Look at the competitive name audit you gathered and map what the category has settled on: descriptive compounds, short Latinate words, invented constructions, founder surnames.
Run a collision scan on anything that feels promising before you spend time on it. A search engine, a domain lookup, an app store check and one trademark database will cost you ten minutes and will remove the candidates that are already spoken for. Then check how names behave outside English — the Coca-Cola transliteration into Mandarin Chinese had to be redone after the first sound-alike carried unintended meanings in some dialects, and that rework happened after the brand was already live in market.
Also read the category conventions critically. In B2B software an abstract name is normal; in a grocery aisle a purely invented name costs you explanation at the shelf. Notice where every competitor sits and pick the space they have left.
Generate Diverse Name Ideas
Broad, varied generation is what makes the shortlist defensible. Set three to five naming territories first — distinct routes rather than three flavours of the same idea — then generate inside each one. A useful grid crosses approach, from descriptive to abstract, against construct, from real words to coined ones. Four territories might be descriptive-plus-real, suggestive-plus-real, abstract-plus-real, and coined.
Then run the workshop properly. Ninety minutes is enough if you hold the structure. Assign a facilitator whose only job is to keep time and enforce the rules, a scribe who writes every name up without comment, and a participant list of five to eight people who actually buy or sell the product. Keep the founder, the legal counsel and the loudest opinion out of the generation session and give them the shortlist instead — they will optimise rather than create.
Inside the room: open with the brief and the kill criteria, ten minutes. Silent individual ideation for fifteen minutes so the senior person does not anchor the room. Round-robin sharing, scribe only, no discussion, twenty minutes. Then a break, because the useful names arrive in the second half once people stop being polite. Work each territory separately with a prompt per territory — customer language, product benefit, personality trait, place, material, metaphor — and aim for 30 to 50 names per territory. Close with a silent, ranked vote that produces a longlist of 20 to 30 and no argument.
Rules that matter: no discussing a name until it is on the board, no killing during generation, no falling in love. If someone dominates, give everyone the same number of sticky notes and cap them at three.
AI tools fit into this phase as a volume generator and nothing else. Useful for producing fifty candidate constructions in a minute. Not useful for judging them: the output skews generic, it will not tell you whether a name is registered or taken, and any availability claim it makes needs checking yourself.
Screen Names for Meaning and Usability
Screening turns the longlist into a shortlist by applying the brief’s criteria, and it is where research beats instinct. Five tests carry most of the weight.
- Memorability: can someone recall it after one hearing? Names people can repeat back after a single exposure survive contact with customers.
- Pronounceability: if a customer cannot say it, they cannot search for it or recommend it.
- Spelling consistency: if the sound does not predict the spelling, expect misspellings forever and budget for the domain variants.
- Distinctiveness: arbitrary and coined marks protect better than descriptive ones, which is a legal point and a marketing one.
- Extensibility: can you add a product line, enter a second market or sell to a second audience without the name contradicting you?
Add a linguistic pass. Short names with clear vowel structure read as more familiar, which is the processing-fluency effect documented by Lowrey and Shrum, and hard consonant clusters read as harsher. Sound symbolism work by Klink and later by Argo, Popa and Smith connects certain sound pairings to concepts like softness or speed, whether or not you believe it consciously. Check every finalist in translation, ideally with a native speaker rather than a tool.
The Coca-Cola case is the standard warning: a transliteration that read as something other than “happiness in the mouth” in certain dialects had to be replaced after launch. Screen in every language you will ever need, not the ones you have today.
Test the Shortlist With the Audience
Testing is not a vote on which name people like best. It is a measure of how people interpret the name and what they attach to it, since consumers invent meaning for anything you give them.
Start qualitative. Two or three conversation sessions of six to eight target customers, ten minutes each on a name, no brand context, no logo. Ask what they think the company does, what the name suggests about it, and what kind of product it would fit. You are listening for misinterpretations and for associations you cannot control. If someone hears “cleaning” in a name you meant as “speed”, that is data.
Then run a quantitative pass if the decision is close: a survey presenting the shortlist, measuring unaided and aided recall, appeal, perceived fit with the category and stated preference. Keep the recall task separate from the appeal task, because a name can be appealing and still vanish from memory, and memory is what builds brand equity over years.
Do not run consumer testing before legal screening on the finalists. There is no point measuring emotional response to a name you cannot register.
Check Legal and Practical Availability
Screen continuously rather than once at the end — but distinguish two different activities, because confusing them is what causes people to file first and search later.
A knockout screen is cheap and fast. Run it during generation on anything the room likes: a search engine check, a domain and social handle lookup, an app store check and one public trademark database search. It takes minutes and it removes obvious dead ends before anyone gets attached.
Full clearance is neither cheap nor fast. It is a comprehensive search by a trademark professional across the relevant Nice classification classes, in the jurisdictions you trade in, plus common law rights checks for unregistered use in your space. That last part surprises people: a business already trading under a name in your category may have rights even without a registration, so an empty registry result is not the same as a clear result.
The defensible rule is simple. Knockout screen everything you like, all the way through. Run full clearance only on the two or three finalists, before you file and before you print anything. If a search turns up a problem, generate again against a different territory rather than trying to rework the same root words into something available.
Check the practical items in the same pass: the domain, every social handle you would plausibly want, the app store listing, and whether the company name can be registered where you are incorporated. Nothing here is legal advice, and for a name you intend to build a business on, get a trademark attorney to review the search result before you commit.
Present and Approve the Final Recommendation
The last job is turning two or three favourites into one documented decision, so the choice survives the person who made it leaving the company.
Score the finalists on a weighted rubric agreed back in the brief, not after people have seen the names. Weight it so that a legal blocker overrides everything, and keep the weights visible. A workable version looks like this.
| Criterion | Weight | What a top score looks like |
|---|---|---|
| Legal and availability | Pass or fail | Clear search in target classes, domain and handles secured |
| Strategic fit | 25% | Matches positioning and personality with no contradiction |
| Memorability and recall | 25% | Recalled after one unaided exposure in testing |
| Usability | 20% | Pronounceable and spellable on first hearing |
| Extensibility | 15% | Works for the second product and second market |
| Consumer reaction | 15% | Correct interpretation in qualitative sessions |
Have the scoring happen anonymously and in writing before the group discusses. Practitioners repeatedly flag falling in love with a name before the trademark search as the most common self-inflicted failure, and a show-of-hands meeting is exactly the environment where that happens. If the winner is the name someone proposed in week one, that is a reason to slow down, not a reason to skip the check.
Then write the recommendation up: the strategic rationale, the evidence from testing, the risks, the pronunciation guidance in plain phonetic form, and the implementation list. Register the trademark, buy the domains and secure the handles. Give everyone who will say the name out loud the same pronunciation instructions.
Common Mistakes
Almost every failed naming project fails the same way, and the fixes are cheap.
Naming before the strategy is settled. If you cannot state positioning in one sentence, you have no criteria to score a name against. Fix: the brief comes first, signed off by whoever signs the budget.
Falling for a favourite before screening. It is human and it is the most expensive habit in this work. Fix: run the knockout screen on anything the room reacts to, immediately, and agree kill criteria before generation rather than after.
Checking availability at the end. Teams discover conflicts after the logo is designed and the domain is printed. Fix: knockout screen continuously, full clearance on finalists only, before any spend on identity.
Assuming an empty domain search means the name is free. Domains and trademarks live in different systems, and unregistered use in your category can still create problems. Fix: treat the domain and handle search as a practicality check, never as legal clearance.
Going too descriptive. A name that spells out the product explains itself once and then limits you forever. Fix: pick the most descriptive name in your territories, not the least, and only if the category rewards plainness.
Generating from one idea. A single territory produces variations of one thought, and the shortlist ends up thin. Fix: three to five territories, 30 to 50 names each, and a rule that each territory produces at least three longlist entries or gets dropped.
Two habits pay off across every project. Screen phonetically as you screen legally — say each name out loud, in a phone call, to someone who has not seen it. And keep the anonymous pre-scoring in place, because it is the cheapest protection you have against your own enthusiasm.
Frequently Asked Questions
How long should a brand naming process take?
A team working from a written brief can run the sequence in two to four weeks: brief, research, one generation workshop, screening and a knockout pass. Add consumer testing and full trademark clearance and you are looking at six to ten weeks. Agency-led projects take longer because legal clearance depends on how much conflict the search surfaces. If you have three weeks, drop consumer testing rather than legal screening.
Should a brand name be tested with customers?
Yes, when the name is expensive to change later or when you are choosing between close finalists. Test how people interpret the name and what they associate with it, not which one they like best, since likeability is close to random without context. Conversation sessions reveal misinterpretations that a survey hides. Always run testing after legal screening, so you never measure reaction to a name you cannot register.
What is the difference between a brand name and a product name?
A brand name identifies the company and is meant to last as the portfolio grows. A product name identifies one offering inside that portfolio and usually carries a function or benefit signal. The distinction matters because the brand name should be abstract enough to extend, while the product name can be descriptive. Run the naming process separately for each, with the product name designed to sit under the parent brand.
When should trademark checks happen in the naming process?
Continuously, but at two different levels. Run a cheap knockout screen during generation — a search engine check, a domain and handle lookup, an app store check and one public trademark database — on anything the team reacts to. Then run full clearance on the two or three finalists only, through a trademark professional, before you file or commit design spend. Checking once at the end is what produces costly surprises.
How many brand name ideas should a team generate?
Between 100 and 300 in a first round, across three to five naming territories, with 30 to 50 names per territory. That narrows to a longlist of 20 to 30, then a shortlist of 5 to 8, then two or three finalists. The volume matters more than the polish early on, because variety beats quality at generation time and quality is what the screening phases add. Fewer than 50 usually means the territories were too narrow.
Can a brand name be changed later if the business evolves?
Technically yes, practically it is expensive. A rename means new domains, new trademark filings, packaging and signage, and every piece of search equity and word-of-mouth you built so far starts again. Some renames are forced — a market entry where the name reads badly, or a business expanding past what the name implies — but those are exceptions. Choose for extensibility in the brief so the question never comes up.
Conclusion
A naming process works because each phase produces something the next one needs: a brief, territories, a longlist, a shortlist, a cleared finalist. Skip a step and the argument happens at the end instead of the beginning, where it is cheaper.
Start by writing the naming brief and agreeing the criteria and kill criteria with everyone who has a vote. Once that single page is signed off, the rest is sequence — and it takes weeks rather than the months a stalled naming debate usually costs.
