WIPO’s World Intellectual Property Indicators 2025 counts 93.2 million trademarks in force worldwide in 2024. A new name lands in that pile, and most of the short, obvious candidates are already spoken for.
Naming a brand is mostly elimination. A written brief goes in, and one name comes out that the company can own, people can say out loud and a named owner has signed off on.
That is why trademark clearance starts while the shortlist is still long, not after someone falls for a favorite.
What Are the Stages of the Brand Naming Process?

Agency guides count the stages differently, from 5 steps at Lexicon Branding to 9 at Motto, but the order barely moves between them. Each stage hands the next one something concrete, and most of them cut names from the pool.
It starts with a written brief covering criteria, tone, exclusions and who gets to decide. Research follows, usually a competitor name audit plus a few naming territories to explore. Generation then produces a long working list sorted by concept, and scoring trims that to a ranked shortlist against the brief’s criteria.
Trademark clearance comes next, a prescreen first and a written legal opinion on the last few names after that. Domain, social handle and language checks for the target markets sit alongside it. Testing puts the cleared names in front of stakeholders and customers, and the whole thing ends with one signed-off name and an application to register it.
Naming sits inside the wider work of building a brand identity. Motto’s 2026 guide starts naming only after brand strategy and positioning are fixed, and moves to visual identity once a name is chosen.
The front end decides how painful the back end gets. Inkbot Design argues that firms that set decision rights before any name exists avoid re-arguing taste at every stage.
How to Write a Naming Brief

Before a single name exists, the brief fixes what is being named, what the name has to communicate and how candidates will be judged. It is stage one, and every later score points back to it. Most teams rush it, which is a mistake.
Motto’s brief template is long, and only some of it changes outcomes. Say exactly what is being named (company, product or sub-brand). List the territories the name should point toward and the ones it should stay away from. Pick a tone: literal or symbolic, playful or serious, familiar or unexpected. Describe the audience and the competitors whose names yours must not resemble. Then write down the hard requirements, such as domain needs, trademark screening rules and languages to clear.
Positioning sits upstream of all of it. A brief is only as sharp as how the brand is positioned, so a missing positioning statement is a reason to stop and write one first.
Criteria get fixed here as well. Published guides converge on distinctiveness, pronunciation, memorability, fit with the brand’s meaning and room to grow.
The brief sets the weights. An app sold in several countries ranks pronunciation and cross-language safety above cleverness, while a local service business can weight them lower.
Decision rights belong in the brief too. Name who votes, who holds a veto and who breaks ties. Inkbot Design recommends one named owner who applies the rubric and records the audit trail, with gut feel kept as a tiebreaker only.
Taste never leaves a naming project. The brief just gives people something other than taste to point at when they disagree.
Which Type of Brand Name Fits the Business?

The choice among descriptive, suggestive, arbitrary and invented names mostly trades how easily a name explains itself against how well the law protects it. The more a name says about the product, the less of it a company can own.
| Name type | Example | Registrability | Main risk |
|---|---|---|---|
| Descriptive | Vision Center (eyewear) | Only after acquired distinctiveness | Hard to own, easy to copy |
| Suggestive | Coppertone (sunscreen) | Inherently distinctive | Examiner may read it as descriptive |
| Arbitrary | Apple (computers) | Inherently distinctive | Meaning must be built through marketing |
| Invented | Xerox (copiers) | Inherently distinctive, broadest protection | Carries no meaning at launch |
The sort order comes from the Abercrombie spectrum, named for Abercrombie & Fitch Co. v. Hunting World, Inc., which places generic terms at the bottom as words that can never work as a trademark. The Federal Circuit restated the working rule in Real Foods Pty Ltd. v. Frito-Lay North America (2018): suggestive, arbitrary and fanciful terms are inherently distinctive, and descriptive ones are not.
Construction types cut across the table. Founder names (Ford, Disney), acronyms (IBM) and compound names (Microsoft, Mailchimp) describe how a name is built, not how strong it is, and each lands on the spectrum wherever its meaning puts it.
A founder name ties the brand to one person. An acronym shortens a longer phrase and means little until the company is established. Compounds join two words to hint at the product.
Apple drew skepticism at first and became iconic because it was distinctive, according to Motto’s 2026 guide.
Match the type to what the launch budget can teach. A small budget in a crowded category favors a suggestive name, while a company funding awareness across several markets can carry an arbitrary or invented one.
Descriptive names save money on explaining. They cost more on protection.
How to Generate and Score Name Candidates

Generate a long list without editing, score every name against the criteria fixed in the brief, and send only the survivors to legal screening. The strongest names rarely show up early, so volume comes first.
Generating names
Mixed workshops produce a lot of the volume, with namers, strategists and the client’s own team all contributing, which is how Motto runs it. Word sources add more: dictionaries, thesauruses, mythology, geography, science and other languages. Koto has a trick for dodging category clichés, which is to name the brand as if it were a restaurant, a band or a fashion line. Keyword-based name generators add raw volume too, though a person still has to judge what comes out of them.
Keep the list sorted by concept, not alphabetically, so patterns show up. Do not cut during generation, because a weak idea is sometimes the bridge to a strong one (Motto, 2026).
How the list narrows
Motto’s 2026 guide puts a full generation round above 200 ideas, cut to 1 to 3 names by the time a full legal search starts. Two other published funnels have the same shape.
Bullhorn Creative starts at 200 or more names, drops to about 50, then about 25, then 10 to 15 after small-group testing, and sends the top 2 or 3 to clearance. The UCL IBIL pharmaceutical naming presentation (2021) runs 200+ names, then 50 to 70, then about 15, then 4 to 6 names including backups.
All three start from 200 or more names and end with a single-digit set for full clearance. Budget the legal work for that set, not for the long list.
Scoring against criteria
Score each name 1 to 5 on distinctiveness, pronunciation, memorability, fit with the brand’s meaning and room to grow. Weight those from the brief, then rank. Gut feel settles ties only (Inkbot Design).
Alexandra Watkins’ SMILE and SCRATCH test (Hello, My Name Is Awesome, second edition, 2019) works as a second filter. A good name is Suggestive, Memorable, uses Imagery, has Legs and is Emotional.
SCRATCH covers the seven faults: Spelling-challenged, Copycat, Restrictive, Annoying, Tame, Curse of Knowledge and Hard to pronounce. The book cites Xobni, Svbtle and Doostang as names that look like typos.
How to Run a Trademark Clearance on a Shortlisted Name
Clearance runs in two tiers, a quick knockout prescreen on the whole shortlist and then a full attorney search on the last one to three names. The prescreen exists so nobody falls for a name that cannot be owned.
The prescreen is a fast search of the registers for identical and similar marks in the relevant classes. Motto sorts results into Out (high risk), Maybe (medium) and In (low), and only In and Maybe names move on. The full search is attorney-run and ends in a written registrability opinion for each name.
Guides disagree on when to start. Inkbot Design treats clearance as a live filter from the first shortlist, not a final checkbox, and the USPTO timing data supports that.
The figures below come from the first half of fiscal 2026 unless noted. First-action pendency is 4.45 months, down from 5.6 months (USPTO, May 2026). Total disposal pendency is 10.03 months, down from 11.7 months in fiscal 2025 (USPTO, May 2026). The base application fee is $350 per class, in effect since January 18, 2025 (USPTO).
Agency guides, Motto’s included, quote roughly 10 months for USPTO review. That is the total disposal figure, not the wait for a first examiner response, which is where a conflict first shows up.
Most US work happens in USPTO Trademark Search, which covers every US application and registration. Beyond that, EUIPO TMview runs a single search across national and regional offices (the USPTO and WIPO included), and the WIPO Global Brand Database holds international registrations filed through the Madrid System.
Many guides still point to TESS. The USPTO retired it on November 30, 2023 and replaced it with Trademark Search, so older instructions describe a tool that no longer exists.
Search the classes the company will file in, plus related ones. The Nice Classification sorts goods and services into 45 classes (34 for goods, 11 for services), but confusion is judged by the thirteen du Pont factors, which include how the marks look and sound and how related the goods are.
The factors come from In re E. I. du Pont de Nemours & Co. (CCPA 1973). No single factor decides by rule, since the court held that the weight of each shifts from case to case.
A clean prescreen is not clearance. Motto notes that even names rated low risk carry no guarantee, and no attorney can promise an outcome at the USPTO.
How to Check Domain, Social Handle and Language Availability
Check the domain first, then social handles, then run a language screen with native speakers in each target market. The first two take minutes, so they cut names before the slower language screen has to.
The Domain Name Industry Brief (DNIB.com, sponsored by Verisign) counted 166.6 million registered .com domains at the end of June 2026. A name built from an ordinary word rarely has a clean .com left.
If the .com is taken, adding a prefix or suffix keeps it and lengthens the name (Koto’s first suggestion). A different extension such as .io, .co or .net keeps the name short, but the .com owner collects any visitor who types the habitual ending. Where the .com belongs to an active competitor, treat it as a clearance problem, not a domain problem.
Handles follow the same logic. Check the exact name on every platform the brand will use with a handle checker such as Namechk, because mismatched handles make a brand harder to find.
ICANN oversees the domain name system, and accredited registrars such as Namecheap and GoDaddy sell the registrations. Either registrar’s search box shows availability in seconds.
Language screening is the step no search box can do. Lumen Learning’s marketing course (SUNY, 2016) recommends having native speakers from each operating country review the final shortlist.
Buick shows what a screen can and cannot settle. GM sold the LaCrosse in Canada as the Allure from 2005 because “la crosse” is Quebec French slang, then returned to the LaCrosse name with the 2010 redesign after Quebec dealers asked for it, according to CBC News as republished by The Auto Channel (2009).
A screen flags the risk. Someone with authority still has to price it.
How to Test Names With Stakeholders and Customers
Only names that have cleared legal screening should get tested, and the order matters: a stakeholder vote against the rubric, then a small-group test, then a consumer survey. Each answers a different question, so skipping one leaves that question open.
For the stakeholder vote, every voter scores the names against the criteria from the naming brief, alone, before any discussion. Lexicon Branding makes avoiding groupthink its first naming step, and independent scoring is the simplest guard against it.
The small-group test puts a handful of people from the target audience in front of the names in context. Koto points out that a name never appears as a bare word on a page, so show each finalist on a mood board with imagery and sample copy.
The consumer survey is quantitative and needs respondents who match the buyer profile, not colleagues who already know the backstory, as SurveyMonkey’s name-testing guidance stresses.
Focus groups add depth. SurveyMonkey notes they are expensive and rarely produce statistically significant input, so treat them as the small-group step, not as proof.
Vendors split on survey design. SurveyMonkey shows each respondent a single name (a monadic test), which removes the order bias that creeps in when people compare a list.
Lab42’s case study for a national fast food chain went the other way, with a sequential design. 600 respondents each rated all 8 names, the current one included.
The current name won overall, which Lab42 tentatively tied to the equity it already carried. That is a familiarity effect, and any name change has to overcome it.
Test results inform the decision. The named owner from the brief still makes it.
How Long Does Brand Naming Take and What Does It Cost?
Naming runs weeks, not days, and the price ranges from a few thousand dollars for a micro package to six figures for enterprise work. The route you pick sets both.
| Route | Price | Best fit |
|---|---|---|
| In-house | Staff time plus attorney fees | One market, low trademark risk |
| Micro package | $6,000 (Eat My Words) | Teams of 3 or fewer, outside crowded trademark classes |
| Specialist firm | $17,500 to about $50,000 (Eat My Words) | Several markets or a crowded class |
| Enterprise project | $50,000 to $199,999 (Clutch review band) | High-stakes, multi-market names |
Clutch review data backs up the middle row. For two naming firms, Tanj and NameStormers, the most common project size sits in the $10,000 to $49,999 band.
On time, Catchword Branding puts a full naming process at up to 6 weeks and tells DIY teams to allow at least 4 to 6. Assembling the team and writing the brief take a visible bite out of the start.
Two clocks sit outside that calendar. The USPTO wait after filing (covered in the clearance section) starts only once a name is chosen, and stakeholder review time stretches every stage.
Clay’s 2026 timeline guide, written about full branding projects, blames review latency for the gap between working time and calendar time.
Attorney and filing fees sit on top of every route. Motto, for one, bills deep-dive clearance and trademark filings separately from its naming fee.
As a rule, go in-house when the name serves one market, the category is not crowded with trademarks, and one person owns the decision. Hire a specialist when the name must clear several markets or a crowded class, or when stakeholders have stalled before.
Even a specialist’s budget tier has limits. Eat My Words excludes software and other crowded trademark classes from its micro package.
When Does the Brand Naming Process Break Down?

It breaks down when a name is committed to before it is cleared, when a conflict sits outside the registers, and when the category adds a second approval gate. It also stops fitting when the thing being named is already in market.
Committing too early is the classic failure, and Inkbot Design names falling for an early favorite as the cause. Mozilla’s browser shows the cost, renamed from Phoenix on April 14, 2003 after a trademark dispute with Phoenix Technologies.
The replacement, Firebird, collided with a database project that had used the name since 2000, according to LWN.net (2003). Mozilla argued the two products lived in different spaces, dropped Firebird anyway, and shipped as Firefox by February 2004.
A register search covers registered rights only. Search the web, app stores and open source projects for same-name products too, because a community objection can force a rename even when the lawyers disagree.
Drug names get a second approval gate. They pass a trademark office and a health regulator, and the two reviews run independently of each other. Cooley’s drug naming guide puts the road from name creation to final approval at more than three years.
The same stages apply to these names. Clearance is just no longer the last gate.
A name already in market is a different problem. Renaming something people already use is rebranding, which brings difficulties a new-name project never meets. The audience knows the old name, so the new one starts behind.
Mozilla’s users saw the name change twice in under a year.
On the other end, a short-lived campaign name needs a brief, a quick search and a decision, not eight stages.
Brand Naming Process FAQ
How is brand naming different from product naming and company naming?
The steps are nearly identical. What differs is lifespan, since a brand name has to survive expansion and so stays broader than any single product. Product naming labels one product or line, while brand naming covers the whole company and its identity.
Can AI name generators replace the naming process?
No. A generator adds candidates, which covers one stage of the process.
The brief, the trademark clearance and the testing still need people. Smashbrand’s guide treats AI naming tools as time savers that free the namer for research and testing.
Which naming conventions work for sub-brands and product lines?
Set the rules before the first sub-brand exists. Martin Roll’s naming toolkit recommends fixed conventions for variants, extensions and sub-product names.
One example is an umbrella brand whose sub-brands all take alphabetical names, which makes each later name a quick decision.
Does a good brand name have to be short?
No. Motto’s 2026 guide points to Habitat for Humanity, Banana Republic and Victoria’s Secret as names of five or more syllables that work.
Meaning and distinctiveness outweigh length, though a name that is hard to spell or pronounce still fails the SMILE and SCRATCH test.
Is it legal for two companies to use the same brand name?
Yes, when the goods are unrelated enough that confusion is unlikely. The USPTO weighs the thirteen du Pont factors, which include how related the goods are.
Identical names can coexist in distant classes. Similar names selling related goods risk a refusal.
Why shouldn’t you pick the name everyone likes best?
The most liked names are often the safest ones. Motto’s 2026 guide says popularity does not predict effectiveness.
A name can also feel odd at first and grow on people as it becomes familiar. Score candidates against the brief, not against applause.
What to Register First After the Name Is Chosen
Registrations follow a fixed order once a name is chosen. The trademark application goes in first, on decision day, and the domain and social handles are secured the same day. An in-use application follows after launch.
Motto’s 2026 guide favors the intent-to-use route because it holds a place in line while the identity is still being finalized. The in-use filing then completes registration.
The trademark clock is the longest one. USPTO processing runs 7 to 11 times longer than a naming project, so filing on decision day starts it first.
The trade-off is exposure, since a filed application becomes searchable in public trademark databases before launch day.
The USPTO targets nine-month disposal by fiscal 2028 (USPTO, May 2026), which shortens the wait but leaves the order unchanged.
With the filings in, the work moves to visual identity, where the name gets its logo, typography and color.
- The Windsurf Logo History, Colors, Font, And Meaning - 10 October 2026
- How to Remove Background From Product Photos (No Photoshop) - 10 October 2026
- How Designers can Build a Stronger Online Creative Identity - 9 October 2026