2026年、ブランドに名前をつける方法 — クリシェの罠を避けて

A bad name is the cheapest mistake a founder can lock themselves into for a decade. The naming work itself is rarely expensive — three to six weeks of strategist time, a five-figure legal clearance budget, some native-speaker testing. The actual cost shows up later. Every time a sales rep has to spell the name on a call. Every time customer support gets a Google search for a competitor with a similar word. Every time the marketing team has to spend extra media dollars to disambiguate the brand from a near-homonym in a different category. At Focus Point we've named over 40 brands across luxury, DTC, B2B SaaS, hospitality and creative industries. The pattern is identical every time. The internal team falls in love with a candidate within the first two weeks. They build emotional commitment to it. They then learn — six weeks in — that the legal team will veto it. They argue. They lose. They start again, demoralised. This article is the order of operations we wish every founder ran before they started.

Step one: run trademark availability before you fall in love

Before you let yourself love a candidate, run it through EUIPO (European Union), USPTO (United States), WIPO (international), and the national registries of your top three target markets. We use a simple rule: if the trademark class you operate in shows even a partial match in any of those, the name is dead. Not 'worth fighting'. Not 'maybe with some adjustments'. Dead. We've watched founders waste 4 months — and €60,000 of agency fees — because they wanted to 'fight for' a name that had a partial mark in Class 35 in Germany. They lost. The German company had been operating quietly since 2009. The founder's emotional commitment to the name had nothing to do with the legal reality.

Step two: the three naming frameworks that actually work

1. Founder-driven (Tesla, Hermès, Dior, Bezos's Amazon-as-personal-vision). Only works if the founder is the story. If you cannot make the founder photograph well, this framework is closed to you.
2. Coined word (Spotify, Verizon, Kodak). Strongest possible IP. Single most defensible in trademark court. But it requires a real marketing budget to build meaning — coined words mean nothing until you teach them.
3. Compound (Salesforce, Facebook, Snowflake). Easiest to land internally because everyone understands it on first hearing. Hardest to register because the components are usually common-word. Half the legal effort is making the compound distinctive enough to qualify.

There is a fourth shape that gets pitched a lot — the metaphorical name (Apple, Amazon, Patagonia). We do not list it as a framework because it cannot be replicated by deliberation. It works when the metaphor is so unexpected that it becomes its own meaning. It fails when it's a tired Latin or Greek root dressed as poetry. If you are choosing between a brave metaphor and a sturdy coined word, take the coined word. The metaphor will look obvious in five years; the coined word will look like yours.

Step three: the linguistic check that catches expensive mistakes

Names look one way written and another way spoken. They mean one thing in your home market and another thing in your launch market. The mistakes are catalogued — the Chevrolet Nova that supposedly didn't sell in Spanish-speaking markets (the story is partly apocryphal but the cautionary tail is real), the Honda Fitta that had to be renamed Jazz in Sweden, the Vicks brand that had to become Wick in Germany. We test every shortlist candidate spoken aloud, in the top three target languages, with three native speakers per language. We are not looking for perfection. We are looking for the disqualifying laugh — the moment a native speaker laughs uncomfortably and says 'no, you can't use that here'. That is a real veto. Half of strong-on-paper candidates do not survive this test.

Step four: the soundability test most teams skip

A name has to survive being spoken on a phone call, shouted across a noisy room, and said three times in a row by a customer support agent without becoming a tongue-twister. We test this with a simple exercise. We ask three internal team members to repeat each shortlisted name ten times in a row, fast. Names with hard consonant clusters fail. Names with vowel-on-vowel collisions fail. Names with three syllables of identical stress fail. The winners are usually 2 or 3 syllables, end in a vowel or a soft consonant, and have a single stressed syllable. This sounds trivial. It is not. The names that fail this test cost more in customer support training over five years than the entire naming sprint cost.

The fifty pitfalls — in a single condensed list

• '-ly' suffixes (Datedly, Quickly, Easily). The category exhausted itself in 2015. Investors now read them as derivative.
• Generic word + adjective (Pure Energy, Bright Capital, Smart Health). They sound like every other brand and trademark badly.
• Latin invented words (Synergia, Vivantia, Optimus). They signal 'we hired a consultancy' rather than 'we built something'.
• Names with silent letters or non-obvious spellings (Aevum, Kynd, Qixx). Customer support nightmare, SEO nightmare, voice-search nightmare.
• Names that contain the founder's first name unless the founder is the story (rarely is, in 2026).
• Names borrowed from mythology that have already been used twice in your category.
• Names that require an accent in your home language but won't render in URLs or English-language press.

Step five: the budget and the team

A naming sprint that does the work properly costs between €25,000 and €60,000 of agency fees, depending on the markets you're clearing and the depth of strategic upstream. The legal clearance budget is separate — typically €5,000 to €15,000 for the standard global cleanup in three to five markets. The domain acquisition is separate again and unpredictable — it can be €500 or €50,000 depending on who owns the .com. Build the full envelope before you start. If the available envelope is only €15,000 total, you cannot do this work properly and you should either delay the project or commit to a single-market launch with a much narrower clearance footprint. We've seen too many founders spend €15,000 and end up with a name they have to abandon at year two.

The team: founder, head of legal, one external strategist

Naming dies in committees. We have a hard rule about who is in the room when the final shortlist is reviewed: the founder (or CEO), the head of legal (or external counsel), and one external strategist. That's it. Three voices. No committee of seven. No 'let's get input from the broader marketing team'. The reason is mathematical: every additional voice in the room multiplies the surface area of consensus required, and naming is a domain where the lowest-energy candidate always wins by attrition. The names that survive committees are forgettable. The names that survive the three-voice room are sharp.

What to do next

If you're about to start a naming project, do three things this week. One: list your three target markets and brief your IP counsel on a clearance check for any name you bring to the table. Two: write a one-page brief that defines the strategic territory the name has to live in — audience, positioning, category code, and one sentence about what you absolutely do not want it to feel like. Three: book a 45-minute call with a brand strategist (us, or someone else — but with someone who has named multiple brands). That call will save you the cost of a wrong direction. We do these for free, no pitch attached. Use the contact page or the quote form, and we'll get back within a working day.