How to Name a Business: A Founder's Playbook

To name a business, run a repeatable process instead of waiting for a flash of inspiration: define your positioning, brainstorm 40 to 50 candidates, shortlist against fixed criteria, test the finalists out loud, then clear the domain and the trademark before you commit. The name you keep should be brandable and available on both the .com and the trademark register — not a compromise you accepted because everything good looked taken.

Most founders do this backwards. They fall in love with a word first, then discover the domain costs five figures or a competitor already owns the mark. This playbook flips the order so you end at a name you can actually own, defend, and grow into.

Start with positioning, not wordplay

A name is a container for a promise, so decide the promise first. Write one sentence: who you serve, what you do, and why it is different. That single line tells you which type of name fits.

• Descriptive (e.g. "General Motors"): instantly clear, but hard to trademark and easy for rivals to copy.
• Suggestive (e.g. "Salesforce"): hints at the benefit without spelling it out — the sweet spot for most startups.
• Abstract or coined (e.g. "Kodak," "Sonos"): a blank canvas that is easy to protect but costs more marketing to fill with meaning.
• Founder or place: personal and credible, but harder to sell or scale beyond you.

Pick the lane before you generate words. A suggestive or coined name is almost always easier to clear legally and to own online, which matters enormously once you reach the domain and trademark stages below.

A repeatable way to name a business: brainstorm wide

The first step in how to name a business is volume. You cannot judge and create at the same time, so separate them. Generate far more candidates than you need — aim for 40 to 50 — before you eliminate a single one.

1. Root words: list 20 nouns, verbs, and feelings tied to your promise. Include the problem you remove, not just the product you sell.
2. Combine and bend: merge two roots (Face + book), add a suffix (-ly, -ify, -io), or borrow from Latin, Greek, or another language for a fresh sound.
3. Metaphors: what in nature, mythology, or everyday life behaves like your product? Amazon and Nike came from this.
4. Sound first: say nonsense words aloud until a rhythm feels right, then reverse-engineer a meaning.

AI name generators are useful fuel here, not a verdict — they spit out dozens of starting points fast, but they do not check whether a name is legally clear or emotionally right. Treat every suggestion as a raw candidate to test, never a finished decision.

Shortlist against fixed criteria

Now switch to judging. Score each candidate against the same checklist so you compare on merit, not on which word you happened to think of last night. A strong name usually passes all of these:

• Say-able: a stranger can pronounce it correctly on the first try.
• Spell-able: they can spell it after hearing it once, with no "is that with a K or a C?"
• Short: ideally one or two syllables; long names get abbreviated by customers anyway.
• No baggage: no unfortunate meaning in the languages of your target markets, and no ugly acronym.
• Extensible: it still fits if you add products or expand beyond your first city or niche.
• Distinctive: it does not blur into three competitors with near-identical names.

Cut anything that fails two or more. You should land on five to eight finalists — few enough to test properly, varied enough to have a real choice.

Test the finalists before you fall in love

A name that looks great on a slide can fail in the real world. Put each finalist through three quick tests.

• The radio test: say "Welcome to [Name], how can I help you?" out loud. If it is a mouthful or gets misheard, it will cost you every phone call and podcast mention.
• The spelling test: text the name to five people and ask them to type it back. Watch where they hesitate — that friction becomes lost traffic.
• The reaction test: show it to a handful of people in your target audience, not friends who want to please you. Ask what the company does. If they guess right, or intrigued, you are close.

Resist the urge to poll everyone. Wide surveys reward the blandest, most familiar option — the opposite of a distinctive brand. Use small, honest tests to catch real problems, then trust your positioning to break the tie.

How to name a business people can actually find: clear the domain

A name you cannot own online is not a name yet. Before you commit, check the exact-match .com. Google's John Mueller has confirmed the search engine treats all generic top-level domains equally for ranking, so .com wins on trust and click-through, not on algorithm favoritism — people still type and expect .com by default.

Fresh registrations run roughly $10 to $12 a year at at-cost registrars, but memorable one-word .coms are usually already taken. That is normal, not a dead end. The domain aftermarket exists precisely for this: the average reported sale sits around $2,000 to $3,000, and .com accounts for roughly 80% of all sales volume. A premium, brandable .com is an asset, not an expense — it can outlast three website redesigns and every marketing campaign you ever run.

Two rules protect you when buying. First, use an escrow service such as Escrow.com (fees run about 0.9% to 3.25%) for anything over $1,000 — never wire a stranger directly. Second, treat automated appraisals as rough signals only; check recent comparable sales before you name a price. A curated marketplace like Atom pairs each listing with an appraisal and a ready-made logo, so you can see a name fully formed and buy it through a secured transaction rather than gambling on a cold DM.

Clear the trademark before you commit

Domain availability is not legal clearance. A name can be free to register and still belong to someone else as a trademark, which means a cease-and-desist letter the day you gain traction.

1. Knockout search: search your finalists in the official trademark database for your country (the USPTO's system in the US) plus a plain web and social search. Rule out obvious conflicts fast.
2. Check the class: trademarks are registered by category. A conflict usually only matters if another business uses a similar name for similar goods or services — the "likelihood of confusion" standard.
3. Mind common law: in some countries a business earns rights simply by using a name, even without registration, so a clean database is not the whole story.
4. Get a professional filing: once one name survives, have an attorney run a full search and file. It is far cheaper than rebranding after launch.

Do these two clearances — domain and trademark — together, before you print anything. The perfect name that fails either check is not perfect; it is a liability you have not been billed for yet.

When the best name is taken, buy — don't settle

The founder who ends up with a weak name almost always did so at this exact moment: the ideal name was unavailable, so they bolted on a spare syllable, a hyphen, or a quirky new-gTLD and called it done. Those compromises quietly tax you forever in mistyped URLs and lost trust.

The better move is to decide what the right name is worth. A great .com is a one-time purchase that pays back across the entire life of the company. If your top choice is on the aftermarket, price it honestly against comparable sales, use escrow, and buy it. If you would rather start from a shortlist that is already clear and brandable, a curated gateway like DominantBrand helps you find, appraise, and acquire a premium domain — with the purchase handled safely through Atom's marketplace, logo included. Either way, you end at a name you own outright, not one you talked yourself into.

Frequently asked questions

How long should a business name be?

Shorter is almost always stronger. Aim for one or two syllables and a name people can spell after hearing it once. Long names get abbreviated by customers anyway, so you lose control of the brand. If it fails the radio test or the text-it-back spelling test, it is too long or too awkward.

Do I really need the .com?

For a serious business, yes. Google ranks all generic TLDs equally, so .com is not about SEO — it is about trust and click-through. People default to typing .com, and a .net or new-gTLD version quietly sends traffic to whoever owns the .com. If the exact .com is taken, buying it on the aftermarket is usually worth it.

Should I check the trademark or the domain first?

Check both together, before committing to anything. A name can have a free domain but an existing trademark, or a clean trademark but a five-figure domain. Run a knockout trademark search and a domain check side by side on your finalists, then only pursue a name that passes both.

What if the name I want is already registered as a domain?

That is normal for good one-word names and not a dead end. Most were registered years ago and many are for sale. Check the aftermarket, compare recent sales for similar names, and buy through an escrow service for anything over $1,000. Never wire money directly to a private seller.

Are AI business name generators worth using?

They are excellent for volume — they produce dozens of starting points in seconds. But they do not verify trademark conflicts, domain availability, or how a name feels in your market. Use them to fuel the brainstorm stage, then run every candidate through your own shortlist, testing, and clearance steps.