A startup name checklist worth actually running through

Most naming advice is a list of adjectives: memorable, meaningful, unique. That does not help you choose between three shortlisted names on a Tuesday afternoon. What helps is an ordered set of tests, run in the right sequence, where the cheap tests come first and kill the weak candidates before you spend money on the expensive ones.

Here is the order that works: disqualifiers, then spoken tests, then legal, then domain, then everything else. Run it in that order and you will not waste a trademark search on a name that fails the phone test.

Start with the disqualifiers, not the qualities

You cannot rank names on how good they feel. You can eliminate names that break something. Do that first, because it is fast and it is objective.

• Does it collide with a well-known company? Not just in your sector. If a name is one letter from a household brand, you will spend years being mistaken for them.
• Does it contain a word that changes meaning across a border? If you have any intention of selling outside your home market, run each candidate past a native speaker of your two most likely expansion languages. This takes ten minutes and has saved a lot of rebrands.
• Does it lock you into a product you might not sell in three years? "Wellness" in the name is fine until you pivot to logistics. Narrow descriptive names are cheap now and costly later.
• Is it hard to say out loud with confidence? If a co-founder hesitates before pronouncing it, so will a customer.
• Does it look like something else in lowercase, in a URL, with no spaces? Compound names occasionally produce an unfortunate second reading. Write every candidate out in lowercase with no spaces and read it cold.

Anything that survives that list is a real candidate. Usually you go in with eight and come out with three.

Test it as speech, not as text

Names get typed less often than founders assume and spoken more often. The tests that matter are auditory.

The phone test. Say the name to someone who has never heard it and ask them to write it down. If they get the spelling wrong, you have not necessarily failed — but you have accepted a permanent tax. Every mis-typed URL is a lost visitor. Names with silent letters, doubled consonants, or a choice between -er and -r endings will fail this consistently.

The radio test. Say "we're at [name] dot com" out loud. If you find yourself adding "that's [name], spelled...", the name costs you three extra seconds every time it is spoken. Over the life of a company that adds up to a real drag on word of mouth.

The voice assistant test. Say the name into a phone dictation field. If the transcription is wrong, the name will be wrong in a lot of other places too.

The noisy room test. Names built from soft consonants (f, s, th) blur in a bar or a conference hall. Hard consonants (k, t, p, b, g) carry.

Do the legal check yourself before you pay anyone

You will eventually want a solicitor or trademark attorney. Before that, an hour of your own searching will remove the obviously doomed candidates.

• Search the trademark registers. The UK IPO, EUIPO and USPTO all have free public search tools. Search your exact name and near-identical variants, in the classes that cover your product. A mark in an unrelated class is usually survivable; a mark in your class is usually not.
• Search the companies register in the countries where you will incorporate. Company name availability is a separate question from trademark, and both matter.
• Search plainly on the web, in quotes, with your industry as a second term. You are looking for an existing business using the name unregistered — that can still cause you grief under passing off or common-law rights.
• Check the app stores. Apple and Google both refuse listings that duplicate an existing app name closely.

If a name survives all four, hand it to a professional for a proper clearance search before you print anything. Self-searching finds the obvious problems, not the subtle ones.

Answer the domain question honestly

This is where most shortlists collapse, and where founders make the most expensive mistakes in both directions.

The mistake in one direction is refusing to consider any name whose .com is taken. That eliminates almost every real word and most short coinages. The mistake in the other direction is choosing a name and assuming the domain problem will be solved later with a hyphen, a "get" prefix, or a .io. Those work, but they carry a cost in typo-traffic loss and in perceived seriousness, and the cost lands on you every day.

The useful middle position: decide the domain question before you commit to the name, and price it properly.

• An unregistered .com costs registration fees only. Rare for real words, achievable for invented ones. If you want a coined name that is definitely free, a tool like keyword domain search for brands is built to only surface names that are actually registrable — the Name Studio checks each invention against the live registry before showing it.
• A registered but unused .com may be for sale. Prices range from a few hundred to well into six figures depending on length, dictionary status and how many other people want it.
• A .com in active use by a functioning business is effectively not for sale. Move on.

If you are shopping the aftermarket, browse before you fall in love with anything. Working from a list of domains that are already priced and available reverses the usual order in a helpful way: you pick from what exists rather than searching for something that does not. On Names.com the transfer runs through escrow, so neither side has to move first, and buyers pay no fee — the 15% commission comes out of the seller's side on completion.

One unglamorous note: leasing a strong domain is a real option if the purchase price is out of range this year. It is not free, and you do not own the asset, but it beats launching on a name you will replace.

Clear the rest of the namespace

Once the domain is settled, spend twenty minutes on the surrounding real estate. You will not get an exact match everywhere and that is fine, but you want consistency.

• Handles on the two or three platforms you will actually use. A consistent prefix or suffix across all of them beats an exact match on one and chaos on the rest.
• A usable email address. If the name is long, check that [email protected] does not run to thirty characters.
• The obvious typo variants of your domain. You do not need to buy them all on day one, but know what they are and whether anyone already owns them.
• Search results for the bare name. If page one is dominated by something unrelated and stronger than you, you will be fighting for your own name for years.

The last three checks before you commit

Write it in a sentence you will actually say. "I'm the founder of ___." "We use ___ for that." "___ raised a seed round." If any of those feels awkward in your mouth, it will feel awkward in everyone else's.

Sit with it for 72 hours. Novelty wears off. A name that felt clever on day one and still feels fine on day three is probably fine. A name that felt clever on day one and slightly embarrassing on day three is a bullet dodged.

Ask whether you would say it to a stranger without wincing. That is the whole test. If you would hesitate to say it at a dinner party, your sales team will hesitate too.

Then stop. Naming has a strong pull toward infinite deliberation, and the marginal value of the eleventh shortlist round is close to zero. A good-enough name that ships beats a perfect one that arrives after your competitor.