Choosing the first domain for a new business

A first domain is a single payment and a decade-long commitment, which is an odd combination and probably explains why founders handle it so badly. It matters less than they fear and more than they act. No domain has ever rescued a weak business. But changing one later means reprinting everything, rebuilding email, resetting search rankings and re-teaching every customer who already knows where to find you. Get it roughly right at the start and the question never comes back.

What follows takes an afternoon. Not a month.

Work out what the name has to do before you fall for a word

Businesses want contradictory things from a domain, and the conflicts are what trip people up. Read this and be honest about which lines actually describe you.

• People will hear it before they see it. Word of mouth, podcasts, radio, someone shouting it across a trade show floor. That rules out clever spellings and anything with a hyphen in it.
• People will find you by searching a category. A local firm competing on "emergency plumber Leeds" gains from plain clarity in a way a venture-backed platform does not.
• You expect to broaden later. A name that describes today's product becomes a cage the moment you sell something else. Plenty of companies have outgrown their own domain.
• You are raising money or selling to enterprises. A weak domain reads as a weak company to people who have twenty minutes to form an impression and no reason to be generous.

First-time founders reliably overrate cleverness and underrate that first point. If you cannot say the name once, over background noise, and have a stranger type it correctly, it is losing you traffic every day you trade.

Five checks, and the good names survive all of them

Get to a shortlist of five to ten, then push every one through the same gauntlet. Candidates die fast here. That is the whole function.

Say it to someone who has never heard it

Do they spell it right first time? Doubled letters, silent letters, dropped vowels and numbers all fail this, repeatedly. "Flickr" worked because a company spent real money teaching the public how it was written. That budget is not available to you.

Fifteen characters, and a phone keyboard

Under about 15 characters is comfortable. Two syllables beat four. Type it on a phone a few times before you commit, because anything awkward under the thumb becomes an email address people get wrong forever.

Read it as one unbroken string

Joining words together produces readings you did not intend. Look for the other word hiding in the middle of yours. Five seconds of work; a lot of saved embarrassment.

Search the trademark register before you spend anything

Check your national register, and the relevant international ones, for anything close in your class of goods. Owning a domain does not give you the right to trade under the name. If a candidate is genuinely in the running, a short conversation with a solicitor costs a fraction of a forced rebrand.

Find out what the domain did before you got there

Search the exact name in quotes. Look into whether the domain has been used before, and for what. Something previously spammed or abandoned can carry baggage. Rarely fatal, always worth knowing first.

Take the .com if the money is there, and be straight with yourself if it isn't

In most people's heads .com is still the default. When someone half-remembers your name, .com is what their fingers produce. Take .io or .co while an active business sits on the matching .com and you are permanently handing some slice of your direct traffic to a stranger.

Now the unglamorous half. For many small businesses that leak costs less than chasing it would. A local trade, a single-site shop, a consultancy that gets work by referral: none of these is losing meaningful revenue to a mistyped extension. Spending a five-figure sum on a .com when your whole launch budget is smaller than that is simply a bad trade.

A hierarchy that holds up:

• An exact-match .com within reach. Best outcome available. Take it.
• A slightly different .com. A short prefix or suffix, or an invented word, usually beats an alternative extension on your first-choice word. Invented .coms are the most underrated option going, and the brand keyword search exists for precisely this: feed in the words that matter and get back combinations that are actually available.
• A strong country extension. Serving one market, .co.uk, .de or .com.au reads as native rather than as consolation.
• A recognised alternative. .io, .ai, .co and .studio are understood well enough in their own contexts. Accept that a small part of every conversation goes on clarifying the ending.

No hyphens, no numbers. Both fail the spoken test.

What the market actually charges

There are no fixed rules in domain pricing, but the shape of it is consistent enough to plan around.

• Unregistered .com names cost a standard registration fee at any registrar. Almost always invented, or several words long.
• Owned but unremarkable domains, the long and obscure ones nobody is competing for, tend to change hands in the low hundreds to low thousands.
• Short, pronounceable, brandable .coms typically sit from the mid four figures to the low five figures.
• One-word dictionary .coms start in the five figures and climb well past it. Genuine scarcity, priced accordingly.

The figure to keep in mind is what a rebrand costs you in two years: signage, packaging, email, lost search equity, and the confusion tax levied on every customer you already have. Against that, paying once for the right name is often cheaper. Against a business that may not exist by autumn, it isn't. Decide which of those you are running.

Working to a fixed budget, browsing a small business domain category beats open-ended searching, because you are looking at priced, available inventory instead of a list of names you cannot have.

Buying a name someone already owns

Most good names are taken. Taken is not the same as unavailable; it means the price isn't zero.

Two routes exist. A marketplace lists the name with a price or invites offers, the transfer runs through escrow, and nobody has to move first. On Names.com the buyer pays no fee and the seller pays 15% on a completed sale. Or you approach an owner cold, which is slower, has no fixed process and usually ends in silence, though it occasionally turns up something nobody thought to list. The owner-direct listings sit between those two.

Either way, use escrow. A transfer takes time, and there is no moment in it where trust is naturally safe. Escrow deletes the question.

The point where naming becomes procrastination

Naming is a very comfortable way of not launching. Three weeks in, you are well past diminishing returns and you probably know it.

The test for good enough: spellable when heard, on a .com or a strong country extension, clear of anyone's trademark, not embarrassing in five years, and payable without endangering the business. Clear all five and buy it. Then go and build the thing that makes the name worth something, because the causation runs that way and not the other.