Picking a Business Name That Passes the Phone Test

The test that kills most name ideas in four seconds

Say your domain out loud to somebody who has never heard it. Do not spell it. Do not write it down. Just say it once, the way you would on a phone call, and watch what happens.

If they type it correctly, you have a good name. If they hesitate, ask you to repeat it, or start a sentence with “is that with a…”, you have a name that is going to cost you money for as long as your business exists.

That is the phone test, and it is the single most useful filter we know of. It is free, it takes four seconds, and it eliminates more bad name ideas than any research tool. We tag names in our own portfolio by whether they pass it, because it is one of the few things about a domain that is both entirely objective and genuinely predictive of how much grief it will cause.

Why this matters more than it sounds like it should

A name that needs spelling is not a one-time inconvenience. It is a small recurring tax on every interaction your business has, forever.

Every phone call where you have to say “that’s S as in Sam.” Every radio spot where you burn three seconds of paid airtime spelling instead of selling. Every truck door somebody reads at a traffic light and mistypes at home. Every time a happy customer tries to tell a friend and gets it slightly wrong, sending them to a competitor or to nothing at all.

None of these show up in any analytics dashboard. You never see the customer who typed it wrong and gave up. That invisibility is exactly why the cost gets ignored, and it is why people who have actually run a service business tend to care far more about this than people who have only run websites.

The most expensive thing about a hard-to-spell domain is that you never find out how much it cost you.

The six rules

1. No hyphens

Hyphens do not hurt your site technically. They hurt every conversation. Nobody says “hyphen” out loud naturally, half your audience will forget it, and the non-hyphenated version — which somebody else owns — gets the traffic you paid to create.

There is one legitimate exception: when the hyphen is part of how the term is actually written in the real world. A compound term that is conventionally hyphenated reads correctly to the audience that uses it. That is a narrow exception and most hyphens are not it.

2. No numerals

Same problem, worse. Is it “4” or “four”? “2” or “to” or “too”? Every numeral forces a decision on the listener that they have a real chance of getting wrong. If a numeral is genuinely core to the brand, own both spellings and point one at the other.

3. Short enough to say in one breath

Past about four syllables, recall drops off quickly. Long exact-match names have real value in some contexts, but understand the trade you are making: you gain clarity about what you do and you lose memorability. That is often the right trade for a lead-generation site and rarely the right trade for the name going on your building.

4. No ambiguous spellings

Words with two accepted spellings, homophones, deliberately clever misspellings. Every one of them splits your traffic between the version you own and the version you do not. Clever spellings are the most common self-inflicted wound in naming, because they feel distinctive to the person who invented them and feel like a typo to everyone else.

5. Say it out loud before you commit

Not in your head. Out loud, to another person. Words that look fine written down sometimes run together when spoken in ways that are memorable for the wrong reasons. This has happened to enormous companies with enormous budgets. Five seconds of saying it aloud would have caught every one of them.

6. Check the trademark before you register

Search the mark databases before anything else. Registering a name that shadows a live trademark is how people lose the name and the money together, and it is entirely avoidable with a search that costs nothing.

The trade-off nobody explains: descriptive or brandable

Two legitimate strategies, and the mistake is not choosing one — it is choosing without knowing you were choosing.

Descriptive

The name says what you do. A stranger who sees it in a search result or on a van knows what the business is before they engage with anything else. Nothing to explain, ever. The cost is that it can box you in: a name built around one city and one service is a liability when you expand to two cities and four services.

Brandable

No keyword. It works as a name people remember rather than a phrase people search. It travels anywhere as you grow and it can become genuinely distinctive. The cost is that it means nothing until you spend money making it mean something, and you will be explaining it for the first few years.

Neither is wrong. What is wrong is buying a descriptive name locked to one city when you plan to be in six, or buying an invented word when you have no budget to build meaning into it. Decide which trade you are making, deliberately.

What a good name will not do

We sell domain names and we are going to say this anyway, because the alternative is a sale we have to defend in six months.

A keyword in your domain will not rank you. Google removed the ranking benefit of exact-match domains years ago and has restated it since. It has gone further and cautioned publicly that keyword-stuffed domains can read as spam and leave a business without a brand anybody can search for on purpose.

A domain will not get you cited by AI engines either. There is no study showing it does. The research points at content clarity, page speed, third-party mentions and presence on review platforms. One large analysis found that broad descriptive titles and URLs earned roughly twice the citations of heavily keyword-optimized ones — the opposite of the pitch. We wrote the long version with sources.

So what does a good name actually buy you? Four things, all real.

• Clarity before the click. Your domain shows in the search result. A clear name gets chosen more often than one that needs explaining.
• Survival in the physical world. Billboard, truck door, radio, phone call. A name that survives all four is worth money every day.
• Demand that belongs to you. When people remember your name and search for it directly, nobody outbids you for it. That is the closest thing to an owned channel.
• Scarcity. There is exactly one of each. Once you have the plain-language name for your service in your city, the competitor who wanted it cannot have it.

.com, or something else?

The alternative extensions are genuinely better than they used to be and the argument for them is stronger than it was. It is still usually the wrong call for a business selling to the general public, for one stubborn reason: the .com reflex has not gone away.

People type .com without deciding to. They hear a name, they type it, and their fingers add .com. That reflex is strongest in exactly the audiences most small businesses sell to — homeowners, older buyers, anyone who is not in technology. It is weakest among developers and startup people, which is precisely why alternative extensions feel more normal than they are to the people advising you.

The practical cost of not owning the .com is that some meaningful share of your direct traffic goes to whoever does. If that is a parked page, you lost a customer. If it is a competitor, you funded one.

Two honest exceptions. A local country extension where you sell in that country carries real signal, and Google treats country extensions as a geographic signal in a way it does not for generic ones. And a specialized extension aimed at an audience that lives inside a niche can work when that audience genuinely recognizes it.

Notably, none of this is an algorithm argument. Google has said it ignores keywords in top-level domains, and a 2026 analysis of over a million cited URLs found .com, .org, .io and .ai citing at roughly the same rate in AI answers. Choose your extension for humans. The machines are close to indifferent.

Naming a local service business specifically

Most naming advice is written for software companies and then applied, badly, to a roofing business. The constraints are not the same, and three of them matter enormously if your business puts a name on a vehicle.

It has to work spoken, at speed

Your customers hear the name on the radio, from a neighbor, or from you on a call with a bad connection. This is not a screen-first medium. A name optimized for how it looks in a logo and never tested out loud is a name that will underperform in the channel that actually drives your work.

It has to fit on a door

Physically. A twenty-eight-character name on a van door is either tiny or wraps to two lines, and either way it gets read at forty miles an hour by somebody who then has to remember it. Length has a hard physical constraint in the trades that it does not have online.

Geographic names are a trade, not a free win

A city in the name is genuinely powerful. It tells a homeowner you are local before anything else does, and local matters enormously in home services. It also welds you to that city. If you expand to the next metro in three years, you are either running a second brand or explaining why the company named after one city is working in another.

This is a real decision with real costs either way, and the right answer depends entirely on whether you plan to expand. Plenty of excellent businesses have run for decades on a city-locked name because they never intended to leave. Plenty of others wish they had chosen differently.

How to test a name before you commit

Four tests, none of which cost anything, all of which people skip.

The phone test

Covered above, and still the most valuable. Say it to three people who have not heard it. Watch what they type.

The radio test

Say the full name and the extension out loud the way you would in a thirty-second spot. Time it. If it takes four seconds and needs a spelling clarification, that is a meaningful chunk of paid airtime spent on logistics rather than persuasion, every single time the spot runs.

The search test

Search the name. What comes up? If there is an established business with a similar name, even in a different industry, you will be fighting for your own brand searches forever and you may have a trademark issue. Find that out now.

The email test

Write it as an email address and say it aloud. Names that read fine as a domain sometimes become unwieldy or unintentionally funny with a name in front of them and an at-sign in the middle. You will say this address thousands of times.

Once you own it, do these four things

A name nobody protects is a name you can lose, and the loss is usually self-inflicted.

• Turn on auto-renew and check the card on file. The most common way to lose a valuable domain is an expired credit card attached to an account nobody opens. It happens to real businesses constantly and it is entirely preventable.
• Turn on privacy protection and the registrar lock. Privacy keeps your home address out of a public database. The lock prevents an unauthorized transfer. Both are usually free or close to it.
• Use an email address that is not at the domain itself. If your registrar contact email is at the domain in question and something breaks, you have locked yourself out of fixing it. Use a separate address you will still control in five years.
• Put something on it. A registered name pointing at nothing does nothing. Even a single page beats a blank. Our hosting page covers what a name needs behind it, and if you would rather have the whole thing built, that is a conversation too.

Should you buy the near-misses too?

Sometimes, and the rule is narrower than the people selling you domains would like it to be.

Buy the variant when the mistake is one a real person would genuinely make. The plural when your name is singular. The singular when it is plural. A common misspelling of a place name that people outside the area get wrong. Register those, point them at the main name with a redirect, and you have closed a leak for the price of a registration.

Do not buy thirty permutations of one idea. It is the most common way new business owners waste money on domains, it creates an annual renewal bill for assets nobody will ever type, and it does nothing for you. Two or three genuine near-misses is protection. Thirty is a hobby.

There is also a defensive argument that is legitimate in competitive local markets: if the plain-language name for your service in your city is available and cheap, owning it means the competitor down the road cannot. That is not an SEO play, it is a scarcity play, and it is one of the more honest reasons to buy a second name.

One thing to check before you buy variants: make sure you own the one people will actually type first. Owning six near-misses and not the obvious version is a surprisingly common and very expensive arrangement.

A practical sequence

1. Write down ten candidates. Not one. Ten. First ideas are rarely best and you need something to compare against.
2. Say each one out loud to another person. Cut everything that fails the phone test. This usually removes half the list and it removes the right half.
3. Check availability. Some of what is left may simply be unregistered, and that is the cheapest possible outcome. Check before you assume — people overpay for variations of names that were free the whole time. Our registration page covers this.
4. Check the ones that are taken. Taken is not the same as unavailable. Find out who holds it, because plenty of them are for sale.
5. Look at what already exists. There are hundreds of exact-match names in our portfolio, owner-direct. Two minutes of searching beats three months of negotiating with a stranger.
6. Run the trademark search. Before money moves. Every time.
7. Grab the obvious near-miss. If the plural or the common misspelling is cheap, take it and point it at the main name. It costs less than losing the traffic.

Then stop optimizing and go build something. The name is a sign on a building. It matters, it is worth getting right, and it is not the business.