How to choose between two names you both like

Most naming projects do not end with a bad name. They end with two good ones and a room that has run out of arguments. Everyone has said their piece twice, the founder likes one and the designer likes the other, and the decision has quietly become a question of who gives in first. That is the worst way to choose, because the person who gives in remembers it.

The way out is to stop arguing about which name is better and start asking which name is better at something. Two names that tie on taste rarely tie on the tests below, because the tests measure different things from the ones a shortlist measures.

First, check the tie is real

Before any of this, make sure both names are actually available at the same price of effort. If one .com is for sale and the other belongs to a company that will never sell, you do not have two finalists; you have one finalist and a wish. What to do when the .com is taken is a separate question, and it should be answered before the tiebreak, not after. A surprising number of deadlocked rooms are deadlocked over a name nobody can have.

1. The phone test, run properly

Ring somebody who has not seen either name written down. Say each one once, at normal speed, and ask them to type it into a message. Do it with five people, not one. The name that comes back spelled correctly more often wins this round, and the round matters more than it seems: every customer who hears your name before they see it will go through exactly this.

Invented names lose this test more often than real words, but not always. A coined word with one obvious spelling (four or five letters, no doubled consonants, no silent anything) can beat a real word with two spellings. Do not guess; ring the five people.

2. The invoice test

Put each name at the top of an invoice, in plain black type, at the size an invoice uses. No logo, no colour, no tagline. Then put each on an app icon at the size a phone shows it. Then on a lanyard.

One name will look like a company and the other will look like a project. The difference is usually letter shape: a name with a mix of tall and round letters holds its shape small, and a name that is all one height goes grey. This is the test the designer wins with, and it is a fair win, because these are the sizes the name will spend most of its life at.

3. The sentence test

Write the three sentences the name will live in most: "Hi, this is [name]", "They got it from [name]", and "[Name] raised a round". Say them out loud.

Names that end in a vowel sound tend to sit better in front of a verb. Names that are also common words trip over themselves in a sentence, because the listener has to work out whether you mean the word or the company. If one name needs "the" in front of it to sound right, that is a fact about the name that will follow it everywhere.

4. The five year test

Describe the company in five years in one line: what it sells, to whom, and what it has stopped selling. Then ask whether each name still fits. A name that names the first product is a good name for a product and a bad name for a company. A name that means nothing fits everything, which is its strength and its weakness: it will never be wrong and it will never do any work on its own.

The useful middle is a name that suggests the category without stating the product. If one of your finalists is descriptive and the other is invented, the difference is worth a piece of its own, and it usually decides this round.

5. The regret test

Imagine each name has been chosen and it is a year later. Somebody at a conference mishears it, or a competitor with a similar name launches, or the company pivots. Which name would you be more annoyed to have picked? Regret is a better guide than enthusiasm here, because enthusiasm is what got the room stuck in the first place.

Scoring it

Do not add up points. Five tests with equal weight produce a tie as easily as taste does. Instead, before you run any of them, agree which one matters most for this company. A business that will be recommended by word of mouth weights the phone test. A business that lives in an app store weights the invoice test. A company that expects to pivot weights the five year test. Agree the weighting first, while nobody knows which name it favours; that is the whole trick, and it is the same trick that makes a naming session end in a decision rather than a longer list.

The tiebreak

If the tests genuinely split, take the shorter name. Shorter names are easier to say, spell, fit and remember, and the advantage compounds with every use. This is why the short names on Webba are a collection of their own: five letters or fewer is a property that keeps paying after everything else about the decision has been forgotten.

And if the tests split and the names are the same length, take the one the founder likes. Somebody has to say the name a thousand times in the first year, and it should be the person who will not get tired of it.

Two finalists, from the collection

Four names that sit in the same categories and would be sensible finalists against each other. Run the tests on a pair and see which round decides it.

Before it comes down to two, how long should a business name be? is a quick way to separate a shortlist.

Short .com namesThe tiebreak, as a collection.