How to run a naming session that ends in a decision

Naming sessions rarely fail for lack of ideas. They fail because nobody agreed in advance what a good answer would look like, so every name gets judged against every other name, the list grows, and the meeting ends with a promise to sleep on it. Four weeks later the company is named after whichever option was least objectionable to the person who was most tired.

The fix is structural and it is mostly about doing things in a different order.

Write the brief before anyone suggests a name

Half a page, agreed by everyone who has a say, before a single option exists. It answers four questions.

• What does the company do, in a sentence a customer would recognise? Not the vision. The thing people pay for.
• Who says the name out loud, and where? A name said by a nurse to a patient, a driver to a dispatcher, and a founder on a stage are three different briefs, and the third is the least important one.
• What must the name not sound like? This is the most useful line in the brief and the one most often skipped. Every category has a sound it defaults to, and deciding to avoid it is a real constraint.
• How far does it have to travel? One country, one language, one industry — or not.

The brief is what candidates get judged against. Without it, the only available comparison is name against name, which is a matter of taste, and taste arguments are won by whoever cares most rather than whoever is right.

Agree the decision rule while nobody has a favourite

Decide, in the same meeting as the brief, who chooses. One person deciding with advice is a good rule. A small group with a named tie-breaker is a good rule. A vote is not a rule, because a vote reliably selects the option with the fewest enemies, and names that nobody objects to are usually names nobody remembers.

Write the rule down. It is uncontroversial before there are candidates and impossible to agree afterwards.

Generate apart, judge together

Group brainstorming is a poor way to produce options and an excellent way to produce consensus around the first plausible one. Send the brief out, give everybody a few days, and ask for their own list privately.

Use whatever helps: a thesaurus, other languages, the vocabulary of the industry, a generator. Tools produce quantity cheaply and it is worth taking the quantity, because the work that matters comes next and is not something a tool does.

Then pool everything into one list with no attribution. Names that arrive attached to the person who suggested them get discussed politically.

Check availability before anybody falls in love

This is the change that saves the most time, and almost nobody makes it.

The usual order is: shortlist, present, agree, then check — and the check kills the winner. Now the room has spent its enthusiasm and has to restart with a worse feeling about the whole exercise. Do it the other way round. Before the shortlist, take every surviving candidate and spend two minutes each on the .com, the company register, the app stores and the trademark registers.

Two minutes is enough to sort them into three piles. Clear. Complicated but possible. Dead. Only the first two pile go into the shortlist, and the middle pile goes in labelled — a name whose .com is owned by somebody who might sell is a live option with a cost attached, and the room can weigh it honestly if it is told.

There is a longer piece on what a taken .com actually means, and one on the checks worth running before committing to a name, in the Journal.

Shortlist to five, and make the five different

Five is enough to feel like a choice and few enough to hold in the head. What matters more than the number is that they are not five versions of one idea. If four are invented words and one is descriptive, the descriptive one will lose on unfamiliarity rather than on merit, and the room will never have compared the two approaches.

A useful shortlist has at least one name from each direction the brief permits: something plain, something invented, something borrowed from another field or language. Then the decision is about which kind of name the company wants to be, which is the real question, rather than which arrangement of letters is prettiest.

Judge them somewhere unflattering

Not on a slide, in large type, centred, on a dark background. Every name looks good there and the comparison is worthless.

Put the five on the least glamorous things the company will ever produce: an invoice footer, a delivery label, a lanyard, a browser tab. Say each one down a phone to somebody who has not seen it written and ask them to type it back. Read each one out in a sentence that a real customer would say, which is usually some version of "we use [name] for that".

Names fail differently under those conditions than they do on a slide, and the failures are the ones that turn up later at cost.

Decide, then write down why

The person named in the decision rule decides, in the room, on the day. Then, before the enthusiasm fades, somebody writes a paragraph: what the name is meant to do, what is known to be wrong with it, and what the company decided to live with.

Every name has something wrong with it. A company that has written down what it accepted is a company that can answer the question in eighteen months when somebody new asks why it is not called something clearer. A company that has not will hear that question as criticism and may act on it.

The four ways the room fails

1. No brief, so every name is judged against the other names and the argument becomes taste.
2. A vote, which selects the least objectionable option rather than the best one.
3. The availability check left until last, which kills the favourite after the room has already committed to it emotionally.
4. Somebody with a veto who was not in the room. Find out who that is on day one, not at the end.

When to stop

A naming process that has run for more than about three weeks is no longer producing better names; it is producing fatigue and a larger list. If the shortlist has been through the tests above and one of them is fine, take it. Companies are remembered for what they did and the name acquires its meaning afterwards, which is why so many good ones look arbitrary in hindsight.

The exception is a name that failed a check rather than a preference — a trademark collision, a bad reading in a market you sell to, an address you cannot own. Those are reasons to keep going. Not liking any of them yet is not.

Working alone rather than as a team? How to come up with a business name sets out the same path for one person.