How to name a new market category

On March 20, 1986, McDonald's opened in Rome, in Piazza di Spagna. Protesters gathered outside, and among them was a journalist from Piedmont, Carlo Petrini, who moved through the crowd handing out plates of penne. That same year he founded Arcigola, the association that would later become Slow Food. Three years after that, at the Opéra-Comique in Paris, delegates from fifteen countries signed the movement's manifesto. The name they signed was in English, and it was built as the exact reverse of the thing it opposed.

Two words, one of them flipped. Anyone who has heard “fast food” understands “slow food” without an explanation, and also understands which side it is on.

Why does the category name matter more than the brand name?

Because it is the word customers search with. Nobody, the first time around, searches for the name of a company they don't know. They search for the name of the thing they need. And the name of that thing decides who you will be compared with. If a customer describes you as “a marketing agency,” they will compare you with marketing agencies, at marketing agency prices.

In our guide to Category Design we explained that a new category starts from a problem the market has not named yet. The name is the moment that problem becomes a word people can repeat. A brand name helps people recognize you; a category name helps them recognize the problem. You are the one who uses the first. Everyone else has to use the second, or it doesn't work.

When do you choose the name of a new category?

At the end. First you define the problem, then who has it, then how the solution differs from the ones that already exist. Only then do you look for the word.

The order matters because people who start from the name almost always end up describing their own product. The product is what the company knows best, so it is the first thing that comes to mind. But customers don't know the product. They know their problem. A name that describes the product forces them to translate, and almost nobody does.

The lakehouse case shows this well. On January 30, 2020, five authors at Databricks, an American data software company, published a post titled “What is a Lakehouse?”. The text opens with an observation about customers: a new way of organizing data that had emerged independently at many of them, in different situations. The name comes after the phenomenon has been described, and it is defined as an open system that combines the best elements of data lakes and data warehouses.

What makes a good category name?

Looking at the names that have lasted, a few traits keep coming back. None of them is a written rule. This is our reading of what they have in common.

It makes sense in words the customer already knows. “Lakehouse” works because people who work with data knew both halves. James Dixon, then chief technology officer at Pentaho, introduced “data lake” on October 14, 2010, with a comparison anyone could follow: if a store of organized data is bottled water, clean and packaged, a data lake is a large body of water in its natural state. Data warehouse, the store for organized data, was an even older term. Databricks didn't have to teach anyone a new word. It glued two old ones together.

It says what it is pushing against. Slow food says fast food without naming it. “Inbound marketing,” the term coined by Brian Halligan, who founded HubSpot with Dharmesh Shah in June 2006, does the same thing with traditional marketing, in our reading: “inbound” means incoming, customers who arrive on their own instead of being chased. A name with an implied opponent does half the explaining, because the customer already knows the problem and links it to the word you are setting yourself against.

It can be used like any other word. On December 17, 1997, Jorn Barger coined “weblog.” In May 1999 Peter Merholz, as a joke, split the word into “we blog.” In 2004 the American dictionary Merriam-Webster chose “blog” as its word of the year. In less than seven years the word had been shortened, turned into a verb and taken to the top of a dictionary by people who had nothing to do with whoever invented it. A category name you can't shorten, conjugate or drop into a casual conversation rarely makes it out of company presentations.

It leaves room for others. To a customer, a category with only one company in it is a product. The name has to work for anyone selling something similar, and for the people who assess the market. In 2005 two Gartner analysts, Mark Nicolett and Amrit Williams, introduced the acronym SIEM for a type of cybersecurity software. It combined two categories that already existed, security information management (SIM) and security event management (SEM). The name didn't belong to any vendor, and our reading is that this is exactly why every vendor could adopt it.

Is it better to invent a word or combine existing ones?

It is almost always better to combine words that already exist. A word invented from scratch has to be learned first, then understood, then remembered. A word built from familiar parts skips the first step.

The history of the Italian language says the same. “Agriturismo” is agriculture plus tourism, and anyone hearing it for the first time understands what it means. With Law no. 730 of December 5, 1985, titled “Rules on agritourism,” the word entered national law. When a category ends up in the title of a law, it has stopped being someone's idea and become the way an entire country names a thing.

Invented words make sense when the thing looks like nothing people know and there is time, and budget, to teach them. For a midsize company selling to customers who are short on time, the math usually doesn't work.

Can the category name be the same as the brand name?

It is the strongest temptation, and the most expensive one. The idea is simple: if the category is named after me, every time someone mentions it, they are advertising me. In practice the opposite often happens.

In 1900 the inventor Charles Seeberger registered the word “escalator” as a trademark for his moving staircase, and the following year he sold the rights to the Otis Elevator Company. Decades later a competitor, the Haughton Elevator Company, challenged the registration before the U.S. Patent Office. In 1950 the office ruled in its favor: the public recognized the word as the name of a moving staircase, not as an indication of who made it. “Escalator” became everyone's.

The reverse path leads to the same place. In the 1990s Psion had registered the trademark “Netbook,” tied to one of its older devices, the Netbook Pro. When the word began to describe small, inexpensive laptops from many manufacturers, Psion defended it as its own. In February 2009 Intel filed suit to free the word, and in June 2009 Psion voluntarily withdrew all its registrations. A company owned the name of the category, but not the category.

Our reading is this: the category name and the brand name do two different jobs, and it pays to keep them apart. The category has to be free, because the more people use it, the more real it becomes. The brand has to stay yours, and it grows strong when the market associates it with the category as its best example.

How do you test a name before using it?

Before you print a name on a website, it is worth putting it through a few simple checks. These are the ones we use at Project Mantra, and they don't require any special tools.

1. The sentence test. The name has to fit inside an ordinary sentence: “We're looking for a…,” “Have you ever tried a…?”. If it doesn't fit without sounding odd, nobody will use it.
2. The phone test. Say it out loud to someone in your industry, without explaining. If they ask you to repeat it or spell it, you have a pronunciation or spelling problem.
3. The competitor test. Picture your main competitor using your name in their advertising. If that feels natural, the name describes a category. If it feels like theft, you have written a product name.
4. The opposite test. Ask yourself which existing word your name stands against. If there is no answer, the customer will have nothing to compare it with, and you will have to explain everything from scratch.
5. The search test. Type the name into a search engine. If the first results are about something else, or if the name is already someone's registered trademark, start over.

Then there is a test you can only run later: the name works when other people use it. Halligan and Shah's book, “Inbound Marketing,” came out in 2009. By July 2011 it was in its seventh printing, with 40,000 copies sold and translations in nine languages, and the founders had built a movement around the term, including a conference, INBOUND. The name had left the company.

What are the most common mistakes in naming a category?

Using an acronym only you understand. SIEM worked because it combined two acronyms the industry already used. A new acronym the customer can't decode is one more obstacle.

Naming a feature instead of a problem. “Faster,” “smart,” “advanced” describe a better product inside a category that already exists. The customer hears an adjective and compares it with the old product, at the old product's price.

Adding a number or a prefix to an old category. “2.0,” “next generation,” “smart” admit that the category is still the one it was before. If it really is, that's fine. But then it isn't a new category, and the signs that tell you so usually show up already in the reasons you lose customers.

Changing the name every year. A category name works by accumulation: every time someone repeats it, it becomes a little more real. Whoever changes it after the first campaign that doesn't work starts again from zero, and leaves the word free for whoever comes next.

Falling for the name before the thing exists. If the name promises a new category and the customer gets the same product as before, at the same price and with the same service, the name becomes proof that you overreached.

Does the name have to be in Italian?

It depends on how your customers already talk about the problem. Slow Food was born in Piedmont with an English name, and our reading is that this wasn't an affectation: the opponent, fast food, had an English name, and the wordplay only worked in the same language. Agriturismo was born in Italian because it spoke to people who lived and worked in the Italian countryside.

For a company that sells mainly in Italy, to customers who don't work in English, a clear Italian word is more likely to be repeated. In an industry that already speaks English, such as software or finance, an Italian word can sound like a translation and get dropped. The question to ask is simple: what words does the customer use today to describe the problem, when you aren't there listening?

The 1950 Otis decision added a detail worth remembering for anyone who wants their category to carry their own name. In the ruling, the office noted that Otis itself had used “escalator” as a generic term, both in several of its own patents and in its own advertising.