The resource How Brands Are Built honestly records this confusion: on the internet, one can find classifications with 5, 6, 7, 10, 13, and even 18 categories, and all of them claim completeness. The reason is simple — authors of articles often combine in one list features of different natures. Some terms describe approach to meaning — how directly the name speaks about the product: from descriptive (speaks directly) through suggestive, associative (hints) to abstract (says nothing). Other terms describe construction of the word — what it is physically made of: a real word of the language, a composite of two roots, or a completely invented neologism.
Professional naming agencies do not dispute the "correct" number of types, but work with a two-dimensional grid where these axes intersect. A name can simultaneously be, for example, associative in meaning and a real word in construction — this is Apple. Or it can be descriptive in meaning and a neologism in construction — this is closer to "Coffeemania." Here is how this matrix looks in practice:
| Approach / Construction | Real word | Composite | Neologism |
|---|---|---|---|
| Descriptive | «Moloko» for a dairy product | Aeroflot | «Cakepops» |
| Suggestive / associative | Apple, Amazon, Nike | Facebook, Instagram | Häagen-Dazs |
| Abstract | — | — | Kodak, Xerox, Yandex |
It is this very grid that explains the discrepancy in the numbers: if you count only rows — you get 3 type, if only columns — also 3, if you multiply them — 9, and if you add subtypes like eponyms and abbreviations, it is easy to reach 18. All classifications are locally correct and wrong when they claim to be the only true one.
The author of the book «Naming. How Playing with Words Becomes Business» suggests not stretching the matrix to infinity, but rather fixing five practically occurring categories that are most often encountered in the work of agencies:
| Category | Essence | Example |
|---|---|---|
| Constructed | Assembled from word parts or morphemes | Facebook, «Gazprom» |
| Existing words | A common word of the language in a new meaning | Apple, Patagonia |
| Neologisms | A made-up word without dictionary meaning | Kodak |
| Classical / eponyms | The name of a person or founder | Chanel |
| Shortened / abbreviations | Shortening of a long name | INTERPOL |
This classification is convenient because it describes origin rather than meaning — and it almost does not intersect with the axis of «approach» mentioned above. Therefore, it should be perceived as a complement to the matrix, not as an alternative.
The difference between these types is not philosophical, but quite monetary — it determines whether a trademark will be registered. A descriptive name directly indicates the type of product, its properties, or purpose, and precisely for this reason Article 1483 of the Civil Code of the Russian Federation prohibits registering such designations as trademarks: they must be available to all market participants, not monopolized by a single company.
A telling example is the attempt to register the word «cakepops» (cake pops, a type of dessert on a stick). The patent office rejected it: the word directly describes a type of product, not an image around it, and any competitor has the right to use this term to describe their product. Associative names work differently: they require the consumer to «fill in the blanks» — there is no direct connection between the word and the product; the imagination creates that link. It is precisely this semantic distance that makes such names legally protectable: examination sees not a product description, but an original image.
From this practical conclusion follows: if a business plans to aggressively protect the brand and invest in recognizability for years to come, a descriptive name is a risk of losing legal grounds exactly at the moment when the brand begins to become valuable.
Nike, Amazon, Apple, Monocle — all of these are associative names built on real words of the language taken from another semantic field. Nike is the name of a goddess of victory, not literally connected to athletic footwear. Amazon is a river, not a store. This is not a coincidence, but a conscious strategy: such words become 'empty vessels' into which the brand pours the necessary meaning at the start through advertising, design, and user experience.
This works more effectively than descriptive names for one reason — flexibility. The word 'Milk Farm' is rigidly tied to a category and prevents the company from, for example, entering the production of premium cheese under the same name. The word 'Apple' originally promised nothing specific, so the company could grow from computers to phones, watches, and services without semantic conflict with the name.
But the metaphorical approach has a price. At the start, such a name tells nothing about the product — the consumer does not understand what 'Amazon' sells until they see advertising or encounter the brand in practice. This means mandatory marketing investments in the first years of the brand's life: without them, an associative name remains just a beautiful set of letters. Indeed, this is where the difference between the three functions of a commercial name becomes apparent — nominative (simply distinguishing the brand from others), informative (communicating the product's properties), and advertising (creating an emotional image). Descriptive names are strong in the informative function and weak in the advertising function, while metaphorical names are the opposite.
There is no universal 'best' type, but there are working criteria by which you can narrow the choice down to one or two cells in the matrix.
In practice, choosing the type of name is not about finding the perfect term from a list, but solving a specific business task: to protect the mark, grow beyond the category, or quickly explain the product to the market. The approach matrix and constructions help see what each cell pays for and what the brand buys for it, instead of blindly going through other examples.
The resource How Brands Are Built honestly records this confusion: on the internet, one can find classifications with 5, 6, 7, 10, 13, and even 18 categories, and all of them claim completeness. The reason is simple — authors of articles often combine in one list features of different natures. Some terms describe approach to meaning — how directly the name speaks about the product: from descriptive (speaks directly) through suggestive, associative (hints) to abstract (says nothing). Other terms describe construction of the word — what it is physically made of: a real word of the language, a composite of two roots, or a completely invented neologism.
Professional naming agencies do not dispute the "correct" number of types, but work with a two-dimensional grid where these axes intersect. A name can simultaneously be, for example, associative in meaning and a real word in construction — this is Apple. Or it can be descriptive in meaning and a neologism in construction — this is closer to "Coffeemania." Here is how this matrix looks in practice:
| Approach / Construction | Real word | Composite | Neologism |
|---|---|---|---|
| Descriptive | «Moloko» for a dairy product | Aeroflot | «Cakepops» |
| Suggestive / associative | Apple, Amazon, Nike | Facebook, Instagram | Häagen-Dazs |
| Abstract | — | — | Kodak, Xerox, Yandex |
It is this very grid that explains the discrepancy in the numbers: if you count only rows — you get 3 type, if only columns — also 3, if you multiply them — 9, and if you add subtypes like eponyms and abbreviations, it is easy to reach 18. All classifications are locally correct and wrong when they claim to be the only true one.
The author of the book «Naming. How Playing with Words Becomes Business» suggests not stretching the matrix to infinity, but rather fixing five practically occurring categories that are most often encountered in the work of agencies:
| Category | Essence | Example |
|---|---|---|
| Constructed | Assembled from word parts or morphemes | Facebook, «Gazprom» |
| Existing words | A common word of the language in a new meaning | Apple, Patagonia |
| Neologisms | A made-up word without dictionary meaning | Kodak |
| Classical / eponyms | The name of a person or founder | Chanel |
| Shortened / abbreviations | Shortening of a long name | INTERPOL |
This classification is convenient because it describes origin rather than meaning — and it almost does not intersect with the axis of «approach» mentioned above. Therefore, it should be perceived as a complement to the matrix, not as an alternative.
The difference between these types is not philosophical, but quite monetary — it determines whether a trademark will be registered. A descriptive name directly indicates the type of product, its properties, or purpose, and precisely for this reason Article 1483 of the Civil Code of the Russian Federation prohibits registering such designations as trademarks: they must be available to all market participants, not monopolized by a single company.
A telling example is the attempt to register the word «cakepops» (cake pops, a type of dessert on a stick). The patent office rejected it: the word directly describes a type of product, not an image around it, and any competitor has the right to use this term to describe their product. Associative names work differently: they require the consumer to «fill in the blanks» — there is no direct connection between the word and the product; the imagination creates that link. It is precisely this semantic distance that makes such names legally protectable: examination sees not a product description, but an original image.
From this practical conclusion follows: if a business plans to aggressively protect the brand and invest in recognizability for years to come, a descriptive name is a risk of losing legal grounds exactly at the moment when the brand begins to become valuable.
Nike, Amazon, Apple, Monocle — all of these are associative names built on real words of the language taken from another semantic field. Nike is the name of a goddess of victory, not literally connected to athletic footwear. Amazon is a river, not a store. This is not a coincidence, but a conscious strategy: such words become 'empty vessels' into which the brand pours the necessary meaning at the start through advertising, design, and user experience.
This works more effectively than descriptive names for one reason — flexibility. The word 'Milk Farm' is rigidly tied to a category and prevents the company from, for example, entering the production of premium cheese under the same name. The word 'Apple' originally promised nothing specific, so the company could grow from computers to phones, watches, and services without semantic conflict with the name.
But the metaphorical approach has a price. At the start, such a name tells nothing about the product — the consumer does not understand what 'Amazon' sells until they see advertising or encounter the brand in practice. This means mandatory marketing investments in the first years of the brand's life: without them, an associative name remains just a beautiful set of letters. Indeed, this is where the difference between the three functions of a commercial name becomes apparent — nominative (simply distinguishing the brand from others), informative (communicating the product's properties), and advertising (creating an emotional image). Descriptive names are strong in the informative function and weak in the advertising function, while metaphorical names are the opposite.
There is no universal 'best' type, but there are working criteria by which you can narrow the choice down to one or two cells in the matrix.
In practice, choosing the type of name is not about finding the perfect term from a list, but solving a specific business task: to protect the mark, grow beyond the category, or quickly explain the product to the market. The approach matrix and constructions help see what each cell pays for and what the brand buys for it, instead of blindly going through other examples.