Brand Naming · Frameworks
The best brand naming frameworks.
The world has four canonical brand-naming agencies whose published frameworks define the craft. Lexicon in Sausalito, founded by David Placek, is responsible for Pentium, BlackBerry, Subaru, Swiffer, Sonos, Dasani, and a dozen other names Americans say every week without thinking about them. Operative Words in Brooklyn publishes the cleanest public taxonomy of creative moves. Eat My Words in Sausalito, run by Alexandra Watkins (author of Hello, My Name Is Awesome), gave the industry the SMILE/SCRATCH mnemonic. Catchword in Oakland anchors on a clarity-and-distinctiveness model that has produced two decades of named entities at scale.
Each framework illuminates a different surface of the same shape. Reading all four together is the closest a founder can get to the encoded craft of the profession without paying the six-figure agency engagement. This article unpacks each, with the worked examples that show where each framework is most useful.
1. Lexicon — the phoneme-association framework
David Placek's thesis is that brand names work the way poetry works: the phonemes prime an association before the word's denotative meaning has time to register. The K of Kodak primes a click. The P of Pentium primes power. The long vowel of Dasani primes calm and luxury. The choice of phoneme is the choice of feeling, made upstream of conscious cognition.
The framework names a small set of phoneme-association categories — speed (hard plosives), comfort (soft labials), luxury (long vowels and liquid consonants), tech (hard fricatives and edge consonants), warmth (nasals). The naming work, in this framework, is to identify the desired feeling and then search the phoneme space that primes it.
Where it's most useful: the moment a founder has three finalists and is choosing between them. The Lexicon framework lets you score each finalist on the phoneme-association axis and see which one is priming the feeling closest to the brief. The Sapir-1929 / Köhler-1929 / Maurer-2006 / Ćwiek-2022 line of research underneath this framework is summarized in Sound Symbolism for Brand Naming. The phonosemantic axis Etymolt scores is the operationalization of Placek's model.
Worked example: Linear. The /l/ onset is liquid (calm, smooth), the /i/ short vowel is precise, the /n/ medial is nasal (warm, internal), the /ər/ close is soft. The aggregate signal is calm-precise-warm — the exact register a software product targeted at engineers wants. The name does the work of a brief in two syllables.
2. Operative Words — the creative-move taxonomy
Operative Words' Brooklyn shop publishes the cleanest taxonomy of creative moves in the industry. The moves are the meta-categories of where a name comes from — the answer to the question how did the namer arrive at this candidate. There are roughly nine, and most professional shortlists span four or five of them by design.
- Metonym — a part of the thing stands for the whole. Stripe (the magnetic stripe stands for the payment), Anchor (the visual stands for the audio host).
- Gerund — an -ing word that names an activity. Tinder (the verb form of tinder, lighting), Glassdoor.
- Verb — the imperative or bare verb that describes what the user does. Slack, Zoom, Plaid.
- Compound — two words mashed into one. Snowflake, Facebook, Tinybird.
- Mythology — Greek, Roman, Norse, Egyptian names. Apollo, Hermes, Athena.
- Minimalism — short, often three-letter names. Cal, Mux, Bun, Neon.
- Found-object — a real word from an unrelated domain, deliberately repurposed. Stripe, Bolt, Loom.
- Translation — a non-English word. Asana (Sanskrit), Zara (likely Hebrew origin).
- Misspelling — phonetic respelling of a real word. Lyft, Tumblr, Flickr.
Where it's most useful: the brainstorm. A brainstorm that ignores the taxonomy returns sixty candidates that cluster in three of the nine moves. A brainstorm that uses the taxonomy as a checklist returns candidates across all nine. The shortlist is more diverse and the founder makes a real comparison instead of comparing apples to apples to apples.
3. Eat My Words — the SMILE and SCRATCH rubric
Alexandra Watkins' Hello, My Name Is Awesome is the bestselling naming book in the industry, and the SMILE/SCRATCH mnemonic is the most-used rubric in the field. SMILE names what a great name does; SCRATCH names what a bad name suffers. The rubric is fast to teach and fast to apply.
SMILE — what makes a name great:
- S — Suggestive. Suggests a benefit or property of the product without being descriptive.
- M — Meaningful. Connects to a real concept a customer understands.
- I — Imagery. Triggers a picture in the reader's head.
- L — Legs. Has room to grow with the business (not over-narrow to the first product).
- E — Emotional. Evokes a feeling.
SCRATCH — what makes a name suffer:
- S — Spelling. Hard to spell from hearing it.
- C — Copycat. Sounds like an existing brand.
- R — Restrictive. Boxes the business into one product or geography.
- A — Annoying. Cute, clever, or twee in a way that wears on repeated use.
- T — Tame. Forgettable, generic, blends in.
- C — Curse. Inadvertent profanity or slur in a relevant market.
- H — Hard-to-pronounce. Pronunciation is ambiguous from the spelling.
Where it's most useful: the post-brainstorm cull. Score each candidate on every SMILE letter (1-5 scale) and flag every candidate that triggers any SCRATCH letter. The candidates that survive are the ones worth verifying. The Curse axis maps almost exactly to the cultural verification Etymolt runs — see The Five Ways a Brand Name Dies for the cultural failure-mode frequencies in our corpus.
4. Catchword — the clarity-and-distinctiveness model
Catchword's Oakland shop has published a model anchored on a single trade-off:clarity (the buyer immediately understands the category) versus distinctiveness (the buyer remembers and the lawyer can defend). Every credible name sits somewhere on this axis, and the right position depends on the cohort and the category.
The trade-off is real and brutal. A maximally clear name ("Online Bookstore") is unprotectable and unforgettable in the wrong direction. A maximally distinctive name ("Xobni") is protectable and forgettable. The Catchword view is that the right answer lies somewhere in the middle and that the position depends on the business's tolerance for category-education spend.
Where it's most useful: evaluating founder shortlists that lean too far in either direction. A shortlist of nine real English words and one coined word probably needs more distinctiveness. A shortlist of nine coined words probably needs more clarity. Catchword's contribution is the explicit articulation of the trade-off as a slider, not a binary.
How Etymolt scores against the frameworks
Etymolt's methodology operationalizes the testable claims of all four frameworks. The phonosemantic axis is Lexicon's phoneme-association framework, calibrated against the Kawahara & Shinohara corpus. The creative-move tagging is the Operative Words taxonomy. The Curse and Hard-to-pronounce SCRATCH flags map onto the cultural and pronunciation axes. The clarity-distinctiveness balance shows up in the §2(e)(1) descriptiveness probability and the trademark distinctiveness score.
The verifier returns a single Clearance Confidence Score (0-100) and a verdict (PROCEED, STRATEGIC, ABANDON). Every flag traces to a record number. The full methodology lives at /methodology and is recalibrated weekly. The case studies show the frameworks in action: Arq PROCEED is a textbook all-four-frameworks-aligned candidate.
How to use the four frameworks together
The frameworks are complementary, not substitutable. The best naming process uses each at a different stage of the work.
At the brief stage, use Catchword's clarity-and-distinctiveness model to pick the slider position. The founder commits to a target — distinctive coined word, semi-clear metonym, clear real-word — before the brainstorm. The commitment prevents the shortlist from drifting across the clarity axis as the brainstorm proceeds.
At the brainstorm stage, use the Operative Words taxonomy as a checklist. Run six rounds of ten candidates, one round per creative move (Metonym, Verb, Compound, Mythology, Latinate, Found-object). The shortlist that emerges spans the taxonomy and gives the founder a real comparison.
At the cull stage, use Eat My Words' SMILE/SCRATCH rubric. Score each candidate on every SMILE letter (1-5) and flag every SCRATCH trigger. The candidates with the highest aggregate SMILE and zero SCRATCH flags advance.
At the choosing stage, use Lexicon's phoneme-association framework to evaluate the final three. Score each on the phoneme-feeling alignment with the brief. The candidate whose phonemes prime the briefed emotion is the answer.
What the agencies do not publish
None of the four agencies publishes a complete verification framework — and this is the gap Etymolt was built to close. The four frameworks above describe how to generate a good name. They do not describe how to attest that a specific candidate is clear. The attestation requires running USPTO records, domain registries, social-handle APIs, foreign-language lexicographic sources, and acoustic round-trip pipelines, with every flag traced to a record number.
Etymolt's methodology operationalizes the four frameworks into testable claims, then layers the attestation. The verdict is the conjunction of the framework score and the attestation receipts. A name that scores well on every framework but fails the trademark axis is not cleared. A name that clears every attestation but fails the phonosemantic framework is not optimal. The verifier composes both views.
Related reading
- What Makes a Good Brand Name — the 6 Criteria — the operational rubric that synthesizes all four frameworks.
- SaaS Brand Naming Patterns — the creative-move taxonomy applied to the 32-name SaaS canon.
- Sound Symbolism for Brand Naming — the empirical research underneath Lexicon's framework.
Disclaimer
Etymolt operates under the Bureau Model. We surface clearance signals across the five axes; we do not provide legal advice. A trademark opinion belongs to a licensed attorney in your filing jurisdictions.
Take the next step
Score a name against every framework — in three seconds.
The verifier operationalizes the testable claims of Lexicon, Operative Words, Eat My Words, and Catchword. Five free per IP. Every flag traces to a record number.
Verify a name