More insights.
The learning never stops.
Strategy, creative and numbers feed each other every week. What the loop teaches us lands in your inbox.
Most brand identity is a deliverables list.
Logo. Palette. Two typefaces. A sixty-page book with clear-space diagrams. Every agency sells roughly this list, and the list is the problem: it can be completed, invoiced and handed over without a personality ever being found. You can spot the list-built identities instantly. They're tasteful, current, and could belong to any of forty companies in the category, because the process started with references instead of a person.
Twelve years of identity work taught us that the work that lasts starts with a hunt, not a moodboard. Here's the method, including the questions to answer before any designer opens a file.
Answer five questions before any pixel exists
1. Who is the customer, psychologically? Not demographics. What they trust, what they scroll past, what register makes them lean in and what makes them leave. Pull this from reviews, comments and the content they actually consume, not from a workshop.
2. What does the customer need to see before they believe? This is where the aesthetic gets decided, before design. Glutone's buyer wanted certainty and minimalism in one breath: comfortable with science, allergic to clutter. That read produced the look, a natural palette, humans against open sky, white space carrying the trust, before any designer touched a file. BigMe's young cohort buys with its eyes and distrusts anything medicinal, so the identity went modern and indulgent, closer to culture than pharmacy. Somma had to outrank a hemp ingredient's modern reputation, so the identity reached backwards into Sanskrit and the Himalayas, then forwards into tinted glass and emboss. Three brands, three aesthetics, one method. A style is what the designer likes. An identity is what the customer needs to see.
3. What must the name do? When the naming is ours, it's the first strategic act. Liquify is a verb in a category of vault-words, and the entire anti-institution personality lives in that one choice. A name is the cheapest place to install a personality and the most expensive place to miss one: get it generic, and every downstream asset works overtime compensating.
4. What worlds are we choosing between? Never present bare logo options. Build two or three complete personalities: each with its textures, applications, voice lines and motion already mocked. We ran this on our own rebrand, three full worlds, a glitch-pixel mark with neon renders, a desi-pop wordmark living on Indian street photography, and a stretched technical wordmark with a Devanagari companion. The choice wasn't the prettiest mark. It was the world we wanted to live in.
5. What rules make it reproducible? Decide what gets written down before deciding what gets designed. An identity only its original designer can execute isn't a system, it's a dependency.
You can't judge a logo. You can only judge a world.
This is the operating rule behind question four, and it's worth its own weight. A mark in isolation tells you nothing about how the brand feels on a business card, in a feed, on a hoarding, in motion. Presenting bare options forces the client to imagine all of that, and imagination is where good decisions die. Presenting worlds lets everyone feel the personality before committing. If you take one practice from this article, take this one: whatever identity decision you're facing, build the candidates out to world level before choosing.
Then write the rules down
Finding the personality is half the work. Making it reproducible is the other half. The rulebook covers: what each colour is for and where it's allowed to appear, not just which colours. Which register each typeface carries and when the secondary face takes over. How the motion behaves. How the photography is lit. And the shortest, most useful section: what the brand never does. These rules are what let a third designer, a contractor, six months from now, produce work that still sounds like the same personality. Our deepest system to date ran over a hundred pages, every page protecting one thing: the brand staying itself on a surface nobody had met yet.
Test it in the feed, not the boardroom
Most identities get judged at presentation scale, on a wall, with everyone nodding. Brands live in a feed, at thumbnail size, against a thousand competitors, with half a second to register. So the final test is pressure: does the identity hold at thumbnail? Can it produce two hundred ads without fatiguing? Does the register stop the specific customer it was built from? Building identities inside an agency that also runs media means we know from spend data which visual registers stop which people. A personality too fragile for performance creative isn't a personality. It's an art project.
The principle
An identity is a personality, and a personality has to be hunted: in the customer's psychology, in the name, in complete worlds built and compared, in rules that make it reproducible, in the feed where it has to fight. The deliverables list is what you hand over at the end. It was never the work. A brand book describes a personality that already exists; if the hunt didn't happen, the book documents nothing.
If you want to see what this looks like applied to specific brands, the work is here. If you want to talk about whether it fits what you're building, book a call.



