This article argues that distinctiveness has quietly become the most defensible advantage a brand can hold, precisely because artificial intelligence has made everything around it easier to copy.
Distinctiveness is recognition, and it is not the same as being different
Two words get used as if they mean the same thing. They do not. Differentiation is a claim about substance: this product lasts longer, this service is faster, this platform integrates more cleanly. Distinctiveness is a claim about recognition: this is the brand with the yellow arches, the tick, the six-note sound. The Ehrenberg-Bass Institute, the largest centre for marketing science in the world, draws the line sharply. Differentiation is about features that set a brand apart in the customer's judgement. Distinctiveness is about assets that let a customer identify the brand at all.
The reason the distinction matters is commercial, not academic. Byron Sharp and Jenny Romaniuk, whose work at Ehrenberg-Bass reshaped how serious marketers think about growth, argue that buyers rarely choose brands because they have reasoned their way to a meaningful difference. They choose the brands they can remember and find easily at the moment of need. Distinctive brand assets, the colours, logos, characters, and slogans that trigger recognition, are what make a brand easy to notice and easy to buy. Difference is hard to sustain and easy to erode. Recognition, once built, compounds.
That was already the more evidence-backed position before AI entered the room. What AI does is turn a debate into a decision.
AI has made competence cheap, which makes sameness the default
Here is the shift that reorganises the whole argument. Generative tools have collapsed the cost of producing competent brand work. A passable logo, a serviceable name, a clean set of brand guidelines, a launch campaign in three formats: all of it is now minutes of prompting rather than weeks of craft. The floor has risen. Almost nobody produces embarrassing work any more. The problem is that the ceiling has come down to meet it.
This is not a hunch. In 2024, Anil Doshi and Oliver Hauser published a controlled study in Science Advances, one of the most rigorous journals in the field, testing what happens when writers are given AI-generated ideas. Individual stories became more creative, better written, and more enjoyable, especially for less naturally creative writers. But the stories also became measurably more similar to one another. The authors named the trade-off plainly: generative AI enhances individual creativity while reducing the collective diversity of novel content. The machine helps each person and homogenises everyone.
Apply that to branding and the consequence is direct. If every founder in a category briefs the same three tools with the same category language, the tools return the same conventions, the same palettes, the same reassuring roundness, the same verbs. The result is a market where everyone is competent and nobody is distinct. The banks all look like fintechs, the clinics all look like wellness apps, and the AI answer engine describing the category cannot tell any of them apart, because there is nothing to tell apart.
When competence becomes free, the only scarce thing left is the quality of being unmistakable.
Consider the parallel in another industry. When digital photography made a technically correct image effortless, the value did not disappear, it moved. It moved to the photographers with a recognisable eye, a signature that survived being stripped of the credit line. The same relocation is happening in branding now. Craft that any tool can reproduce stops being an advantage. Craft that produces something no tool would have generated on its own becomes the whole game.
Exhibit 1: The distinctiveness ladder, four rungs from invisible to unmistakable
Distinctiveness is not binary. It is a position on a ladder, and most brands sit lower than their owners believe. The ladder below is a diagnostic Vantage uses to locate where a brand actually stands, tested by a single question: strip the name off, and how far up can the brand still be recognised?
| Rung | State | The test | What most AI-generated brands score |
|---|---|---|---|
| 4. Unmistakable | Recognised from a single asset, no name needed | A fragment of colour, sound, or shape identifies the brand | Rare. Cannot be generated, only built and defended over time |
| 3. Distinctive | Recognised from a combination of owned assets | Two or three assets together make the brand identifiable | Achievable, but only with deliberate ownership |
| 2. Conventional | Recognised only with the name attached | Remove the name and the brand dissolves into its category | The default output of generative tools |
| 1. Invisible | Not recognised even with the name | The brand is confused with competitors regardless of labelling | More common than owners admit |
Most brands that commission work today land on rung two and mistake it for rung three. They have a competent identity that looks professional and looks like everyone else. As shown in Exhibit 1, the work of distinctiveness is the climb from two to three and, for the ambitious, from three to four. That climb cannot be prompted. It is a series of ownership decisions about which assets a brand will commit to for long enough that the market learns them.
Distinctive assets are owned, consistent, and defended, or they are decoration
A distinctive asset only works if it meets three conditions, and most brand elements meet none of them. It must be owned, meaning the brand can credibly claim it rather than borrowing a category convention everyone already uses. It must be consistent, applied for long enough and widely enough that the market forms the association. And it must be defended, protected from the internal pressure to refresh, tweak, and modernise every time a new marketing lead arrives.
The failure mode is almost always the same. A brand builds a genuine asset, then dilutes it. The colour gets adjusted for a campaign. The logo gets flattened for a trend. The tone gets loosened for a younger audience. Each change feels reasonable in isolation. Together they reset the market's memory to zero. Ehrenberg-Bass research on distinctive assets is blunt on this: the value of an asset is a function of how many people link it to the brand and how uniquely they link it, and both are destroyed by inconsistency far faster than they are built by it.
Intel offers the clean example. Five notes, no words, decades of discipline. A brand that resisted every temptation to modernise the sound into irrelevance, and now owns a piece of audio real estate no competitor can occupy. That asset was not clever. It was consistent for long enough to become unmistakable. In a market where a rival can generate a clever new sound in an afternoon, consistency over time is the one advantage the afternoon cannot buy.
The machine now judges distinctiveness too, not just the customer
There is a second reader to satisfy now, and it changes the stakes. Buyers were always the audience for distinctive assets. Increasingly, so are the AI systems that assemble shortlists and answers. When a founder asks an assistant to name the strong brands in a category, the system builds its answer from what it can distinguish. A brand that reads as interchangeable with its category gives the machine nothing to hold on to, and gets folded into a generic description or left out entirely.
This is where distinctiveness stops being a marketing concern and becomes a visibility one. A brand the machine cannot tell apart is a brand the machine cannot recommend by name. The mechanics of how these systems assemble and describe a brand are covered in how to make sure AI describes your brand correctly, and the wider shift in how AI is changing branding. The point here is narrower and sharper: the same homogenising force that Doshi and Hauser measured inside AI output is now also the lens through which AI reads the market. Sameness is punished twice, once by the customer who cannot remember you and once by the model that cannot distinguish you.
Exhibit 2: What to build, what to prompt
The response to AI sameness is not to reject the tools. It is to be deliberate about which layer of brand work they touch. The tools are extraordinary at production and dangerous at definition. The table separates the two.
| Layer | Safe to accelerate with AI | Must be a human ownership decision |
|---|---|---|
| Strategy and positioning | Research synthesis, market scanning | The actual position and the point of view |
| Distinctive assets | Variations, applications, mock-ups | The core assets themselves, chosen to be owned |
| Verbal identity | Draft copy, format adaptation | The voice, the phrases the brand will own |
| Production and rollout | Resizing, versioning, localisation | Nothing. This layer is safe to automate |
The rule that falls out of Exhibit 2 is simple to state and hard to hold: use AI to produce faster, never to decide what is worth producing. The moment the machine chooses the brand's core assets, the brand inherits the machine's central weakness, which is the pull toward the average of everything it has seen.
How distinctiveness compounds across Southeast Asia
Distinctiveness is not a uniform problem across markets, and in Southeast Asia it is unusually acute. A brand here rarely serves one market. It serves Singapore, Malaysia, Indonesia, Vietnam, Thailand, and the Philippines, across at least four working languages and several scripts. Positioning statements, taglines, and value propositions all have to be translated, and translation is where meaning frays and consistency slips. A distinctive visual or sonic asset does not have this problem. A colour is the same colour in Jakarta and Ho Chi Minh City. A shape needs no localisation.
This is why distinctive brand assets are disproportionately valuable to regional brands, and why the fragmentation that stalls so many Southeast Asian companies is partly a distinctiveness failure. When a brand rebuilds its identity market by market, it ends up with six competent local expressions and no single recognisable brand. The deeper version of this problem, and the architecture that solves it, is set out in why Southeast Asian brands struggle to scale. The distinctiveness angle adds one instruction to that analysis: whatever else flexes between markets, the distinctive assets must not. They are the part of the brand that carries recognition across a border that language cannot cross.
There is a regional structural point underneath this. Family enterprises are the backbone of the ASEAN economy, and the EY and University of St. Gallen Global 500 Family Business Index for 2025 found 17 Southeast Asian family businesses among the world's 500 largest, together generating over US$146 billion in revenue. Many of these firms hold decades of accumulated recognition in their home markets that they have never formalised into owned, defensible assets. That latent distinctiveness is an asset waiting to be claimed before a generatively-competent challenger claims the space around it.
When distinctiveness is worth funding, and what it costs
Distinctiveness is worth funding at the moments when the market's memory is about to be set, or reset. That means at founding, at a rebrand, at entry into a new market, and at any point where a category is filling up with competent, interchangeable challengers, which now describes almost every category touched by generative tools. The counter-intuitive timing is this: the best moment to invest in distinctiveness is before it feels urgent, because building recognition is slow and the brands that start early are the ones the market has learned by the time everyone else notices the problem.
Most Singapore branding programmes fall between S$5,000 and S$50,000, with enterprise work higher, and distinctive asset development sits inside that range rather than beside it. Strategy and identity work of this kind is typically eligible for Enterprise Singapore's Enterprise Development Grant, which can co-fund up to 50% of qualifying projects. The more useful reframe is away from cost entirely. The question is not what distinctiveness costs to build. It is what sameness costs to live with, paid every day in the customers who could not remember which brand was yours and the AI answers that never named you.