How Luxury Brands Build Their Unbreakable Tone of Voice
If we look at how luxury brands are typically built, the visual system almost always comes first: the monogram, the colour, the silhouette. Language is treated as decoration wrapped around an image that already does the work. Yet if you take away the logo and the product entirely, hand someone a single sentence pulled from a caption or a press release, and a trained eye can still guess the house behind it. That ability is the result of a communication system built with the same rigour as a print, but expressed through vocabulary, rhythm, and syntax rather than fabric and colour.
This is where most brands stumble. A visual identity can be briefed and produced by an agency in weeks. A voice cannot. It has to behave like a personality rather than a specification; consistent under pressure, and recognisable without a signature (Aaker, 1997). Tone of voice, in this sense, functions less like a style guide and more like a living organism: it adapts to context yet remains unmistakably itself. This matters now more than ever, since AI-generated copy has made competent, on-brief language instantly available to anyone.
Fashion houses that understand this operate at two extremes of the same discipline. The Row constructs its voice almost entirely through omission, treating silence as authority. Balenciaga does the opposite, building identity through rupture and deliberate discomfort. Both succeed for the same reason: neither can be reverse-engineered from a single post, because both are systems, not styles. This article examines how that system is built, from campaign copy to a practical framework for testing whether a brand’s voice holds up once its name is taken away.
The Row and Balenciaga: Two Ends of the Same Discipline
The tone of voice in fashion tends to sit somewhere on a spectrum between whisper and manifesto. At one end, meaning is built through withholding; at the other, through excess. The Row and Balenciaga occupy these opposites almost perfectly, and comparing them shows that a distinctive voice has less to do with volume than with discipline.
The Row’s communication is built on the scarcity of language itself. The brand rarely issues statements, avoids explanatory captions, and lets garments circulate without narrative scaffolding (FashionUnited, 2025). Even its most viral product moments, such as the Margaux bag, were left to spread through organic conversation rather than brand-authored commentary (FashionUnited, 2025). This is not an absence of strategy but a strategy of absence: every unexplained image asks the audience to supply meaning rather than receive it, which is precisely what makes the brand’s rare public statements carry disproportionate weight.
Balenciaga’s voice works through the opposite mechanism: rupture. Its campaign language and visual grammar are built to unsettle expectation, using irony, cultural remix, and calculated discomfort to keep the house at the centre of debate (Latterly.org, 2026). Scholars researching the brand’s 2022 crisis describe this as a deliberate register long before it became a liability, noting that shockvertising is designed to be “startling and offensive by intention rather than accident” (The Conversation, 2025). Provocation, in other words, is not a lapse in Balenciaga’s tone. It is the tone.
What connects these two brands is not proximity in style but consistency in method. Neither voice could survive being copied for a single caption, because neither is really a caption-level choice. Silence reads only as authority when it is sustained everywhere; disruption reads only as intention when it repeats predictably enough to be recognised as a signature rather than an accident. The lesson for any brand watching from the middle of the spectrum is that tone of voice is not about tone at all. It is a rule system, applied without exception.
Language as a Strategic Asset, Not a Supporting Layer
Most brand guidelines still treat language as an afterthought, developing it only after the logo, colour palette, and campaign visuals have already been locked. This order of business reflects a widespread assumption in luxury marketing: that words communicate an identity which visuals have already established. Recent research on brand linguistics challenges this assumption directly, arguing that language does not merely describe brand meaning but actively shapes how consumers identify, experience, and evaluate a brand in the first place (Schmitt, 2017).
This reframing matters because it changes what a copywriter is actually responsible for. Vocabulary, sentence rhythm, and syntax are not stylistic garnish; they are mechanisms that influence recognition and recall in the same way a colour palette does. A brand that consistently favours short, declarative sentences is making a claim about confidence and control, whether or not that claim is stated outright. A brand that writes in long, carefully qualified sentences is making the opposite claim, regardless of what the product itself communicates visually. The discipline that constructs this kind of coherence has existed under its own name for decades: verbal identity, pioneered as a practice distinct from copywriting or advertising language, with its own frameworks, audits, and specialists (Simmons, in Clifton and Blackett, n.d.).
Once language is understood as a structural layer rather than a decorative one, a further principle follows: what a brand chooses not to say is as strategic as what it says. The Row’s near-total silence and Balenciaga’s constant provocation are not just different amounts of communication; they are different linguistic strategies operating on the same underlying logic, where every word, or the deliberate absence of one, functions as a signal that visual identity alone cannot conceptualise.
From Campaign to Caption: Building a Coherent System
A recognisable voice rarely lives in a single document. It is distributed across dozens of smaller, seemingly minor decisions: the phrasing of a press release, the caption under a runway image, the register of a customer service reply, the syntax of a manifesto posted without further explanation. None of these formats carries the brand’s identity on its own. What creates recognition is the fact that all of them are governed by the same underlying rules, even as their tone shifts from one channel to another.
This is where the distinction between voice and tone becomes most significant. Voice is the constant: the personality, vocabulary, and syntactic habits that remain stable regardless of context. Tone is the variable: the way the same personality modulates itself depending on the situation, the platform, or the required emotional register. A brand can sound more formal in a press release and more intimate in a caption without contradicting itself, provided the underlying voice, its values, its vocabulary, and its characteristic rhythm stay traceable underneath both. The failure mode is not variation in tone; it is a voice that disappears entirely once the context changes.
This coherence is not merely an aesthetic preference. Keller’s foundational model of customer-based brand equity argues that a brand’s equity is built from the associations consumers hold in memory, and that those associations are strengthened or weakened by every marketing interaction they encounter (Keller, 1993). A verbal identity that behaves inconsistently across touchpoints does not simply look sloppy; it actively erodes the distinct, favourable associations that constitute brand equity in the first place. Each inconsistent message competes with, rather than reinforces, everything the audience has already learned to expect. In practice, this means a manifesto, a product description, and a one-line Instagram caption should be treated as expressions of a single grammar rather than three separate writing tasks. The words change. The system producing them should not.
The Authenticity Test: Can It Be Recognised Without a Name?
If a brand voice is built rather than merely styled, it should survive a simple test: strip away the logo, the byline, and any visual context, and ask whether the remaining sentence can still be attributed correctly. This is not a hypothetical exercise. It is a method already used in practice: reviewers are given a mix of a brand’s content and its competitors’ with all branding removed, and asked to identify which text belongs to which house (Atom Writer Blog, 2026). A voice that fails this test, one that could have been plausibly written by any competitor in the category, has not actually been thought through. It has only been decorated.
This framework gives the abstract idea of “tone of voice” something closer to an empirical standard. A genuine verbal identity should be identifiable through vocabulary alone, through characteristic sentence length, through the presence or deliberate absence of certain words, and through rhythm that persists whether the text in question is a manifesto or a two-line caption. Brands whose voice has become genuinely distinctive are already discussed in these terms, described as recognisable in a blind test even without any visual cue attached (Vista Social, 2025). The Row would likely pass this test through restraint alone; a fragment of unexplained, unadorned description would already feel unmistakably its own. Balenciaga would likely pass through friction; a sentence built to unsettle rather than reassure carries its own fingerprint regardless of context.
What this test ultimately measures is not creativity but discipline. Imitation is easy to produce at the surface level, a wide sentence here, a knowing irony there, but it becomes far harder to sustain the same restraint or rupture across hundreds of unrelated pieces of copy, written under different pressures, for different purposes, over years rather than weeks. A voice that cannot be counterfeited is therefore not the most original one. It is the one applied with enough consistency that imitation collapses under its own inconsistency the moment it is tested at scale.
Building the System: Where Emerging Brands Should Start
For a brand without a century of archive behind it, none of this is out of reach; it is simply unbuilt. The starting point is not a tagline or a mood board of adjectives, but an audit of existing language: pull every caption, product description, and email sent in the last six months, and ask which words repeat, which sentence lengths dominate, and whether a stranger could sort them from a competitor’s without seeing the logo. Most young brands discover the answer is no, not because their writing is bad, but because it has never been treated as a system with rules, only as a series of one-off decisions made under a deadline.
From there, the useful exercise is not reconstructing a voice from scratch but deciding on a clear set of verbal boundaries, since constraints create distinctiveness faster than aspirations do. A brand that commits to never using superlatives, never explaining a reference, or never posting without a verb in the first sentence has already built a stronger voice than one with a thirty-page style guide full of approved adjectives. This is also where a founder’s personal register, if genuine rather than performed, can substitute for the institutional memory a young house lacks. A standardised way of describing a fabric or a process becomes, with enough repetition, as recognisable as an heirloom monogram.
The final step is treating that voice with the same protectiveness applied to a logo or a print. If a copywriter, an intern, or a generative tool can produce a caption indistinguishable from the brand’s own without reference to a style guide, the guide has failed. Consistency across turnover, across agencies, and across the pressure to sound like whatever performs well that week is what eventually turns a set of preferences into something closer to a personality, and personalities, unlike taglines, are what audiences actually learn to recognise.
Conclusion
A logo can be licensed, a colour palette pulled from a Pantone chart, a silhouette reverse-engineered within a season. None of this requires understanding the brand, only observing it. A voice resists this kind of extraction because it is not a surface-level feature. It is what emerges when values and vocabulary are applied across thousands of small decisions that no single competitor ever sees in full. Copying the words without copying the discipline behind them produces only a plausible facade, not the product itself.
This has particular urgency now. As generative tools make competent, on-brief copy instantly available to anyone, fluency stops being a differentiator. Far fewer brands can sound like themselves under pressure, across every channel, without a name attached. For emerging brands, this is an opening rather than a disadvantage: a voice built with real constraints and enough repetition can pass the same test as a house with a century behind it, because the test was never about age. It was always about discipline.
The question worth asking of any brand is not whether its voice sounds distinctive in a single campaign, but whether it would still be recognisable with the name removed. That test, more than any style guide, is what separates a voice from an imitation.
References
Aaker, J.L. (1997) ‘Dimensions of brand personality’, Journal of Marketing Research, 34(3), pp. 347–356. Available at: https://journals.sagepub.com/doi/10.1177/002224379703400304 .
Atom Writer Blog (2026) How to measure brand voice consistency: KPIs, scoring, and automated frameworks. Available at: https://www.atomwriter.com/blog/measure-brand-voice-consistency/.
FashionUnited (2025) The Row and strategic power of silent marketing. Available at: https://fashionunited.com/news/fashion/the-row-and-strategic-power-of-silent-marketing/2025122969797.
Jones, C.W. (2025) ‘Balenciaga’s controversial new campaign and the long history of “shockvertising”’, The Conversation. Available at: https://theconversation.com/balenciagas-controversial-new-campaign-and-the-long-history-of-shockvertising-195778.
Keller, K.L. (1993) ‘Conceptualizing, measuring, and managing customer-based brand equity’, Journal of Marketing, 57(1), pp. 1–22.
Latterly.org (2026) Balenciaga marketing strategy: provocative campaigns, celebrity hype, and virality. Available at: https://www.latterly.org/balenciaga-marketing-strategy/.
Schmitt, B. (2017) ‘Brand linguistics: A theory-driven framework for the study of language in branding’, Journal of Consumer Psychology, 27(4), pp. 508–531. Available at: https://www.sciencedirect.com/science/article/abs/pii/S0167811617300058 .
Simmons, J. (n.d.) cited in Clifton, R. and Blackett, T. (eds.) Brands and Branding. Available at: http://www.culturaldiplomacy.org/academy/pdf/research/books/nation_branding/Brands_And_Branding_-_Rita_Clifton_And_John_Simmons.pdf.
Vista Social (2025) Brand voice guidelines examples: Why consistency matters. Available at: https://vistasocial.com/insights/brand-voice-guidelines-examples/.
Leave a Reply