Names are a design decision
Naming tends to happen late. By the time a product, service, or programme gets to the point of needing a name, there are usually a dozen more pressing things on the critical path. Naming gets squeezed into a meeting that’s already running over, or delegated to whoever happens to have an opinion. Someone suggests something. Someone else doesn’t like it. A compromise is reached, usually in the direction of blandness or bureaucratic safety. The name ships. Nobody is quite satisfied. Nobody says anything, because the next sprint has already started.
This is how a lot of things end up called what they’re called — not through considered judgment, but through exhaustion and the need to move on. And it matters more than the circumstances of its creation usually suggest.
A name is not just a label. It’s a signal — and typically a figurative one rather than a literal one. It conveys how an organisation thinks and what it values. It sets expectations before the product or service has had any opportunity to do anything. It opens the conversation with a user and persists through every subsequent interaction. A name that signals the wrong thing, or signals nothing in particular, creates a deficit that everything downstream has to work against. Marketing, customer service, onboarding — all of it starts from behind.
In public services especially, naming carries unusual weight. Services that people have to use — for benefits, permissions, healthcare appointments, council tax payments — carry names that can feel either human or bureaucratic, welcoming or hostile, legible or opaque. “Universal Credit” says something about values and intentions before any transaction has occurred. So does “GOV.UK”. So does almost everything that carries a name into a civic or institutional context. Names are decisions, even when they’re treated as if they’re not.
The difficulty with naming is that it sits at the intersection of several things that don’t naturally cooperate: creative judgment, legal availability, cross-cultural linguistic check, organisational politics, and the need to avoid whatever trend happened to peak eighteen months ago. An invented name protects its trademark but surrenders immediate meaning. A descriptive name communicates its function but offers no distinction and dates quickly. A location-based or founder-based name is legible but limiting. Every option involves a trade-off, and trade-offs require criteria, and criteria require prior agreement about what the name is actually supposed to do.
This is the step most naming conversations skip entirely. Before generating options, it’s worth establishing whether the name is primarily describing a function, signalling a value, establishing a family relationship with other products, or marking a deliberate departure from an existing identity. These are not the same goal, and a name that succeeds at one will often fail at another. A practitioner handbook on naming — one of the more candid accounts of what the process actually involves — sets out four factors that consistently matter in evaluation: semantic value (does it convey the right idea or attitude?), strategic fit (does it align with what the organisation is trying to become?), phonetic structure (is it easy to say, spell, and remember?), and legal availability. Of these, semantic and strategic are hardest to agree on — because they require agreement about intent, which is really a question about organisational identity, one level up from naming.
The political dimension of naming is consistently underestimated in design processes that are otherwise quite good at managing politics. Naming projects fail not usually because the creative work was poor, but because the conviction to defend a good name wasn’t there when it met resistance. Consignia was not a beautiful name, but it also never had a fair chance — launched without adequate explanation of what it was for, it collapsed the moment the press objected. Centrica, equally invented and equally abstract, survived because the reasoning was clear and the commitment held. What names need, above a certain threshold of quality, is not perfection. It’s backing — and backing requires that the people doing the backing understand what the name is doing and why.
Design can help with naming, not by making it feel more creative, but by applying the same discipline it brings to other problems: clarify the problem before generating solutions, test options in context, make the criteria explicit, separate the subjective response from the strategic evaluation. The conversation will still be political — naming always is. It will at least be better structured.
The name is the first thing most people encounter. It will be the last thing changed. Treating it as a branding problem to be solved downstream is how organisations end up with names that nobody defends because nobody chose them, for reasons that nobody quite remembers, at a moment when the more important work had already moved on.