Somewhere in your shared drive there is a tone-of-voice guide. Someone was paid to write it. It says the brand is bold, human, authentic and expert. There is probably a page on the values behind the words, and a nice pull quote about speaking like a person.
Almost nobody opens it.
Not because the people who wrote it were careless, and not because your team is undisciplined. It goes unread because of what it is. A list of admirable adjectives tells you who you would like to be. It does not tell you what to write. The person drafting a refund email at five o’clock on a Friday does not need to be reminded that the brand is human. They need to know whether to say sorry, how sorry, and whether “we appreciate your patience” is the house line or the thing the house has banned.
The adjective problem
Watch what an adjective actually does when someone sits down to write. It approves almost everything.
A blunt, defensive refund reply can be argued as “bold”. A vague, padded one can be argued as “human”. An email stuffed with jargon can be defended as “expert”. The words are so wide that two people can take opposite drafts to the same guide and both come away vindicated. That is the tell. A guide that cannot reject a bad sentence is not steering anything. It is decoration with a confident font.
This is why the workshop where the guide is born feels so good and changes so little. Choosing the adjectives is pleasant. Everyone agrees the brand should be warm and clear and credible, because who would argue for cold and confusing. The agreement is real, but it is agreement about aspiration, and aspiration is not a decision. The hard part was never deciding to be clear. The hard part is the hundred small choices, made by different people on different days, that add up to whether you actually are.
A guide should resolve arguments, not start them
The test of a tone-of-voice guide is simple. When two people disagree about a sentence, does the guide settle it?
Most guides cannot, because they were written to inspire rather than to adjudicate. A useful one is built the other way round. It is full of decisions someone already made, so the writer at five o’clock does not have to make them again under pressure. Do we use contractions? Yes, always, even in formal notices. Do we say “customers” or “clients”? Do we open a complaint reply with an apology or with the fact? When we have made a mistake, do we explain why it happened or just fix it? When we cannot say yes, how do we say no without sounding like a legal department?
None of those are adjectives. They are rulings. And a guide made of rulings does a thing the adjective guide never can: it removes the argument from the room. The decision was taken once, calmly, by people with the whole brand in view, instead of relitigated every time by whoever happens to be writing.
The most operational guides go further and show the work, side by side, in the words the company actually uses:
- The sentence you might write, and the one the house writes instead, with one line on why.
- The specific phrases you reach for and the ones you have retired, so “we apologise for any inconvenience caused” is on the page with a line through it.
- The constructions you avoid even when they are technically fine, because they make you sound like everyone else.
- The one or two things you do that nobody else in your category does, written down so they survive a change of staff.
Put that in front of a writer and the guide earns its place, because it answers the question they actually have. Put four adjectives in front of them and they will write what they would have written anyway, then reach for the word that justifies it.
Build it from the writing you already do
Here is the part most guides skip. The voice is not waiting to be invented in a workshop. It is already scattered across everything you have published: the support reply a customer thanked you for, the launch post that sounded like your founder, the one explainer that finally made a regulated product feel human.
A real guide is reverse-engineered from that. You gather the writing that worked, you name what it was doing, and you turn that into rules other people can follow. The guide becomes a record of your best instances of yourself, not a wish list assembled from competitors and a mood board. It is shorter, too, because it only contains what is true about you, not everything that is true about good writing in general.
For a fintech this matters more than for most. Your writing carries a regulated product, and the gap between “clear and trustworthy” and “hedged and frightening” is decided sentence by sentence, by people who are not writers, often at speed. For an NGO it is the same pressure in a different shape: the difference between a message that moves a supporter and one that sounds like every other appeal is small, specific and entirely a matter of execution. Adjectives do not survive that pressure. Rules and examples do.
The honest limit
A guide cannot cover every case. That is true, and it is worth saying plainly rather than pretending a document can pre-decide every sentence your organisation will ever write. Some abstraction is unavoidable, because language is open-ended and your writers will meet situations nobody anticipated.
But “it cannot be complete” is not an argument for staying vague. It is an argument about what the guide is for. Its job is not to be a rulebook so exhaustive that judgement is never required. Its job is to make the next decision easier than it would have been without it. A writer who has internalised ten real rulings and twenty before-and-after examples will handle the eleventh, unforeseen case far better than one holding four adjectives, because they have absorbed how the brand thinks, not just how it wants to be described. The examples teach a pattern. The pattern travels to cases the guide never listed.
That is also why length matters. A guide nobody can hold in their head is a guide nobody uses. If it runs to forty pages, the writer at five o’clock will not consult it, and you are back to adjectives by default. The discipline of keeping it short forces the real question: of everything we could say about how we sound, which few rules would change the most sentences? Answer that, and you have a guide. Skip it, and you have a document.
What this is really about
A tone of voice is not a statement of identity you publish once and admire. It is the sum of thousands of small decisions made by people who are not you, when you are not in the room. The only question that matters about your guide is whether it is making those decisions easier or leaving them to chance.
If you read yours honestly and it is a list of things you would like to be, it is not working, however good it looks. The fix is not a better workshop or a longer document. It is to write down the decisions, draw the examples from your own best work, and keep it short enough that the person under pressure will actually use it.
That is most of what we do when we take on a brand: turn the aspiration into rules a whole team can follow without us, then keep them honest as the organisation grows. It is the unglamorous half of communications, and it is the half that decides whether you sound like yourself everywhere a customer meets you. If your guide is being ignored and you want it to do some work, let’s have a conversation.