Never Say List

A never say list is the bouncer at the door of your brand voice. It names the exact words and phrases that never get published no matter how accurate they might be. Each banned term comes with its replacement so writers and AI generators know precisely what to say instead. This tool sits at the heart of any operational voice system because it turns gut feelings about corporate sludge into hard rules. In 2026 most brands still rely on vague advice like avoid jargon. The never say list replaces that with entries like never say utilize always say use. It works for human writers who need reminders and for AI that needs explicit training data. The article on brand voice shows how Mailchimp Discord Headspace Innocent and Hey each maintain their distinct personalities by knowing exactly what they refuse to sound like. Without it even the best voice attributes get diluted by the defaults that writers and machines fall back on when under pressure from deadlines or legal teams. The list becomes training data that makes your voice stick across every surface from marketing headlines to support documentation.

It is not a style guide full of positive suggestions or a loose set of preferences that writers can interpret. Never say lists do not read like here are some ideas or try to avoid these when you can. They are blunt instruments. This is not the place for corporate speak or hedging about best practices in brand voice. If your list says things like minimize jargon it fails the test because it still requires the reader to decide what counts. Real never say lists name the exact poison and the exact antidote. It is also not a replacement for the other layers of your voice system. Voice attributes define the positive direction like opinionated but not combative. The tone spectrum shows how that voice flexes in an error state versus a social post. Applied examples give side by side rewrites of real sentences from your own site. The never say list simply guards against the things that undermine those positives. It is not a one time exercise from a branding workshop that gathers dust in a shared drive. Brands that treat it that way watch their voice slowly drift as new writers introduce fresh corporate habits and AI tools fill in the gaps with generic output. Innocent never allows serious defensive or corporate language because that would break the playful honest contract they have with their customers on every juice carton and every web page.

Concrete examples prove why this tool outperforms vague voice documents. Mailchimp built their reputation on a never say list that banned pressure tactics and formal jargon back in 2015. They refused to say things like maximize your ROI or leverage our industry leading solution. The replacement was always direct useful language that treated the user as competent. The wrong sentence We help you leverage our powerful tools to maximize your email marketing ROI became the right sentence Send better email. That three word punchline captured their warm pragmatic archetype perfectly and the never say list protected it for over a decade. Hey takes it further by making their list a positioning weapon. In their 2024 voice JSON they banned neutral hedged and deferential phrases. Never say perhaps or in our humble opinion. Always say this is broken or we fixed it. Their concrete rewrite took a bland corporate version Our innovative platform provides seamless integration for your email needs and delivers exceptional value and changed it to Most email is a mess. We made one that is not. No spying. No algorithms. Just the stuff you need. This example got fed to their AI tools and kept every generated response sharp and opinionated. For Headspace the never say list protects the calm deliberate experience. They ban hype words like unlock your potential achieve inner peace or transform your life. Those create urgency that fights the product. Instead they use be kind to your mind or lets take a breath. One error state example from their app in 2025 showed the difference. The wrong version Your session has timed out. Please restart your mindfulness journey became Your session ended. Take a moment whenever you are ready. Discord uses theirs to stay casual inclusive without sliding into sloppy or exclusive slang. Banning corporate announcements and formal instructions keeps every notification sounding like it came from a friend. Innocent lists serious corporate defensive language as banned. They replace any defensive response with radical honesty wrapped in a joke. A customer complaint response that might have said We apologize for any inconvenience caused by our supply chain disruption becomes Sorry the van broke down. Here is a free bottle for the hassle. The never say list made sure their playful personality stayed intact even in tough moments. These examples from real brands in specific years show the never say list is not abstract. It is the tool that makes voice scalable across teams and AI systems.

Use the never say list whenever you bring a new person onto the team or hook up a new AI generator. Drop the full list into your system prompt before the first output. Reference it during copy QA to catch drift before it ships. It is most critical for the surfaces that usually get ignored by brand teams like product UI strings error messages and transactional emails. Those are written by engineers or ops teams who default to cautious corporate language. The list gives them clear boundaries without requiring a full voice workshop. Run an audit by collecting 20 live pieces of copy and checking each against the list. You will spot the slow accumulations that make your brand start to feel off. Paste the JSON version straight into Claude projects or ChatGPT custom instructions right next to your voice attributes and example rewrites.

Do not bother with a never say list if your brand voice is supposed to sound like a traditional corporation or law firm. Some terms that read as sludge to a startup are table stakes for trust in certain industries. Avoid it if you have no one assigned to enforce the rules in actual reviews. A list that no one checks becomes theater. Keep the list under 25 items so it stays usable. Do not revise it every month based on trends or it loses its role as a stable reference. Update once a year when you review live copy and add any new intruders that have appeared.

A never say list turns your voice from an accident into a deliberate system that survives writers AI tools and time.