Words That Build, Words That Break
"We're sorry if you feel that way."
Five words. Zero accountability. The customer service equivalent of a shrug wrapped in professional courtesy. This is how brands lose customers they spent thousands acquiring—not through product failures or price increases, but through language that says: your problem is your perception.
Every brand communicates. Few are trusted.
The difference shows up in the support email someone reads at 2 a.m., the error message during checkout, the CEO's crisis statement, the tooltip in your product interface.
According to Edelman's 2024 Trust Barometer, consumers now weigh trust as heavily as product quality when deciding to buy. Language delivers that trust—or destroys it.
Yet most brands still treat language as decoration rather than infrastructure.
This article is for marketing leaders who understand that trust is emotional before it's rational. Here's how to audit, refine, and systemise the words your brand uses so they consistently signal credibility, respect, and emotional safety.
Trust is a feeling before it becomes a metric.
It's the subconscious calculation: "Does this brand see me as a person or a transaction? Will they protect me when something goes wrong? Do their words match what they actually do?"
Words carry that feeling. They're signals of intention, competence, and respect. Two versions of the same message:
Version A: "We apologise for any inconvenience this may have caused."
Version B: "We messed this up, and I'm sorry. Here's exactly what we're doing to fix it for you right now."
The rational information is nearly identical. The emotional impact is entirely different.
Version A diffuses responsibility. Version B owns the mistake and signals immediate agency. One phrasing shift changes whether someone feels managed or genuinely cared for.
Certain patterns consistently increase perceived trustworthiness. Not because they're stylistically appealing, but because they reduce cognitive friction and signal authenticity.
Take clarity. Ambiguity breeds anxiety. Transparency reduces it. Compare "Leveraging cutting-edge synergistic solutions to drive paradigm-shifting outcomes" with "Here's how we'll help you save 3–4 hours a week." The first sentence requires mental translation. The second delivers information directly.
Specificity works the same way. Concrete details feel honest because they can be verified. "Results in weeks" is inherently suspect because it can't be disproven. "Most customers see measurable improvement within 21–45 days" creates accountability, which builds credibility.
Then there's ownership. Active voice plus "we" signals accountability. Passive voice signals evasion. "Unfortunately, a processing error occurred" versus "We dropped the ball on your order. Here's the fix, plus a credit for next time."
When Satya Nadella became Microsoft's CEO, a key cultural change was replacing defensive language with ownership language. Memos shifted from "Issues were identified in the deployment process" to "We shipped this too fast and broke things for customers," boosting employee trust in leadership by 27% in six months. The language change led the cultural shift.
The empathy question is tricky. Saying "We're sorry if you feel upset" implies blame, as if upset is unreasonable. In contrast, "This sucks, and it's our fault. I'd be frustrated too. Let me make it right" shows genuine empathy—acknowledging feelings to address the issue, not just ending the conversation.
Jargon creates distance; human phrasing builds closeness. When Domino's launched their "Pizza Turnaround" campaign in 2010, they didn't say "We're enhancing flavour profiles through optimised ingredient matrices." They said, "Our pizza was crap. We fixed it." Revenue increased 14% that quarter.
These patterns consistently erode trust, though most brands don't notice the damage until customers stop returning.
Vague, rehearsed commitments top the list. "Your call is important to us," while you wait 47 minutes. "We take your privacy very seriously" on a site selling data to 39 third parties. These phrases are so common they've become transparent. They're what brands say when they don't want to promise anything specific.
Over-hedging does similar damage. "Save up to 50%" means "probably save nothing." "Results may vary" means "results probably won't happen for you." Appropriate hedging shows intellectual honesty. Excessive hedging shows a lack of confidence.
There's that performative empathy again. "We're sorry you feel that way." "We regret that you're disappointed." "We apologise for any inconvenience caused." All three shift blame back to the customer's emotional response instead of owning the action that caused it.
Manipulative urgency teaches customers to distrust everything else you say. "Last chance!" for the 47th time this month. "Limited stock—127 people viewing right now!" when there are thousands available. These tactics trigger an immediate response by exploiting anxiety, but once someone realises they've been artificially pressured, scepticism becomes permanent.
Distance-creating language, like third-person, passive, bureaucratic phrasing, signals you're dealing with a system, not a person. For example, United's 2017 statement about "re-accommodating" a forcibly removed passenger sounded formal but dehumanising, exemplifying how corporate language can betray values. In contrast, using "you" and "we" creates a more human connection: "We'll find you a seat on the next flight and cover your hotel tonight."
People trust what feels familiar and human, not engineered.
Convoluted, hedged, or overly polished language adds mental work, seeming like hiding. Clear, direct, imperfect language—using contractions, first-person, and colloquialisms—feels more authentic and trustworthy.
Tone mirroring matters more than most brands realise. Customer writes in lowercase with emojis? Match the energy within reason. Customer formal? Mirror that formality. Linguistic alignment creates a connection faster than any other variable.
Emotional congruence matters more than perfect grammar. Language must match the emotional reality of the situation. If someone is frustrated, cheerful corporate-speak feels dismissive: "Thanks for reaching out! We're so glad you contacted us! 😊" If someone is celebrating a milestone, formal bureaucratic language feels cold: "Your request has been processed. Reference ID: 847392."
A small typo in a run-on sentence that feels human works better than flawless corporate prose that feels robotic.
Treat this like a security audit, not a copy edit.
Start by collecting 50–100 customer-facing messages like support emails, error messages, sales scripts, UX copy, leadership posts, and terms updates—everything customers encounter. Score each on a 1–5 trust pattern scale: Clarity, Ownership, Specificity, Emotional tone, Action steps. Messages below 3.5 need urgent review.
Map emotional friction points. Where do customers feel confused, pressured, dismissed, or sold to? Review the language used at those specific moments. Does it clarify and reassure, or compound anxiety?
Here's a practical test: Pull one message from each channel—marketing copy (homepage or ad), support communication (email or chat transcript), legal/compliance (terms, privacy policy), executive communication (CEO post or internal memo). Read them back-to-back. Do they sound like the same company?
Most brands discover massive inconsistency. Legal sounds threatening, marketing sounds hyped, support sounds robotic, executives sound like they're reading teleprompter copy.
Most brand voice guidelines list adjectives like "fun, friendly, approachable"—just decoration. True voice as behaviour considers how we communicate when things go wrong, discuss limitations, handle disagreement, and acknowledge mistakes.
Guidelines built around behavioural principles work differently. We acknowledge problems before explaining solutions. We give specific timelines, not vague promises. We explain why when we say no. We use "I" and "we" rather than institutional passive voice. We match emotional tone to the situation, not corporate defaults.
Individual word choices matter. Sustainable trust requires systematic change.
Most organisations have fragmented language with different voices across departments like marketing, support, product, and leadership. Implementing cross-functional principles ensures clarity over cleverness, with engineers avoiding jargon, marketers steer clear of hyperbole, support refrains from canned scripts, and executives avoid buzzwords. It emphasises specificity through concrete details such as numbers, timelines, and examples. Ownership is prioritised over deflection by using active voice, e.g., "we made a mistake" instead of "mistakes were made. " Empathy is valued over efficiency by optimising for appropriate responses rather than the shortest reply.
Language needs ownership. CMO or Brand Lead sets principles and owns external consistency. CX or Support Lead owns high-stakes emotional touchpoints. Product and UX own in-app microcopy and error states—legal reviews for compliance without watering down ownership language. The CEO models trust language publicly—this cannot be delegated.
Internal culture language has to align with external messaging. If leadership speaks to employees using vague corporate-speak, defensive language, or manipulative framing, that culture leaks into customer-facing communication. The language in all-hands meetings, performance reviews, Slack channels—all of it shapes the linguistic culture that eventually reaches customers. If you want external messaging that sounds human, internal communication must sound human first.
Making trust language a quarterly review item converts aspiration into operational discipline. Pull 20 random customer interactions. Score them against your rubric. Identify pattern failures. Update guidelines. Retrain teams on specific gaps.
Employees mirror leadership. When the CEO uses defensive language, teams follow suit. Clear, accountable communication from the CEO models the same for employees, investors, and customers, building trust. Avoid performative code-switching or different truths for different audiences.
Building trust through language isn't a project with a completion date. It's an ongoing discipline.
It requires questioning every word choice: Does this create clarity or confusion? Does this demonstrate respect or condescension? Does this sound like a real person, or institutional performance?
Trust is fragile. One poorly worded crisis response can undo years of careful messaging. One dismissive support email can override positive brand experiences.
But language is also one of the few trust elements entirely within your control. You can't control market perception, competitive dynamics, or external events. You can control whether your words build or break trust.
The brands that understand this—that treat every phrase as a trust signal, every sentence as a credibility decision—build communication systems that compound trust rather than erode it.
The question isn't whether your brand communicates. Every brand does.
The question is whether your communication consistently signals that you understand your audience, respect their intelligence, and will act in their interest.
Trust isn't built through what you say about yourself. It's built through how you say everything else.
