Brand Voice System · B2B

A voice system your whole team can write from.

Every time a new writer joins, the voice drifts. Every time a content piece ships, a stakeholder argues about whether it sounds right. A brand voice system fixes both: a 25-page enforceable codex, a named voice test any writer can apply, and a training program so every future hire can sound like you within a week.

Hire me on Upwork See investment
25
Pages, enforceable codex
20+
Voice examples per axis
4w
Build timeline
Top 1%
Upwork Top Rated Plus
Voice codex document
Sliding tone scales
On-voice examples
Off-voice examples
Vocabulary do / do not lists
Writer onboarding kit
Voice codex document
Sliding tone scales
On-voice examples
Off-voice examples
Vocabulary do / do not lists
Writer onboarding kit
Problem
The Problem

Most "brand voice" documents are three slides nobody uses.

A PDF with three adjectives. A slide deck with a tone spectrum. A Notion page last edited 18 months ago. None of it survives contact with a new writer under deadline. Six reasons most voice documents fail:

01 Three adjectives are not a voice. "Confident, warm, clear" describes half the companies in any given category. If your voice doc could be used by your competitor, it is not a voice doc. It is vocabulary.
02 No examples of what off-voice looks like. Writers need to see the line between on-voice and off-voice to hold the line. Without the off-voice examples, every borderline sentence becomes a judgment call.
03 No way to test a finished piece. "Does this sound like us" gets answered by whoever is in the room. Different reviewers pass and fail the same draft. Writers learn to write for the reviewer, not the voice.
04 No vocabulary list. Every brand has 30 to 50 words it uses more than other brands, and 20 to 30 it avoids. Without that list written down, new writers use the default vocabulary of their last job.
05 The voice lives in one person's head. One senior writer or the founder somehow knows what the voice is. They edit every piece. When they leave, the voice leaves with them. The company then hires a new senior writer to re-learn what was never documented.
06 No training program for new writers. Every new hire gets the PDF on day one and has to figure it out from there. By month three the writer has drifted into their own approximation of the voice. Nobody audits the drift because nobody has the tool to audit it.
Codex
The Deliverable

A voice codex your writers open. A voice test your editors run.

The codex sets the rules. The voice test enforces them. Every piece of customer-facing copy passes through the same named checks before it ships, so "does this sound like us" becomes a yes-or-no, not a vibe check.

Acme · Voice Codex · v1.2 Section 02 of 08
02 · Voice on four scales
Formal Casual
Terse Expansive
Cerebral Visceral
Reverent Irreverent
On voice
Your forecast missed because nobody owns the pipeline on Fridays.
Nice dashboards. Where is the number?
We shipped this because Sales asked for it twice on the same call.
Off voice
In today's rapidly evolving landscape, forecasting presents challenges.
Our robust analytics suite surfaces actionable insights.
We are excited to announce this powerful new capability.
Voice test · Running Score · 4 / 5
Sample · Blog post intro
Your forecast keeps missing by 18%. The dashboards look fine. The CRM is tidy. But the number is not robust enough to surface actionable insights when it matters.
Voice test · 5 checks
Specific, not category Named an exact number, not "varies."
Pass
Short sentence cadence Three short sentences in a row. On voice.
Pass
No banned vocabulary "Robust" and "actionable insights" flagged.
Fail
Second person, not first "Your forecast." Reader in focus.
Pass
Irreverent, not reverent "Dashboards look fine" lands dry.
Pass
Codex
Inside The System

Six components. One enforceable voice.

Every component exists because something it replaces was failing. Cut any one of them and the voice starts drifting again within three months.

Component 01

Voice codex (25 pages)

Written definition of the voice on four sliding scales, with named tone positions, cadence rules, sentence-length targets, and paragraph structure patterns.

Component 02

Example library (40+)

Twenty or more on-voice examples and twenty or more off-voice examples pulled from your own existing content, annotated with why they land on which side of the line.

Component 03

Vocabulary lists

A "do" list of 30 to 50 words your brand uses more than other brands, and a "do not" list of 20 to 30 banned words that signal off-voice writing. With reasoning, not just lists.

Component 04

Voice test (5 checks)

A named pass-or-fail test any editor can run on any draft. Five checks, each with a specific diagnostic. Turns "does this sound right" into a yes or no in under 5 minutes.

Component 05

Writer onboarding kit

A one-page quick-reference plus a 60-minute onboarding video for any new writer (internal or contractor). Gets a new hire to 80% on-voice output by end of week one.

Component 06

Channel adjustment sheet

How the voice flexes for LinkedIn vs email vs long-form vs support docs. Same voice, named dial adjustments per channel, so nothing sounds rigid or mechanical.

Process
Process

Four weeks. One working codex.

Interviews first, archive mining second, codex drafting third, writer testing last. The order matters because skipping the archive mining is why most voice systems sound invented.

Week 1
1

Capture

60-minute kickoff plus interviews with founders or content leads. Access to your top-performing content, recorded sales calls, and internal docs. Voice archive assembled.

Week 2
2

Mine & map

Archive mined for on-voice and off-voice examples. Tone scales drafted with specific position points. Vocabulary do and do-not lists extracted from actual usage, not invented.

Week 3
3

Draft & test

Full codex drafted. Voice test built from the codex. Drafts run through the test on 6 to 10 real upcoming content pieces. Fix the edge cases that expose codex gaps.

Week 4
4

Train & hand over

60-minute writer onboarding session with your team and any contractor writers. Full document handover. Access transferred. Voice test adopted into your editorial workflow.

Investment
Investment

Build the system. Keep it sharp.

A single flat-rate build that delivers a working codex and a trained team. An optional monthly retainer for companies whose voice evolves as their audience, product, or category position shifts.

One-time build
Brand Voice System
$6,500 flat

Complete 4-week engagement. Codex, example library, vocabulary lists, voice test, writer onboarding kit, and channel adjustment sheet. Paid 50% on signing, 50% on handover.

25-page voice codex covering scales, cadence, and structure
Example library of 40+ annotated on-voice and off-voice samples
Vocabulary do and do-not lists pulled from your actual content
5-check voice test your editors can run in under 5 minutes
Writer onboarding kit plus 60-minute training session, recorded
Channel adjustment sheet for LinkedIn, email, long-form, support
Full ownership transfer. The system is yours to extend.
Keep it current
Maintenance Retainer
$1,800 / month

Optional after the core build. Keeps the codex current as your voice evolves, new channels are added, and new writers join. Month-to-month, cancel anytime.

Monthly codex review against recent published content
New example additions and vocabulary updates every month
One new-writer onboarding session per month included
Async voice-test reviews for up to 8 pieces per month
Quarterly 60-minute strategy call with the content lead
Cancel anytime with 30 days notice. No lock-in.
Scope
Scope

What the system is. What it is not.

A voice system, not an ongoing content production service. The distinction matters for how we engage and what you can expect.

Included

The system and the handover.

  • Voice capture interviews with founders and primary content leads
  • Mining of your existing content archive for voice patterns
  • Full codex document, example library, and vocabulary lists
  • Testing of the codex on 6 to 10 real upcoming content pieces
  • 5-check voice test with editor guidance and fail diagnostics
  • 60-minute writer training session plus recorded onboarding
  • Channel adjustment sheet for the 4 to 6 channels you actually publish on

Excluded

Named up front to prevent scope creep.

  • Ongoing content production (the system exists for your team to use)
  • Positioning, ICP definition, or core message architecture (Messaging Foundation)
  • Visual brand identity, typography, color, or logo work
  • Rewriting your existing content library against the new codex
  • Voice systems for brands with no existing content archive to mine
  • Multi-voice systems with more than two distinct voices in one codex
  • Localization or translation of the codex into non-English languages
Proof
Selected Work

Recent systems. Real consistency.

Three recent engagements. Full case studies available on request under NDA.

Top 1% on Upwork
Top Rated Plus freelancer
Awarded to freelancers with sustained top client feedback, high earnings, and proven reliability on long-term engagements.
Series B B2B SaaS
RevOps platform (NDA)
0
Writers onboarded to the codex in the first quarter

Marketing team had three in-house writers, two contractors, and a founder who "felt the voice." Codex unified them. First quarter post-handover, editing time per piece dropped from 3 rounds to 1 and the founder stopped being the final reviewer on 80% of content.

Consulting Firm
Management consulting, USA
0
Partner voices documented inside a single codex

Three senior partners with overlapping but distinct voices wrote everything from thought leadership to proposals. Codex captured the shared voice floor plus partner-specific modifiers. Proposals now read as consistent firm voice with individual authorial stamps intact.

Content Agency
B2B content agency, EU
0×
Client voice systems built using the same methodology

Agency licensed the methodology (not the codex content) and used it to build four parallel voice systems for four retainer clients. Their editorial lead now runs voice tests in-house without escalation. The build time per client dropped from 8 weeks to 4.

About

A brand voice is not adjectives. It is a set of rules specific enough that a writer you have never met can apply them on Friday afternoon and be 80% on-voice before you ever see the draft. That is the bar. Anything softer and the voice quietly evaporates.

Specialty Enforceable brand voice systems
Industries SaaS, agencies, professional services
Codex length 25 pages, 40+ examples
Build time 4 weeks end to end
Investment $6,500 build plus optional $1,800/mo
Availability Two system builds per quarter
FAQ
Questions

What clients ask first.

Six questions.

Messaging Foundation covers positioning, ICP, category definition, and message architecture: what the brand says. Brand Voice System covers how it sounds: cadence, vocabulary, sentence structure, tone adjustments per channel. Most companies need both, in that order. Messaging Foundation first, Brand Voice System second. If you already have positioning locked and only need to solve voice drift, start here.
You need a minimum archive for this engagement to work: roughly 20 pieces of existing customer-facing content (blog posts, emails, LinkedIn posts, sales decks, newsletters) plus recordings of at least 3 to 5 founder-led calls or presentations. Without that baseline, the codex would be invented rather than extracted, and invented voice documents are exactly what I am arguing against. If the archive is thin, we can sometimes get there by extending the capture phase with founder-led writing sessions, but that is a scope change I will flag before signing.
80% on-voice by the end of their first week, with the codex plus the 60-minute onboarding. Full on-voice parity usually takes 4 to 6 pieces of output plus one feedback loop against the voice test. That is the realistic ceiling. Beyond that, you are paying for taste, and taste takes years regardless of documentation. The system will not eliminate the need for a senior editor. It will eliminate the need for a senior editor on every sentence.
Two voices, yes, with a shared foundation and voice-specific modifiers. Three or more voices in one codex, no. Past two, the document becomes too heavy to use and writers default to ignoring the modifiers. If you have three distinct senior voices that need documentation, we scope it as three separate voice profiles with a shared vocabulary and argument structure layer. Higher scope, higher investment, longer build timeline.
Not inside this engagement. Rewriting the existing library is a separate scope, usually priced based on word count and priority. What this engagement does is give you the rules and the voice test. Your team or your contractors then run the rules against whatever you want to rewrite, usually prioritizing top-performing pages and high-traffic sales touchpoints first. If you want me to execute rewrites, that is my content writing or landing page services.
No. It catches the most common failure modes (category language, banned vocabulary, wrong cadence, wrong person, wrong tone position) in under 5 minutes of editor time per draft. It does not catch subtle structural issues like "the argument is weak" or "the opening line does not earn attention." Those require senior judgment, which no test can replace. The voice test is designed to catch the 80% of failures that waste editor time, not the 20% that require real thinking.

Your voice should be written down, not hoped for.

Tell me your primary voice, your current team size, and whether you have an existing content archive to mine. You will get a scope confirmation and a kickoff date within two business days.

Hire me on Upwork See my portfolio
Scroll to Top