Voice follows positioning
Positioning decides what the brand promises, to whom, and against which alternatives. Voice is how that promise sounds when it is said out loud, one post at a time, to people who did not ask to hear it. A brand positioned on expertise sounds calm and specific; a brand positioned as the friendly alternative in a stuffy category sounds warm and plain; a brand positioned on speed sounds brisk. The voice is not chosen from a menu of personalities. It is derived from the position, which is why a voice exercise that starts with adjectives and ends with "fun, authentic, bold" produces the same voice for every brand.
The distinction between voice and tone matters. Voice is constant: the same brand speaks in a product launch and in a reply to a complaint. Tone flexes with the situation: brisk in the launch, patient in the complaint. A voice guide fixes the first and gives rules for the second.
Voice also has a commercial job. In a feed where most content is skimmed, a recognisable voice is a reason to stop, and stopping is what the ranking rewards. A brand whose posts could have come from anyone is competing on the strength of each individual post; a brand with a voice that people recognise has ReachUnique people who saw an ad. working for it before the post is read, because familiarity earns the moment of attention that decides whether a post is seen at all.
What this guide does not do is set the position. If the brand cannot say in a sentence what it promises and to whom, that sentence has to be written first, and the positioning guide is where it happens. A voice built on an unsettled position changes every time the business changes its mind.
Voice principles and examples
A voice guide works when it is specific enough to settle an argument about a draft. Adjectives alone do not do that; a principle with a "we are, we are not" pair and an example line does.
| Principle | We are | We are not | Example line |
|---|---|---|---|
| Plain | Direct, concrete, short words where they do the job | Jargon, hedging, corporate passive voice | "The new frame is lighter and it costs the same." |
| Warm | Interested in the person, generous with credit, quick to thank | Gushing, over-familiar, full of exclamation marks | "Thanks for the photo, that is a serious climb for a first ride." |
| Expert | Confident about what we know, honest about what we do not | Lecturing, condescending, vague where we should be precise | "We have not tested it in freezing conditions yet, so we will not claim it." |
| Dry | Lightly funny when the moment allows it | Jokes about customers, trends for their own sake, sarcasm in replies | "Yes, it comes in black. Everything we make comes in black." |
The four principles above are a worked example for an imagined brand, not a template to copy. A professional services firm positioned on rigour might replace Dry with Measured; a children's brand might replace Expert with Playful. Three or four principles are enough; a guide with ten principles is a guide nobody can hold in their head while writing a reply.
Deriving the principles is a short exercise. Take the positioning statement and ask how each clause would sound. Collect posts and replies the brand has already published that felt right, and ones that felt wrong, and name what made the difference. Write the "we are not" column first, because it is easier to recognise what the brand must avoid than to describe what it is. Then write one example line per principle in the brand's own words, drawn from a real situation. The example lines carry more of the guide's value than the principles, because a writer under time pressure copies an example and ignores an abstraction.
The content-strategy framework offers a check at this point: the subjects the brand speaks about should match the pillars it owns across its content, so that the voice is exercised on subjects where the brand has standing, and silence is the right answer on everything else.
Tone by situation
Tone is the voice adjusted for what is happening. The guide should name the recurring situations and say how the tone moves for each, because the moments when tone matters most are the ones with the least time to think.
A launch or announcement. Brisk and specific. Lead with what has changed and what it does for the person reading; keep the brand's excitement in the facts rather than in adjectives.
A question. Helpful and complete. Answer the question asked, in the reply, without sending the person to a link unless the answer is genuinely too long for the format. Credit the questioner where it fits.
A complaint. Patient, unhurried and accountable. Acknowledge the specific problem, say what happens next and by when, and move to a private channel once the public reply has shown other readers that the brand responds. Never argue the facts in public; never delete the complaint.
Bad news from the brand. Plain and early. An outage, a recall, a delay or a price change is told in the brand's own words before the audience tells it in theirs, with what the brand is doing about it and without blaming suppliers or customers.
A crisis in the wider world. Quiet. Pause scheduled posts, because a cheerful product post beside grave news reads as indifference, and say nothing unless the brand has something specific and useful to add. Holding statements for these moments are best drafted in these tones before they are needed, as part of a social media crisis plan.
A trend or moment. Only where the brand's voice fits it. A trend is an invitation to say something in the brand's voice using the trend's format; it is not an instruction to adopt the trend's voice. The TikTok guide on posting cadence and trends covers how trends work on the platform where they matter most.
Praise. Warm and short. Thank the person, resist the urge to turn it into a sales message, and ask permission before reusing it.
Response and escalation rules
Replies are where voice is tested, because they are written fast, by whoever is on the inbox, to someone who may be annoyed. The guide fixes a few rules so that the voice survives.
A response window. State it in words the team can keep: within the working day for questions and complaints, faster for anything that mentions safety, a named person or a legal matter. The window applies to the first response, which can be an acknowledgement, not to the resolution. Publish the window in the profile where the platform allows it, so that expectations are set.
Who replies. One person owns the inbox at any time, with cover named for absences. Replies in the brand's name use the brand's voice; replies from named staff accounts use their own voice within the brand's principles.
What escalates, and to whom. Anything legal, anything about safety or a recall, anything from a journalist, anything that mentions a data problem or a named employee, and any conversation that has gone round twice without resolution. The guide names the person for each, and the rule is that escalation happens before the second reply, not after the tenth. Written down with a name and cover for each, this list is the inbox's escalation routeA short written route naming who front-line staff contact, how and how fast, when a problem is beyond what they can decide alone..
Templates and judgement. Saved replies are fine for the questions that arrive daily, provided they are written in the voice and edited for the specific case. A reply that could have been sent to anyone tells the reader that nobody read their message. The tutorial on building a library of saved replies writes them from contact reasons, marks the parts every agent must personalise and gives each reply an owner and a review date.
Taking it private. Move to messages once the public reply has acknowledged the issue, and say so in public, so that other readers see the brand engaging rather than disappearing.
What never happens. Criticism is not deleted, arguments are not won in public, and nobody on the inbox makes a promise the business cannot keep. Moderation of abuse, spam and off-topic content follows a written policy that the community management guide sets out in full; this page fixes only the rules that protect the voice.
Cadence by platform and capacity
Cadence is how often the brand appears, and the right cadence is the one the team can keep in its busiest month. A rhythm set for an ideal team is abandoned within weeks, and an abandoned rhythm costs more than a modest one, because the ranking learns inconsistency and the audience learns absence.
| Platform role | Rhythm to aim for | What fills it | What the inbox needs |
|---|---|---|---|
| Primary | Several posts a week in the native format, at a fixed rhythm the audience can learn | Content made for the platform; the best of the brand's pillars; replies and conversation as content in their own right | Checked through the working day; response window published |
| Secondary | A few posts a week or fewer, adapted from the primary | Repurposed primary content re-cut for the format; occasional native pieces for the platform's particular audience | Checked on a schedule, at least daily on working days |
| Listening | Nothing posted | Profile pointing to the active accounts; a pinned note | Mentions monitored; messages redirected with a saved reply |
| Seasonal burst | A temporary increase around a launch, event or season, planned and ended on a date | Pre-produced content batched before the burst begins | Extra cover arranged for the burst |
Platforms differ in what cadence they reward, and the differences are documented, where they are documented at all, in the platform roots: how often a short-video account should post, for example, is a question for the TikTok posting guide, and the professional-network answer is different. The cross-platform principle is that regularity beats volume. Appearing on the same days each week, so that both the audience and the ranking can learn the pattern, matters more than the count. FrequencyAverage times a person saw an ad. in the paid sense, how many times one person sees the brand, is bought deliberately with ads; organic cadence sets a floor for it among followers and nothing more.
The capacity test is straightforward. Write down the hours available per week for production, replies and reporting, with names beside them. Subtract the inbox time first, because it is not optional. Divide what remains by the honest production time per piece, including review, and that is the cadence. If it is lower than the strategy wanted, either the strategy accepts it or the capacity changes; the cadence does not get rounded up. Batching helps: a half-day of production every fortnight produces more consistent output than an hour snatched daily, and it is how a small team sustains a primary platform. Choosing fewer platforms, which the choosing platforms guide argues for, is the other half of the same decision.
A cadence should be reviewed against its KPIA metric tied directly to an objective. rather than against the feeling that the brand is not posting enough. If reach and engagement per post hold as cadence rises, the rhythm can rise; if they fall, the extra posts are diluting the account and the rhythm should come back down.
A gym chain sets its cadence
A regional gym chain with 6 sites and one marketing coordinator sets cadence for the year. The coordinator has 12 hours a week for social, of which 4 go to replies and messages across both accounts and 1 to reporting. The remaining 7 hours fund production. A short vertical video from a site takes about 90 minutes including editing, and a static post with a trainer's tip takes 30 minutes. The primary account is set at 3 videos and 2 static posts a week, roughly 5.5 hours. The secondary professional-network account, aimed at corporate membership buyers, gets 1 adapted post a week. In January, the busiest month, the plan allows for a 3-week burst at double the video cadence, produced in a batched filming day in early December, and the burst ends on a fixed date rather than whenever the coordinator runs out of material.
The voice guide
Everything above ends up in one document, short enough to be read in a sitting and specific enough to settle a disagreement about a draft. A usable voice guide contains the positioning statement it derives from; the voice principles with their we-are, we-are-not pairs and example lines; the tone-by-situation rules; the response window, ownership and escalation contacts; the cadence by platform role; a short list of words and phrases the brand uses and avoids; and a handful of real posts and replies that show the voice at its best, annotated with why they work.
The guide has an owner, usually the person who owns the social strategy, and it is a living document in a modest sense: it is reviewed each quarter alongside the strategy, updated when positioning changes, and added to when a new situation produces a reply worth keeping as an example. It is the first thing a new team member, agency or freelancer reads, and it is the reference for the review of drafts, so that feedback says "this breaks the Plain principle" rather than "I don't like it". Where drafts start with an AI tool, the same principles and example lines become the core of a voice pack for AI drafting, tested against real briefs. The social voice guide extends the voice and tone sections of the brand's own guidelines, which apply the same principles to every channel alongside the logo, colour, type and imagery rules. The voice principles and the cadence by platform role are then summarised beside the other social decisions in the organic social media strategy template.
A quarterly check keeps the guide honest: read the last month of posts and replies against the principles and note where the account drifted. Drift is normal, especially when several people post, and the check is what brings the voice back without anyone feeling blamed.
Between those quarterly checks, the social post publishing checklist asks for the tone of each post to be read against the guide before it is published.
Common mistakes
- Starting from adjectives. "Fun, authentic, bold" describes nothing and fits every brand; derive the voice from the position instead.
- One voice for all platforms, unadjusted. The voice is constant, but a professional feed and a short-video feed pull it in different directions; the guide says how far it may flex.
- Letting tone follow mood. A complaint answered curtly on a bad day costs more than the complaint did.
- Cadence set for an imaginary team. Count the inbox hours first and set the rhythm from what is left.
- Volume over regularity. Posting daily for a fortnight and then vanishing teaches both the audience and the ranking the wrong thing.
- No escalation path. The most damaging replies are written by someone who did not know whom to ask.
- A guide nobody reads. Keep it short, keep the examples real, and review it with the people who post.
