Social media brand guidelines cover five things: profile setup, voice per platform, approved visuals and templates, rules for replies and comments, and who approves a post before it goes out. Most brands write down the logo and the colors, then leave everyone posting on instinct.
The gap shows up fast. One account bio says "we," another says "I." A product shot gets cropped wrong for a story and nobody notices until it is live. A reply to a complaint reads nothing like the brand that wrote the caption above it. None of this is a design failure. It is a guidelines gap, and it is specific enough to close.
Below is what belongs in each section of a social media brand guide, with the parts most teams skip and why those parts matter most once AI tools are drafting captions and replies alongside people.
Key Takeaways
- Social media brand guidelines need five sections: profile setup, voice per platform, visuals, reply rules, and approval.
- Voice should flex by platform within one set of rules: the words the brand avoids stay fixed, the register can loosen.
- Reply and comment guidance is the section most brand books skip, and the one that gets tested the most often.
- A review date and a named owner matter as much for social guidelines as for the logo file, because platform specs and campaign details change often.
- TrueKeep keeps the voice rules and approved components in one portal and serves them to Claude, ChatGPT, and Cursor through an MCP connection, so a caption drafted by AI still reads like the brand.
What social media brand guidelines are
Social media brand guidelines are the rules that keep a brand recognizable across every account it runs, from the bio line to the reply in a comment thread. They sit downstream of the main brand guidelines template: the logo, the colors, and the voice rules already exist, and social guidelines say how those rules apply to a square image, a 280-character caption, or a reply typed in under a minute.
A brand style guide written for print and web rarely answers the questions a social manager has at 9am: what hashtags to use, how to crop a photo for a story, whether to reply to a joke with a joke. Social guidelines answer those, specifically enough that a new hire or a contractor can follow them without asking.
Profile and format specs every account needs
Every account the brand runs needs the same baseline, written down once and applied to each platform:
- Handle and display name, consistent across accounts where the platform allows it.
- Bio, with the exact wording or a template with one variable (a city, a product line).
- Profile and cover images, at the file size each platform requires, cropped to show the logo clearly at thumbnail size.
- Link in bio, pointed at a page that matches the current campaign, not the home page by default.
- Pronoun and voice for the account, "we" or a named persona, chosen once and never mixed.
These specs are the least interesting part of the guide and the part that breaks first. A new platform launches, a social manager signs up in ten minutes to claim the handle, and the bio gets written from memory instead of from the guide. Naming the exact wording, not a description of it, is what prevents that.
Voice across platforms: one brand, many registers
The brand's voice rules do not change by platform. The register they get expressed in does. A brand that is "plain and direct with customers" can sound plain and direct in a LinkedIn post and in a reply to a comment, but the LinkedIn post runs longer and the reply runs shorter.
A useful social voice section names what stays fixed and what flexes:
Fixed, on every platform: the words we avoid, the claims we never make, how we talk about competitors.
Flexes by platform: sentence length, how many questions we ask, whether we use emoji.
LinkedIn: no emoji, full sentences, one idea per post.
Instagram captions: short sentences, one emoji at most, a question in about half of posts.
Replies: shorter than the comment we are answering, never longer.
This is the same information a brand voice guidelines page covers for the rest of the brand, narrowed to what a caption or a reply needs. The Mailchimp content style guide is a public example written at this level of detail, down to how it handles tone in support replies.
Visual specs: images, crops, and approved components
Social images fail for the same reason logos do: someone resizes a file by eye instead of from a spec. A social visuals section needs the exact dimensions for each platform's profile image, cover image, in-feed post, and story, plus the safe zone where text and logos should sit so a platform's own UI does not cover them.
Approved templates belong here too. If the brand has a social post template, a story template, or a quote-card template, the guide names the file and shows one filled-in example next to one rejected example, the same way the main brand guidelines handle logo and color specs. Alt text belongs in this section as well: the W3C's guidance on writing for accessibility applies directly to a social caption, and a one-line rule ("describe the image, do not repeat the caption") settles most of the guesswork.
Rules for replies, comments, and the moment something goes wrong
This is the section most brand books skip, and the one that gets used the most. A caption gets written once and reviewed before it posts. A reply gets written in seconds, often by whoever is logged in, with no time to check a style guide.
A reply section needs:
- A tone range, not a script: how warm, how brief, how much humor is allowed.
- A short list of situations that escalate, a complaint about the product, a legal question, anything about a person by name, so one person decides instead of whoever is online.
- One line on what never gets said, admitting fault before legal reviews, speaking for a competitor, correcting a customer in public.
None of this replaces judgment. It gives the person replying a rule to apply in the five seconds they have, instead of a blank page.
Who approves a post, and who keeps the guide current
Name one owner for the social guidelines, the same way the rest of the brand guide does. That person approves changes to the voice rules and the templates, and someone, often the same person, approves posts before they go out on at least the accounts that carry the most risk.
Set a review date. Platform specs change, image sizes change, and a campaign detail in a bio link goes stale within weeks, not years. A quarterly check against the live accounts catches most of it.
Make the guidelines reach the person actually posting
A social guidelines document that lives in a shared drive helps the person who already knows where to look. It does nothing for the new hire covering a launch, the agency posting on the brand's behalf, or the AI tool a social manager now asks to draft a first pass of a caption.
TrueKeep holds the voice rules, the approved templates, and the image specs in one portal, set up in four parts: strategy, guide, assets, and components. Connected to Claude, ChatGPT, or Cursor through an MCP connection, the AI reads those rules before it drafts a caption or a reply, and a final pass checks the output against the brand before anyone posts it. The social manager still writes the judgment calls. The guidelines just reach the draft before a person does.
FAQ
What should social media brand guidelines include? Five sections: profile and format specs, voice per platform, approved visuals and templates, rules for replies and comments, and who owns approval. Each one should end in something specific, a file size, an exact bio, a list of escalation situations, not a description of a feeling.
How is a social media style guide different from a brand style guide? A brand style guide covers the logo, color, type, and voice for the whole company. A social media style guide applies those same rules to a platform's exact specs: image sizes, caption length, reply tone, and the handle and bio for each account.
Should every platform get its own guidelines? No. The voice and the rules against certain claims stay the same everywhere. What changes per platform is the register, the image sizes, and a short note on what that platform's audience expects, which fits in a few lines per platform rather than a separate document.
Who should write the reply and comment guidelines? Whoever already handles escalations, usually the same person who owns the brand voice guidelines, working with whoever answers comments day to day. The people doing the replying know which situations actually come up.
How do I keep an AI tool from writing off-brand social captions? Connect it to a source that holds the brand's voice rules and approved components, so it reads them before drafting. With TrueKeep, that connection runs through MCP, and a check runs against the output before it ships, rather than relying on someone pasting the guidelines into a chat each time.
How often should social media brand guidelines be reviewed? Quarterly, at minimum, and after any platform redesign. Image specs and link-in-bio details go stale faster than the logo file does, so the review date belongs at the top of the document where it gets seen.
Building social media brand guidelines that hold up
Social media brand guidelines need five parts: profile and format specs, voice that flexes by platform without changing the underlying rules, approved visuals and templates, reply and comment guidance, and a named owner with a review date. The reply section is the one most teams skip, and the one tested first.
Write the specs, not the adjectives, the same way the rest of the brand guide should. Once the rules exist, the harder problem is reaching the person about to post, especially when that person is an AI tool drafting the first version of a caption. If you want the voice rules and the approved templates in one place your team and their AI tools both read, get started with TrueKeep.