Brand Guidelines

First PublishedMay 25, 2026ByAtif Alam

Looking for the strategic positioning chapter (segments, targeting, differentiation, value-proposition)? That lives at Strategy: STP / Positioning. This page covers the guideline artifact that operationalizes strategy, identity, voice, and story for any team.

The decision this page enables: how to build a brand guidelines artifact that actually gets used — not a 100-page PDF that goes stale in 90 days — and how to operationalize design ops, asset libraries, and AI policy on top of it.

Upstream: Brand Strategy, Identity & Visuals, Voice & Messaging, Brand Storytelling — everything upstream gets codified here. Cross-cutting: Promotion: Martech Stack — the deep treatment of AI and DAM tooling.

What brand guidelines are

Brand guidelines are the artifact that codifies brand strategy + identity + voice + story into a reference that anyone (internal team, agency, partner, advocate) can use to make on-brand decisions without checking with you first.

The guideline isn’t the brand. The guideline is the operating system that lets the brand scale beyond the 5 people who built it. Without one, every new marketer, every new designer, every new agency reinvents the brand a little bit, and you wake up at 100 employees with a brand that looks 100 ways.

A useful test: hand the guideline to someone outside the company and ask them to write a homepage hero or design a social-post template. If their output is recognizable as your brand, the guideline works. If it isn’t, the guideline is incomplete.

A naming clarification — these terms get used interchangeably and shouldn’t be:

• Brand guidelines — the full operating system (strategy + identity + voice + story + ops). What this page is about.
• Brand book — old-school term for a PDF guideline. Falling out of favor as static PDFs go stale.
• Style guide — usually means just the visual side (logo, color, type). A subset of brand guidelines.
• Design system — usually means just the UI/component side (buttons, forms, layouts). A subset, primarily for product teams.
• Identity manual — old term, similar to brand book. Visual-only.

In modern usage, “brand guidelines” is the umbrella; everything else is a subset.

The 4 audiences for brand guidelines

Why this matters: writing for one audience produces a guideline that fails the others.

Audience	What they need from the guideline
Internal teams (marketing, design, sales, support)	Searchable, current, with examples + don’ts. Used daily.
Agencies / vendors	Onboarding-friendly; standalone (they can’t ask follow-up questions easily).
Partners using your brand (co-marketing, integrations, channel)	Permission rules, logo-usage rules, what they can/can’t say. Usually a 1-page partner brand sheet.
Customer advocates	Public-facing version: how to share your brand correctly, how to apply for partner status, what assets are downloadable.

Modern brands often run two versions: a full internal guideline (private) + a public-facing partner-and-advocate version (Brand.ai, Frontify, or a dedicated subdomain).

The 12-section guideline outline

A complete, modern brand guideline. Treat as a starting outline; cut what doesn’t apply to your stage.

#	Section	What goes in it	Source
1	Strategy 1-pager	Purpose, mission, vision, values, personality, promise, archetype	Brand Strategy
2	Logo system	Variants, lockups, minimum sizes, clearspace, don’ts	Identity & Visuals
3	Color system	Primary, secondary, neutrals, semantic + tokens + WCAG pairs	Identity & Visuals
4	Typography	Display + body + monospace; type scale; weights; licensing	Identity & Visuals
5	Imagery	Photography + illustration + iconography style; inclusive-imagery rules	Identity & Visuals
6	Motion	Timings, easings, branded moments, reduced-motion fallback	Identity & Visuals
7	UI / Component library	Buttons, forms, cards, layouts — the design-system layer	Engineering + Design
8	Voice & tone	Voice attributes scorecard, message hierarchy, do-not-use list	Voice & Messaging
9	Storytelling library	Founder-origin story, top customer-as-hero stories, master narrative	Brand Storytelling
10	Applications	Web, social, presentations, ads, email, swag, packaging — concrete examples	Marketing + Brand
11	AI usage policy	Where AI helps, where it doesn’t, review rules, rights and disclosure	This page
12	Legal & usage rules	Trademark, partner usage, co-marketing, asset licensing	Legal + Brand

Some teams add: accessibility section (often woven through 3, 5, 6, 7), localization section (woven through 8 + 10), and a “rebrand history / brand evolution” section for established brands.

Format options and trade-offs

The format is the second-biggest decision after the content itself.

Format	Pros	Cons	When it’s right
PDF	Cheap to produce; shareable; offline	Goes stale in 90 days; not searchable; everyone has a different version	Almost never in 2026; only for one-time partner exports
Notion site / Coda doc	Cheap; searchable; easy to update; familiar to many teams	Hard to control look-and-feel; access management can be tricky	Early-stage; <50-person teams; brand budget under $10k
Figma library + style-guide page	Designers’ native tool; tokens live in the file	Hard for non-designers to navigate; tokens-as-source-of-truth is the right pattern only for product-led orgs	Design-led companies; product teams that use Figma daily
Web-hosted (Frontify, Brand.ai, Brandpad)	Always current; multi-user; can be public or private; integrations with DAM + Figma	Cost ($1–10k/year typically); requires setup investment	Brand spans 5+ surfaces; team of 20+; agency or partner involvement
Custom subdomain microsite (Webflow, Next.js)	Most polished; serves as PR + recruiting asset	Costs to build (~$30–100k) and maintain	Established brands; Series-B+; brand is part of the public reputation

Most modern brands run Figma library + a web-hosted guideline (Frontify / Brand.ai / a custom microsite). The Figma library is where tokens and components live; the web guideline is where strategy, voice, story, and rules live. They cross-reference.

AI in branding

Note: this section covers the policy side of AI in branding. The tooling and technical side lives at Promotion: Martech Stack.

By 2026, generative AI is a permanent part of every brand operation. The question is no longer whether to use it; it’s how to use it without diluting the brand. The policy needs to be specific.

Where AI helps in branding

• Variant generation — produce 10 hero headlines, 30 ad-creative directions, 40 LinkedIn-post drafts. Pick the best; reject the rest.
• Asset production at scale — 100 social-post images from a single template; ad-creative variants for paid testing.
• Length compression — turn a 2,000-word case study into a 300-word version for sales-deck use.
• Audience-specific reframings — same story, rewritten for VP of Engineering vs. VP of Marketing audiences.
• Component / layout generation — Figma + AI plugins generate dozens of card-component variants.
• Internal drafts — internal Slack messages, internal documentation, hiring emails. Low brand-stake, high time-savings.

Where AI doesn’t help (or actively hurts) in branding

• The founder’s voice — only the founder has it; AI imitations are detectable.
• Brand-narrative work — the master narrative, founder-origin story, category narrative. These need human craft, not AI smoothing.
• Legal / pricing / compliance copy — accuracy stakes are too high; AI gets things subtly wrong; brand-tone hallucinations cause real legal exposure.
• Customer-facing case studies (without human edit) — AI invents details and misreads context.
• Brand-voice consistency at scale — AI tends toward generic safe copy; without a strong voice doc + prompts + review, brand voice drifts to AI-default.

Rights and disclosure

• Who owns AI-generated work? In 2026, this is a jurisdictional patchwork. The US Copyright Office has held that purely AI-generated works are not copyrightable; human-edited AI work can be. The EU AI Act and similar regulations are still evolving. Practical rule: don’t ship pure-AI work as a copyrightable asset (logos, taglines, signature visuals). Always add human edit and document it.
• Disclosure — some jurisdictions and platforms require disclosure of AI-generated content (especially in advertising). Set a default policy of disclosing AI use where it’s user-relevant.
• Training-data provenance — increasingly important. Tools like Adobe Firefly that are trained on licensed content are safer for commercial use than tools with murky training-data origins.

The AI-in-branding policy template

ALLOWED with light human review:
- Internal Slack drafts, internal docs, internal hiring emails
- Variant generation for A/B-testing in paid ads
- Length compression of existing approved content
- Internal brainstorm support

ALLOWED with full human review:
- Customer-facing case study drafts (must be edited by brand)
- Blog post drafts (must be edited; ICP-checked)
- Social-post drafts (voice-checked)
- Ad-creative variants (brand-checked before launch)
- Email subject lines + body drafts (voice-checked)

NOT ALLOWED:
- Pure-AI brand-narrative work (founder story, master narrative, category narrative)
- Pure-AI legal / pricing / compliance copy
- Pure-AI customer-facing video where AI-generated faces appear
- Pure-AI logo / signature visual / tagline work
- Customer testimonials or quotes invented or rewritten by AI
- Confidential data, customer PII, or unannounced product details in prompts

DISCLOSURE REQUIRED:
- AI-generated images in advertising (per platform/region rules)
- Substantial AI-authored content in marketing emails (per CAN-SPAM evolution)
- AI-personalized outreach (per FTC AI Act guidance)

TOOLS APPROVED (and their use cases):
- ChatGPT / Claude — copy variants, drafts, editing
- Midjourney / Runway — image variants for paid ad testing only
- Adobe Firefly — commercial-safe image variants (licensed training data)
- Persado — push-notification + email subject-line optimization
- Writer.com — brand-voice-aware copy generation

This policy gets updated quarterly as the tooling and regulations evolve.

Design operations

Design ops is the practice of running the brand engine — asset library, review workflows, version control, sunsetting.

Asset library structure

A useful taxonomy:

/brand-library
   /logos
     /primary
     /monochrome
     /lockups
     /favicons
   /color-tokens
   /typography
   /imagery
     /photography
       /people
       /product
       /environments
     /illustration
     /icons
   /motion-references
   /presentation-templates
     /sales-deck
     /pitch-deck
     /webinar
   /social-templates
     /linkedin
     /instagram
     /tiktok
     /youtube-thumbnails
   /email-templates
   /paid-ad-templates
   /partner-assets

Each asset needs: a clear filename (logo_primary_color_RGB.svg), a description, the use case, the last-updated date, the named owner, and an integration link if it lives in Figma.

Naming conventions for assets

Consistency beats cleverness. The standard pattern:

{type}_{variant}_{color}_{format}.{ext}

Examples:

• logo_primary_color_RGB.svg
• logo_monochrome_white_PNG.png
• template_sales-deck_Q1-2026_v3.fig
• imagery_people_outdoor_Q4-2025_set-12.zip

When the team can find what they need in <2 minutes, the system works. When asset lookup takes 20 minutes, the system has failed.

Review and approval workflows

• Tier 1 (no review) — internal Slack, internal-only docs, A/B variants in paid ads (where the test itself is the review).
• Tier 2 (1-person brand review) — blog posts, case studies, social posts, email campaigns. Brand reviewer signs off in 24 hours.
• Tier 3 (multi-person review) — major launches, brand-narrative work, PR pitches, partner co-marketing. Brand + Marketing + Legal sign off, 3–5 days.
• Tier 4 (executive review) — visual identity changes, major messaging shifts, crisis communications. Brand + Legal + Exec sign off; timeline varies.

The mistake most companies make: putting everything through Tier 3, which becomes the bottleneck. Use tiers; trust the team within their tier.

Version control and sunsetting

• Old assets get archived, not deleted. A /archive folder with the date and reason.
• Logo refresh? Old logo to archive on the rollout date. No mixed assets in active use.
• Quarterly review — what hasn’t been used in 90 days? Either archive it or figure out why no one’s using it (often a discoverability problem, not a relevance problem).

How to build (or rebuild) brand guidelines (step by step)

A practical 6-week process for a 2-person brand team.

1. Audit the current state (Week 1). What guidelines exist? What’s current? What’s outdated? Where are the gaps?
2. Decide the format (Week 1). Notion + Figma for early-stage; Frontify / Brand.ai for growth-stage. Get budget approval.
3. Pull strategy, identity, voice, story content (Week 2). Most of this should already exist; this is where the assembly happens.
4. Write the 12 sections (Weeks 2–4). Start with strategy + voice (the “why” sections), then identity + applications (the “what” sections), then AI + legal (the “rules” sections).
5. Build the asset library (Week 3, in parallel). Structure first, populate second.
6. Define the design-ops workflows (Week 5). Review tiers, naming conventions, ownership.
7. Internal launch (Week 5). All-hands walk-through. Slack channel for questions. Office hours for the first month.
8. External launch (Week 6, optional). Public guideline microsite or partner-facing version.
9. Quarterly review cadence (Week 6 onwards). One person owns the guideline. Quarterly audit + updates. Annual full review.

Templates

Brand-guidelines section outline (12-section default)

1. Strategy 1-pager
   - Purpose, mission, vision, values, personality, promise, archetype

2. Logo system
   - Variants, lockups, sizes, clearspace, don'ts

3. Color system
   - Tokens, hex/RGB, WCAG pairs, cultural notes

4. Typography
   - Families, scale, weights, licensing, technical specs

5. Imagery
   - Photography, illustration, iconography, inclusive-imagery checklist

6. Motion
   - Timings, easings, branded moments, reduced-motion fallback

7. UI / Component library
   - Buttons, forms, cards, layouts (link to Figma + product design system)

8. Voice & tone
   - Voice attributes scorecard, message hierarchy, do-not-use list,
     tone-by-surface guidance, inclusive-language rules

9. Storytelling library
   - Master narrative, founder-origin story (multiple lengths),
     top customer-as-hero stories, story-arc-per-surface matrix

10. Applications
    - Web, social, presentations, ads, email, swag, packaging
      (concrete examples + templates per surface)

11. AI usage policy
    - What's allowed, what isn't, review rules, approved tools

12. Legal & usage rules
    - Trademark, partner usage, co-marketing, asset licensing,
      take-down policy

Asset-library taxonomy

/brand-library
  /logos
  /color-tokens
  /typography
  /imagery
    /photography (people / product / environments)
    /illustration
    /icons
  /motion-references
  /presentation-templates (sales / pitch / webinar)
  /social-templates (linkedin / instagram / tiktok / youtube)
  /email-templates
  /paid-ad-templates
  /partner-assets
  /archive

Each asset: filename, description, use case, last-updated, owner, Figma link.

Naming: {type}_{variant}_{color}_{format}.{ext}
Example: logo_primary_color_RGB.svg

Review-cadence calendar

WEEKLY
   Brand reviewer fields questions in #brand-questions Slack
   Asset library hygiene check (new files, broken links)

MONTHLY
   Brand-consistency audit (sample 10 random marketing pieces)
   AI-usage log review (which tools used for what; flag anomalies)

QUARTERLY
   Full guideline section review (one section per quarter on rotation)
   Asset-library cleanup (archive what hasn't been used in 90 days)
   Voice-consistency audit (sample 20 random pieces of copy)
   Cross-cutting AI-policy review (any new tools? any new disclosure rules?)

ANNUALLY
   Full guideline refresh
   Strategy alignment check (still matches Brand Strategy doc?)
   Format review (is Frontify / Brand.ai still the right tool?)
   ROI review (page-views, asset reuse rate, time-to-asset)

Metrics to track

• Brand-guidelines page-views per quarter — should be high. Guidelines that are referenced are guidelines that are being used. Low view counts = the team isn’t finding it or doesn’t trust it.
• Asset-reuse rate — % of marketing assets pulled from the library vs. created one-off. Target ≥70%. Low rate = the library isn’t comprehensive enough, or its discoverability is broken.
• Brand-consistency audit score across surfaces — quarterly sample, scored against the guidelines. Target ≥85%.
• Time-to-asset — median lookup time. Target <2 minutes. Above 10 minutes = the taxonomy is wrong.
• Guideline-staleness audit — months since last review per section. Target ≤6 months for any section.
• AI-usage log compliance — % of AI-assisted content that went through the right review tier. Target 100% on customer-facing surfaces.
• Partner / agency onboarding time — how long does a new partner take to be productive with the brand? Strong guideline = days; weak guideline = weeks.
• Brand-question Slack volume — total questions about brand per quarter. Should decrease over time as the guideline matures. Sustained high volume = the guideline doesn’t answer questions clearly.

Worked examples

SaaS workspace — Linear-like workspace tool

• Format chosen: Notion-based internal guidelines (Year 1, 80-person team) → migrated to a public Brand.ai microsite at Series-B (also serves as PR + partner-enablement asset).
• Structure: 12 sections as described above + a public “use our brand” section (logo download, partner co-marketing rules, podcast guest kits).
• Asset library: Figma library for components and tokens; Cloudinary for marketing image assets; Frontify for cross-team brand assets.
• AI policy: internal Slack drafts and ad-copy variants OK with light review; blog and case study drafts OK with full brand review; pure-AI brand-narrative work forbidden; tools approved are listed.
• Design ops: 4 review tiers; Tier 2 turnaround SLA of 24 hours; quarterly section review by the 2-person brand team.
• Year-2 metrics: asset-reuse rate 78%; brand-consistency audit 91%; time-to-asset median 1.8 minutes; #brand-questions Slack volume dropped from ~20/week (Year 1) to ~6/week (Year 2). The guideline is doing its job.

Fitness app — Strava-like consumer fitness app

• Format chosen: Brand.ai for internal team (privacy-controlled access for 220-person team) + Frontify for external agencies + a public-facing partner page on the main site for influencer / partner usage.
• Structure: 12 sections + a heavily-developed “inclusive imagery” section (with the explicit casting checklist for all photography work) + a “global brand expression” section per major market.
• Asset library: Cloudinary for image management (integrated with the marketing site for auto-optimization); Figma library for product + marketing components; a Notion-hosted “tone playbook” for support team specifically (same voice doc, framed for support context).
• AI policy: strict. AI variants OK for paid-ad creative testing; never used in-app or on core marketing site without human review; explicit ban on AI-generated bodies/faces in any user-facing surface (consistent with the inclusive-imagery values).
• Design ops: 5 review tiers (extra tier for influencer-partner co-marketing); quarterly Section 5 (Imagery) audit by an external accessibility consultant; quarterly Section 8 (Voice) audit against the inclusive-language do-not-use list.
• Year-2 metrics: asset-reuse rate 82%; brand-consistency audit 88%; #brand-questions Slack volume steady around 12/week (more agency activity = more questions). The guideline scales the team’s strict inclusive-language and inclusive-imagery values to a 5-agency creator network.

Common pitfalls

• PDF brand book that goes stale. Static documents become outdated within 90 days. Pick a format that updates in place.
• No asset library. Everyone redownloads logos from email; every team uses slightly different versions. Centralize, even if the centralization tool is just a shared folder.
• No AI policy until a customer-facing AI flub. Most companies wait until something embarrassing ships before writing a policy. Write it now; you’ll need it before you think.
• Guidelines no one reads. Symptoms: high #brand-questions Slack volume, low page-views on the guideline. Diagnosis: it’s too long, too generic, or hard to find. Fix: shorter, more examples, better search, internal launch.
• Over-prescriptive guidelines that block creativity. “Every CTA button must be exactly this hex and this radius.” This is a design system, not a brand guideline. Brand guidelines set principles; design systems set components.
• Under-prescriptive guidelines that produce drift. “Use the brand colors in a way that feels right.” Now everyone interprets “feels right” differently and the brand drifts surface by surface.
• No review cadence. Guidelines drift from reality in 18 months. Quarterly section review + annual full review keeps them current.
• Brand reviewer as bottleneck. One person reviews everything; they become the gating constraint on velocity. Fix by introducing tiers, training other reviewers, and trusting teams within their tier.
• AI tools without underlying brand data hygiene. AI is only as good as the prompt + the voice doc + the review. Throwing AI at a brand without strong voice guardrails produces generic-sounding output at scale.

Tools / further reading

• Designing Brand Identity (Alina Wheeler) — the operational reference for taking brand strategy through to identity through to guidelines.
• Brand Bible (Debbie Millman) — the canonical book on guideline craft.
• Brand Atlas (Alina Wheeler, Joel Katz) — visual reference for guideline structure + case studies.
• Tools (modern web-hosted): Frontify, Brand.ai, Brandpad — built for living brand guidelines.
• Tools (DAM): Cloudinary, Brandfolder, Bynder, Canto — for asset library management.
• Tools (Figma): Figma + Figma Design Tokens; Tokens Studio plugin; Figma Slides for presentation templates.
• Tools (AI-aware): Writer.com (brand-voice AI), Jasper (copy variants), Persado (push + email AI), Adobe Firefly (commercial-safe image AI).
• Public guideline examples to study: Mailchimp Content Style Guide, Atlassian Design System, Shopify Polaris, Microsoft Writing Style Guide, Vercel Brand. All are free + well-structured.

See also

• Brand Strategy — Section 1 of the guidelines.
• Identity & Visuals — Sections 2–7 of the guidelines.
• Voice & Messaging — Section 8 of the guidelines.
• Brand Storytelling — Section 9 of the guidelines.
• Promotion: Martech Stack — the deep treatment of AI tooling and DAM as part of the martech stack.
• Promotion: Social Media — creator-asset guidelines + partner brand sheets are downstream of this page.
• Internal Branding — the guideline supports the internal-branding rituals.