- Brand guidelines are a decision archive, not a taste document. Their job is to answer questions writers, designers, and vendors ask under time pressure, without a call to the founder.
- A real guide covers ten sections: purpose, logo system, color system, typography, voice, imagery, motion, digital patterns, print behavior, and governance. Skip any one and a whole category of question goes unanswered.
- Write for the person who has never met the brand team and is shipping something at 4 PM Friday. If your guide only works with a walkthrough, it isn't a guide.
- Format matters less than most teams think. A tight thirty-page PDF that gets read outperforms a beautiful microsite that gets abandoned. Choose the format your organization will actually maintain.
- Governance is what turns a document into a living system. Without an owner, a review cadence, and an intake process for exceptions, the guide is stale in eight months.
What brand guidelines actually are
Ask ten designers what belongs in a brand style guide and you'll get ten answers. Ask ten operators — the marketers, sales reps, and vendors who use one — and you get one answer stated a hundred different ways: it should tell me what to do without needing to ask someone. That's the whole job. Every other definition floats above this one.
Brand guidelines are a decision archive. They record the choices someone already made about how the brand looks, sounds, and behaves, so that the next person shipping a slide or a coaster or a pitch deck doesn't have to make those choices again. The value of a guideline document is proportional to how many decisions it makes on the reader's behalf. A twelve-page "guide" that only shows the logo is barely a guide; a thirty-page guide that answers voice, color, layout, and edge-case behavior is a working tool.
The mental model — guidelines as an SLA
Think of the document less like a book and more like an SLA between the brand team and everyone else. On the brand team's side: we've made these decisions so you don't have to. On everyone else's side: if the answer is in the guide, we won't ping you; if it isn't, we will. A well-written guide reduces the number of pings dramatically. A poorly-written one produces more pings, because it says a lot without deciding anything.
What guidelines are not
They're not the brand strategy. Strategy sits upstream — positioning, voice pillars, the reason the brand exists. Guidelines translate strategy into instructions. They're also not the identity system itself. The system is the logos, the colors, the type — the actual assets. The guidelines wrap around the system and explain how to use it. And they're not a marketing document. Nobody is being sold on the brand by reading the guide; they're trying to do their job.
The three failure modes
Guidelines fail in three predictable ways. First — the aesthetic monograph. Beautiful spreads, big photography, no actual instructions. Reads like a coffee-table book, useless when a rep needs to know what background color the logo can sit on. Second — the legalese lockbox. Rules stated so restrictively that everyone ignores them and does what they need to. "The logo must have clearspace equal to three times the height of the wordmark on all surfaces" is the kind of sentence that turns into a violation the first time somebody needs a favicon. Third — the wish list. Long lists of dos and don'ts without the underlying principles, so the moment a real-world situation isn't on the list, nobody knows what to do.
What a working guide feels like
You can tell a guide is working when three things happen. Someone joins the marketing team, reads the guide, and produces on-brand work in their first week without a lot of hand-holding. An outside vendor gets sent the guide and turns in files that don't need three rounds of correction. And the brand team's inbox thins out — the questions they used to answer over and over are now answered in the document. If any of those three things aren't happening after a launch, the guide isn't working, regardless of how it looks.
Who the guide is written for
Before you write a single page, decide who the guide is for. Every design decision downstream — length, format, tone, the level of detail on any given section — flows from that audience choice. Most guidelines fail because they were written for a founder or a design lead's peers, not for the people who will actually use the document day to day.
The five audiences
Internal marketing team — the people producing most of the outbound work. They need the most detail, and they use the guide most often. Sales — they build decks, one-pagers, and proposals under time pressure and they need instructions, not principles. Product and engineering — they use the identity inside the app, and they need patterns that map to code, not just to Figma. External vendors and agencies — they need everything, self-contained, because they don't have a relationship with the brand team beyond the guide. New hires across the company — they need enough context to represent the brand in email signatures, LinkedIn posts, and slide templates without embarrassing themselves.
The ranking that determines format
Rank those five audiences by how much of your guide's readership they'll actually represent. In a mid-market B2B company, the ranking is usually: marketing first, sales second, vendors third, product fourth, new hires fifth. In a consumer brand with a heavy agency footprint, vendors move to the top. In a founder-led startup, sales might be the primary audience because they're generating most of the on-brand output. The ranking matters because the primary audience's needs should shape length, tone, and format choices. A vendor-first guide has to be exhaustive and self-contained. A sales-first guide can afford to be shorter and more prescriptive because the users are internal and can ask questions.
The wrong audience
The audience you should not write the guide for: your design peers, your CEO's aesthetic preferences, or the imagined "world-class brand" you'd like to be someday. Guides written to impress design peers are usually gorgeous and useless. Guides written to defend a CEO's taste read like manifestos and get quietly ignored. Guides written to a future aspirational brand describe a company that doesn't exist and can't produce assets for today's real work.
The 'one story' rule
Before drafting, write a single sentence about the primary reader. "This guide is written for a marketing generalist at Alderwood Coffee who needs to produce a sell sheet for a new SKU by Friday." That sentence becomes a reference point during writing. Every section can be tested against it: does this help our marketing generalist ship the sell sheet by Friday? If the answer is no, cut the section or move it to an appendix.
The vendor stress test
The single best test of a guide is to send the current draft to an unfamiliar vendor and ask them to produce a small piece of work — a business card, a social post, a one-pager. If the vendor can't complete the work without back-and-forth, the guide has gaps. Marqee's coaches call this the "cold hire" test in a different context; the underlying idea holds here. If a competent outsider with only the guide can't produce reasonable output, the guide isn't finished.
Which reader is missing from most guides
The reader most often missing from published guidelines is the operations person — the one shipping the client-facing operational surface. Proposals, invoices, engagement letters, contracts, packaging labels, shipping inserts. In our production data on print work, we see companies that have exquisite guidelines for marketing surfaces and zero guidance on operational surfaces, which means every printed invoice ends up looking like it was designed by a different company. Write the operations person into the audience list from the start.
Section one — purpose and personality
The first substantive section of a working guide is a short, honest statement of what the brand is for and how it behaves. Not the strategy document — the operative summary of the strategy that everything else in the guide will assume. This section is short — three to five pages — and it sets the reader up to make judgment calls in situations the guide can't cover explicitly.
What to include
A one-paragraph purpose statement. Not the marketing tagline. The plain-English answer to "why does this company exist and who is it for." Then three to five voice pillars — the personality traits that show up in every piece of communication. Then the audiences the brand is talking to, described concretely enough that the reader can imagine them. Then two or three explicit "we are not" statements — the adjacent brands or archetypes we are deliberately not.
Why the 'we are not' statements matter
Positive brand definitions get vague quickly. "Warm, professional, approachable" — every brand claims those. What actually differentiates a brand is what it refuses to be. "We are warm but we are not folksy. We are professional but we are not corporate. We are approachable but we are not chatty." The negatives sharpen the positives and give the reader a boundary to work against when writing new copy or designing new assets.
The tone of this section
Write this section in the brand's actual voice, not in the abstracted voice of a design document. If the brand voice is warm and direct, the purpose section should read warm and direct. If the brand voice is precise and technical, the purpose section reads precise and technical. Nothing tells a reader more about a brand than the tone of the document that describes it. If the guide describes a warm brand in cold prose, the reader hears the cold prose louder than the warm claim.
What to leave out
The full brand strategy deck. The history of the founding. The competitive teardown. The customer research summary. Those live in adjacent documents — strategy memos, positioning briefs, research reports — and get linked to from the guide for readers who need them. The guide's purpose section is the operative summary, not the archive.
The 'stranger test'
Read the purpose section to someone who has never heard of the company. Ask them what they think the company does. If they can describe it in a sentence, the section works. If they hedge or ask for clarification, rewrite. The stranger test is the single fastest way to find out whether the section is doing its job, and it's cheaper than any focus group.
When to update
The purpose and personality section should be updated when the strategy changes, not on the calendar cycle that governs the rest of the guide. If the company pivots, this section gets rewritten first, then the rest of the guide catches up. If the strategy hasn't changed, don't touch it. Cosmetic updates to the purpose section confuse readers who thought they knew what the brand was.
Section two — the logo system
The logo section is the one most guides get almost right and still fail. The common failure is showing the primary logo, listing a set of restrictive rules, and calling it done. A working logo section shows the whole system, explains behavior across surfaces, and covers the edge cases that come up in real production work.
What to include
Every mark in the system, at full resolution, downloadable from the guide. The primary logo. The wordmark alone. The symbol alone. The stacked variant. The horizontal variant. The one-color variants. The reversed (white) variants. Any co-branded variants if the company has partnership relationships. The favicon and app-icon versions with their own rules. Missing any one of these creates a support request every time someone needs it.
Clearspace and minimum sizes
Show clearspace visually, using the logo's own geometry as the measurement unit (not "3mm on all sides" — that doesn't scale). Show minimum sizes for print and digital separately, because digital resolution behaves differently. For a wordmark to remain legible on business-card back, you probably want at least 12mm wide. For it to remain legible in a mobile-app tab bar, you need a different set of constraints. Show both.
Backgrounds and contrast
Show the logo on every background color the brand uses, plus white, plus black, plus a photograph, plus a busy pattern. Don't just state contrast rules — show them. "Use the reversed logo on backgrounds darker than X" is instruction; "Here is the logo on backgrounds A, B, C, D — use these variants" is a decision the reader can execute immediately.
Common misuse examples
Every logo section should have a page of "don't do this" examples showing the logo distorted, recolored, given a drop shadow, placed on a busy background, stretched, condensed, and outlined. Show the correct version alongside each mistake, so the reader can compare. The misuse examples are the pages readers actually reference — they answer specific questions from real production work.
The 'clearspace violation' policy
Real production sometimes requires violating the clearspace rule — a favicon at 16px, a merchandise tag, a co-branded lockup where two logos need to sit closer than the rule allows. Rather than pretending these situations don't exist, the guide should state the policy for exceptions. Usually: "When clearspace can't be observed, prioritize legibility of the primary mark and consult brand@company.com for high-visibility exceptions." Guides that acknowledge exceptions get followed. Guides that pretend exceptions don't exist get ignored.
File formats and where to get them
The last page of the logo section is the download index. SVG for web. PNG at three resolutions (72, 144, 300 dpi equivalents for common sizes). EPS or AI for print production. PDF for anything vendor-facing. Include a filename convention so downloaded files stay organized. In our production data on print jobs, the single largest source of file-setup errors is vendors receiving a raster logo when they needed vector; the download index prevents this by making the correct file the obvious one.
Section three — the color system
Color is where guidelines quietly disintegrate. The primary swatch is easy. The secondary swatch is easy. Then someone asks "what background color should the sales deck use" and the guide has no answer. A working color system covers every place color shows up, not just the primary palette.
The four color roles
Every working color system defines colors by role, not just by hex value. Primary — the identity color, the one someone thinks of when they picture the brand. Secondary — the supporting colors that work with the primary. Neutrals — the grays, off-whites, and blacks that carry most surfaces. Accents — the small-quantity high-impact colors used for links, CTAs, warning states, and specific data-viz needs. Every color in the palette should have a role. Colors without roles get misused, because the reader doesn't know when to reach for them.
Color specs to publish
Every color needs six specs published: hex, RGB, HSL (or LCH for modern digital work), Pantone Coated, Pantone Uncoated, and CMYK. Missing any one produces production errors downstream. A designer picking hex without a Pantone match will produce screen-accurate mockups that print differently. A print vendor picking Pantone without CMYK will fail on a digitally-printed job. Publish all six.
Contrast and accessibility
Every text-on-background combination should have a documented WCAG contrast ratio. Don't leave this as a general "be accessible" statement; show the pairs. "Primary blue on white — 6.2:1, passes AAA at 18pt and above." "Secondary teal on primary blue — 1.8:1, decorative only, never text." Real accessibility work is per-pair. Guides that document contrast per-pair produce accessible output; guides that state accessibility as a principle produce inaccessible output with good intent.
Digital vs print behavior
Every color has different behavior in digital and print. RGB is additive (adds light); CMYK is subtractive (adds ink). Some brand colors that look vivid on screen can't be matched in CMYK without a Pantone plate. State this openly: "Our primary blue is Pantone 302 C. In four-color-process print, it will render approximately as CMYK 100/60/0/25, slightly less saturated than screen. For matching-critical print jobs, specify the Pantone." Guides that hide this gap produce disappointed print buyers.
The 'tint scale' every guide skips
Most guides show the primary color and stop. Working guides also show a tint scale — 10%, 25%, 50%, 75%, 100% of every core color — because real design work needs backgrounds and hover states and disabled buttons, all of which need lighter versions of the palette. Publishing tints saves the design team from re-deriving them constantly and prevents subtle drift where every product surface uses a slightly different "lighter blue."
The proportion rule
Beyond individual colors, the guide should describe how the palette is used together. What percentage of a given surface should be primary color versus neutral? What is the typical hierarchy of color usage? The classic 60/30/10 rule is a starting point — 60% neutral, 30% primary, 10% accent — but it should be adapted to the specific brand. A brand that leans heavily on white space might be 80/15/5. A brand with a bold identity might be 40/40/20. Publish the proportions the brand actually uses, so new designers can match them without guessing.
Section four — typography
Typography guidelines fail more often than color guidelines, because type has more variables. Family, weight, size, leading, tracking, casing, hierarchy — each of them a decision. A working typography section decides all of them, not just the family choice.
The type stack, decided
Publish exactly which typeface families the brand uses and in which weights. If the brand uses a primary type family for display and a secondary for body, name both. If the brand's licensed fonts have a web-safe fallback for email or product surfaces where the licensed font can't render, publish the fallback. A stack looks like this: primary display (Halcyon Display Bold), primary body (Halcyon Text Regular and Medium), web fallback (Georgia for serifs, Arial for sans), email fallback (system-ui, -apple-system, Segoe UI, Roboto, sans-serif). Every context that will show text should have a defined stack.
Weight and size decisions
Publish the type scale as a table with specific pixel and point values for every hierarchy level. H1 — 48px display / 40pt print. H2 — 32px / 28pt. H3 — 24px / 20pt. Body — 16px / 11pt. Small — 14px / 9pt. Caption — 12px / 8pt. Line height and letter spacing published alongside. Type scales published as "large, medium, small" without values leave every implementer guessing.
Hierarchy in the wild
Show the type system applied to real surfaces — a sales-deck slide, a blog post, a marketing landing page, a business card, a physical sign. The specific application matters more than the abstract scale. A type scale that works on a website often needs adjustment on print collateral; showing both prevents mismatched execution.
Casing and character rules
Publish casing decisions explicitly. Title case or sentence case for H1s? Are section labels uppercase? Are buttons uppercase or sentence case? Publish the answer, not the debate. Also publish specific character rules — how the brand handles numbers (tabular versus proportional in tables), ampersands (spelled out or symbol), and any special characters that come up in the brand name.
Bad-substitution habits
Every guide should list the fonts the brand explicitly refuses to use as substitutes. Not because they're bad fonts — because they carry the wrong associations. If the brand uses a modern geometric sans, name the fonts people will try to substitute (usually the free Google alternatives) and note explicitly that they don't match. This saves conversations that happen after the fact, when a slide has already shipped with the wrong font.
Type in digital versus print
Digital type and print type behave differently. On screen, hinted rendering can compensate for tight letter spacing. On paper, tight spacing collapses at small sizes. Publish specific advice: "On print collateral at 9pt or smaller, use Halcyon Text Medium rather than Regular, and increase tracking to 5 units." Advice this specific is what makes production output consistent across vendors. In our production data, mismatched type between vendors is second only to color mismatches as the top complaint on multi-piece print jobs.
Section five — voice
Voice is the section most guides do worst. The section usually consists of three or four adjectives and a request to "sound professional but approachable." That's not a voice guide; it's a mood board. A working voice section decides sentence-level questions with enough specificity that a new writer can produce on-brand copy in a first draft, not a fifth.
Voice pillars, made operational
Start with three to five voice pillars — the personality traits that consistently show up. Then make each pillar operational with a triple: what it means, what it looks like in a sentence, and what it explicitly is not. "Warm — we sound like a person who knows the reader's situation. Sample: 'This part usually trips people up; here's the thing that helps.' Not: 'We understand your journey.' Not: 'Cheers, mate.'" Three specifics per pillar beats five abstract adjectives.
Do-and-don't sentences
The single highest-leverage artifact in a voice guide is a table of side-by-side sentence pairs. Left column: how someone might write it. Right column: how the brand writes it. Draw pairs from real production surfaces — the homepage hero, a subject line, an in-product error message, a support reply, a sales-deck headline. Every writer references this table. Everything else in the voice section is scaffolding around this table.
Vocabulary lists
Publish a short list of words the brand uses and words it avoids. Not because the avoided words are wrong — because they carry the wrong tone. "Use: help, walk through, explain, plain-language. Avoid: leverage, unlock, elevate, streamline." The list is short (twenty words in each column at most) and evocative. Guides with fifty-word vocabulary lists get skimmed; twenty-word lists get read.
Contractions, punctuation, and case
Decide the small stuff. Do we use contractions? (Almost always yes.) Do we use the Oxford comma? Do we capitalize titles or use sentence case? Do we use exclamation marks? Ellipses? Semicolons? Ampersands in headlines? Each of these is a small decision that shows up in every piece of writing; publish the answer once so writers don't relitigate it a thousand times.
Voice by channel
Show the voice adapted to different channels — homepage headline, subject line, in-product microcopy, sales email, legal notice. The base voice is consistent; the register shifts. Homepage can be warm and confident. In-product microcopy needs to be terse and unambiguous. Legal needs to be precise. Sample each channel with a side-by-side of a real bit of copy in the wrong register versus the right one.
The 'author test'
The strongest voice-guide test is to have three different writers produce the same short piece of copy — a subject line, a hero headline, a support reply — using only the guide. If the three outputs read like they came from the same brand, the guide works. If they read like three different brands, the guide has gaps. This test is cheap and honest, and it catches voice-guide failures better than any workshop.
Section six — imagery and iconography
Imagery is the section most likely to age badly. Photography styles date fast, and icon systems drift over years as different designers add new symbols. A working imagery section publishes the principles that will survive stylistic evolution alongside the current-execution examples.
The imagery principles
State the principles that don't change: what the imagery is for, whom it depicts, what feeling it's meant to produce, and what it explicitly avoids. "Our photography shows real people in real work environments. We favor natural light, close crops, and moments of concentration rather than moments of celebration. We avoid: staged stock, group high-fives, laptops-with-smiling-face, and any composition where the subject is looking directly at the camera." Principles like these survive when the specific aesthetic evolves.
The current-execution examples
Alongside the principles, show what the current photography actually looks like. Six to twelve images at high resolution, representing the range of subject matter and framing. Update this section annually as the library grows. The principles stay; the executions rotate.
Icons — the two families
Most brands need two icon families. Interface icons — the small monochrome symbols used in product UI and web navigation. Marketing icons — the larger, sometimes multi-color symbols used in feature grids, sales decks, and landing-page illustrations. Publish both families separately. Interface icons need consistent stroke weight, corner radius, and grid alignment. Marketing icons have more expressive latitude. Guides that conflate the two produce inconsistent output because the same designer follows one set of rules for a marketing surface and a different set for a product surface.
Illustration style — decide or don't have one
If the brand uses illustration, publish the style with enough specificity that a new illustrator can match it. Line weight, color palette, composition rules, level of detail, treatment of hands and faces. If the brand doesn't use illustration, state that explicitly so nobody adds it. A missing decision on illustration is what produces the "we hired someone new and now our blog posts have cartoon avatars" moment.
Stock photography — the honest guidance
Most brands use stock at some point. Rather than pretending they don't, publish the rules for when it's acceptable and how to choose it. "Stock is acceptable for internal decks and low-visibility blog headers. For homepage, sales, or paid campaigns, use commissioned or in-house photography. When selecting stock, avoid: obvious staging, generic laptops, and any image where the subject reads as "business person on white background."" Realistic guidance beats aspirational silence.
File formats and naming
End the section with file conventions — image resolutions for web and print, allowed file formats, naming rules. Naming rules matter more than teams realize: consistent naming means images can be found and reused; inconsistent naming means the design team re-crops the same portrait five times because they couldn't find the original.
Turning a guideline into printed reality?
Inkgility handles print production — business cards, signage, packaging, operational surfaces — with the same identity rules across every piece.
See business-card production →Section seven — motion and interaction
Ten years ago, motion sections were exotic. Now every brand ships motion — in product UI, in social ads, in webinar bumpers, in loading states — and the guide needs to answer questions about it. A working motion section covers what the brand's motion feels like and where it belongs.
Motion principles
State the feel of the brand's motion in three or four principles. Duration — is the brand's motion fast (200ms transitions) or unhurried (600ms transitions)? Easing — sharp linear moves, or curved organic ones? Choreography — do multiple elements move simultaneously, or is one element leading and others responding? Register — does the brand's motion tend toward playful bounce or restrained precision? Every downstream motion decision — a hover state, a page transition, a video sting — flows from these principles.
Reference animations
Show reference animations (as GIFs or embedded video in the guide) that demonstrate the principles. Six to twelve short clips — a button hover, a modal open, a page transition, a logo animation, a chart draw-in. Reference animations are more useful than written descriptions because motion is hard to describe in prose but easy to recognize when shown.
Where motion belongs and where it doesn't
Not every surface should have motion. Publish the decision. Motion belongs in product UI transitions, social ads, homepage hero moments, and event graphics. Motion does not belong on printed materials (obviously), in sales-deck slides between substantive content (distracting), or in email (usually broken and slow). Publishing the negative cases prevents the common failure where a well-meaning designer adds animation to every possible surface.
The brand's motion signature
Most brands have one or two "signature" motion moments — the logo animation on the homepage load, the specific transition between product screenshots in a sales video, the way a chart draws in. Publish these signature moments as protected — they should be reused, not reinvented every time. A recognized signature reinforces brand identity across surfaces.
Accessibility and reduced motion
State the reduced-motion behavior clearly. Users who have prefers-reduced-motion enabled at the OS level should see either static versions of animated elements or dramatically reduced motion. Publish which animations degrade to what static state so implementers don't have to guess. This is not a nice-to-have — it's a real accessibility need for users with vestibular disorders, and it's easy to get wrong without explicit guidance.
Motion in print collateral
This sounds oxymoronic but matters. Motion principles influence how a brand handles motion suggestion in static print — the sweep of a photograph, the direction of a diagonal composition, the arrows in a diagram. A brand with fast, sharp motion often has sharper compositional diagonals in print. A brand with restrained motion often has more centered, symmetric print compositions. State the through-line so print work reinforces motion work, and vice versa.
Section eight — digital patterns
Digital patterns are the section product and engineering teams read most and marketing teams read least. It documents how the identity system behaves in the product itself — buttons, forms, tables, navigation, empty states — and it usually needs to be maintained closer to the code than the rest of the guide.
The 'design system' overlap
Some companies have a full design system living in code — components in React or Vue or Swift with published APIs. When one exists, the brand guide should point to it rather than duplicate it. Duplication is where drift starts: two documents describing the same component eventually disagree. Publish the source of truth for each pattern (either the guide or the design system) and link across from the other.
Buttons — the pattern most guides underdocument
Buttons carry more brand weight per pixel than almost anything else. The color, corner radius, weight, hover state, disabled state, loading state, focus ring, and label rules all show up thousands of times a day in a working product. Publish them exhaustively. Include a decision tree for when to use primary versus secondary versus tertiary button treatments. Include the copy rules for button labels (usually action verbs, sentence case, no more than four words).
Forms and inputs
Every product ships forms. Publish the input styling, label positioning, help-text style, validation states, and error messaging tone. Include the accessibility patterns — label association, required-field indicators, error-message announcement. Form patterns are where accessibility either lives or dies in a product, and abstract advice doesn't cut it.
Navigation
Publish the navigation pattern for every surface: top nav for marketing site, sidebar for product, tab bar for mobile, breadcrumbs where applicable. Show the interactions — hover states, active states, mobile collapse behavior. Include the copy rules — sentence case, verb-first navigation labels, plain-language section names.
Empty states, loading states, and errors
The three states every product ships and most guides ignore. Empty states — what does the screen look like before the user has data? Loading states — what does the screen look like while waiting? Errors — what does the screen look like when something went wrong, and how does the brand handle the tone of the error message? These states are where users form durable opinions about the product, and consistent handling matters more than aesthetic polish.
Copy patterns for microcopy
Publish the voice rules specifically for microcopy — the small text inside the product UI. Button labels, tooltips, error messages, confirmation prompts. Microcopy voice usually needs to be terser than marketing voice, but the underlying personality is the same. Include a table of common microcopy patterns and how they should read.
Iconography inside product surfaces
Product icons need special treatment. Publish the icon grid (usually 24×24 for interface, 16×16 for dense contexts), the stroke weight (usually 1.5 or 2), the corner-radius convention (usually 2px rounded), and the color rules (usually inherit from text color for accessibility). Missing any of these produces a UI where different features look like they came from different products.
Section nine — print behavior
Print behavior is the section most modern brand guides omit entirely, on the assumption that everything is digital now. Everything is not digital. Business cards, packaging, printed collateral, event signage, direct mail, and the client-facing operational surface (invoices, contracts, engagement letters) still exist in physical form for most companies, and each of them needs identity rules.
Why print gets its own section
Print behavior isn't just "digital, printed." Ink on paper behaves differently. Colors shift. Type at small sizes needs adjustment. Photographs need different color profiles. Backgrounds that work on screen bleed too dark on paper. Every one of these differences produces failed jobs when the guide doesn't address them. In our production data, roughly a third of print reruns trace back to identity assets prepared for digital being sent to a press without adjustment.
Core specs
Publish CMYK and Pantone equivalents for every brand color (covered above). Publish minimum type sizes for print (usually 8pt for body, 6pt for legal disclosure). Publish bleed and safe-area rules (usually 0.125" bleed, 0.125" safe area from the trim). Publish resolution requirements (300 dpi at final size for photography; vector for logos and type).
Paper stock guidance
The brand should have opinions about paper. Which stocks does the brand default to? A coated cover for business cards? An uncoated text for stationery? A soft-touch finish for premium collateral? Publish the defaults with actual stock names and weights when possible: "16pt uncoated cover, Neenah Classic Crest Natural White, 130 lb" is a real instruction. "Premium uncoated stock" is a vibe that different vendors will interpret differently. The default doesn't lock you in — a specific vendor can substitute if the intent is preserved — but it does anchor expectations.
Finish and special processes
If the brand uses foil, spot UV, letterpress, embossing, or any other special process, publish the rules. Where can foil appear? What color? What logo variants can be foiled? Spot UV is applied to which elements? Letterpress requires bold weights of type (thin weights fill in), so publish the type-weight rules for letterpress separately. These are the details that separate a brand that ships beautiful print work from one that ships "good enough" print work.
Photography for print
Print photography needs different color profiles than web photography. Publish the ICC profile the brand uses (usually GRACoL 2013 for commercial coated print, or SWOP for uncoated). Publish resolution requirements at final size. Include guidance on skin tones (usually more forgiving in print than digital) and blacks (rich black CMYK for large solid areas, black-only for small type).
Operational print — the surface most guides skip
Invoices, statements, contracts, engagement letters, shipping inserts, packing slips. These are the surfaces customers actually see most, and they're almost always designed by whoever set up the accounting software fifteen years ago. Publish the identity rules for operational documents: logo placement, typography, color usage, footer content. When a customer receives a beautifully-branded proposal and a Times-New-Roman invoice, the invoice is the surface they associate with your competence.
Section ten — accessibility and inclusion
Accessibility used to be a compliance topic; now it's an identity topic. A brand that ships inaccessible work is telling a substantial fraction of its audience it wasn't built for them. Modern brand guidelines treat accessibility as a design principle, not a checklist appended at the end.
The three accessibility axes
Visual — contrast, type size, color independence, motion behavior. Cognitive — plain language, predictable navigation, uncluttered layouts. Motor — tap targets, keyboard navigation, focus states. Each axis needs its own rules in the guide, because the failures are different. A brand can score well on visual and fail cognitive because the copy is dense; publishing the axes separately keeps the reader from checking one box and moving on.
Contrast — specific pairs, not general rules
Covered above in color, worth repeating here. Every text-on-background combination the brand uses should have a documented WCAG contrast ratio. Publish the passing combinations and explicitly list the failing combinations as text-decorative-only. When a designer needs to place text on a background, the guide should give an unambiguous answer.
Language and reading level
Publish the target reading level for public-facing copy — usually eighth to tenth grade for consumer surfaces, higher for specialist B2B. Publish the plain-language rules the brand follows — short sentences, active voice, common vocabulary, defined jargon. Guides that publish these rules produce more accessible writing without heroics.
Color independence
State the rule: no information is conveyed by color alone. Every chart legend has a shape as well as a color. Every state indicator has an icon as well as a color. Every link is underlined, not just colored. This rule prevents a whole class of accessibility failures without adding much design work.
Focus and keyboard navigation
Publish the focus-ring style — its color, thickness, and offset. Publish the tab order the brand's UI follows. Publish the keyboard shortcuts the product supports. These are patterns that engineering will implement; the guide's job is to state the visual and interaction rules so the implementation matches the brand.
Inclusion in imagery and voice
Beyond technical accessibility, publish the inclusion principles for imagery and voice. Whose stories does the brand tell? Whose bodies show up in photography? What identities are named in customer examples? Explicit inclusion principles produce more inclusive work than well-meaning vagueness. Publish the categories the brand actively represents — age range, race and ethnicity, gender identity, body type, disability visibility — and the categories the brand avoids stereotyping.
Section eleven — governance and updates
A guideline document that isn't governed becomes stale, then wrong, then quietly ignored. Governance is the section most guides omit and the section that most determines whether the guide is still useful in two years.
The single owner
The guide needs a named owner — a person, not a committee. The owner is responsible for approving updates, resolving edge cases, and running the review cadence. When the owner leaves, the succession is explicit: the new owner is named within thirty days, and the guide's version history logs the change. Guides without an owner drift because no one is empowered to keep them straight.
The review cadence
Publish the review cadence at the front of the governance section. Quarterly for a fast-moving brand; annually for a stable one. The cadence is not "we'll review it when we get to it" — that's how guides become stale. The cadence is calendared and honored, with the review producing either "no changes needed" (documented) or a specific set of updates (published).
The intake process for exceptions
Real work will always produce cases the guide doesn't cover. Publish the intake process for exceptions: who to ask, how quickly they'll answer, what evidence to submit. A working intake process usually looks like "email brand@company.com with the specific case; response within one business day; if approved, the exception gets logged and, if it recurs, added to the guide." Guides that don't publish this process force people to guess or ignore the guide entirely.
Versioning
Publish the version number, the change date, and the changelog. Version 1.0 for launch. Version 1.1 for minor updates. Version 2.0 for major re-writes (usually paired with a brand evolution). The changelog matters — six months later, someone will ask "when did we change the CTA color" and the changelog answers.
Enforcement — soft, not hard
The best guides enforce softly. They provide the right thing to do, make it easy to do, and educate when something drifts. Hard enforcement — insisting every asset go through brand approval before ship — produces bottlenecks and resentment. Soft enforcement — templates, plug-ins, review-on-request, quarterly audits — produces on-brand output without the friction. Publish the enforcement philosophy explicitly so nobody's confused about how strict the guide is meant to be.
Deprecation — when to sunset an asset
Assets get deprecated over time. An old logo variant, a color that's been retired, a photograph that no longer fits. Publish the deprecation policy: how assets get marked as deprecated, how long they remain in use before being phased out, who's responsible for replacement. A guide with clean deprecation avoids the situation where a legacy asset silently persists in one channel for years.
The 'living document' fiction
Most guidelines call themselves "living documents." Almost none of them are. A living document has an owner, a cadence, an intake process, and a changelog — all of which most guides skip. State plainly whether the guide is actively maintained or archived, and update the top of the document accordingly. Honesty about maintenance status is more useful than aspirational claims about being "living."
Format choices — PDF, microsite, or hybrid
Once the content is decided, the format question comes up. PDF or microsite or hybrid? Every option has trade-offs. The right answer depends less on aesthetics than on how the organization will actually maintain the document.
The PDF option
Advantages: portable, versioned, printable, easy to distribute to external vendors. Disadvantages: hard to update in-place (a new version means everyone downloads again), doesn't support motion or interactive elements well, size can balloon. Right when: the primary audience is external vendors, the brand ships a lot of printed collateral, or the organization has a legal or archival requirement for versioned documents. A well-made PDF guide is typically twenty to sixty pages and downloads at under 30MB.
The microsite option
Advantages: easy to update in place, supports interactive components (color pickers, live typography previews, downloadable asset kits), can embed motion. Disadvantages: needs hosting and maintenance, harder to hand to a vendor as a self-contained document, invisible when someone's offline. Right when: the primary audience is internal, the organization has front-end capability to maintain the microsite, or the guide includes many downloadable assets that need version control.
The hybrid
Microsite as the source of truth for internal teams; PDF export generated from the microsite for external vendors. Best of both worlds when done well; more infrastructure to maintain when done poorly. Right when: the organization has design-engineering capacity and the guide needs to serve both internal and external audiences well. In practice, most large enterprise brands eventually land here.
The 'plus assets' packaging
Whatever format is chosen, ship the assets alongside. A Google Drive folder or Dropbox with logos, color chips, type samples, photography selects, template files, and social-media covers. Guides that make assets easy to download get followed; guides that require asking-for-assets get worked around. In our observation, roughly a third of off-brand output traces back to "I didn't want to bother them for the file, so I used the one I already had."
The wrong reasons to choose a format
Do not choose format for aesthetic reasons. "A microsite feels more modern" is not a maintenance plan. "A printed book will impress the leadership team" is not a distribution strategy. Choose format for how the guide will actually be maintained and consumed over the next three to five years.
The 'print the guide' failure
Every couple of years, someone will suggest printing the brand guide as a physical book. Usually a coffee-table object handed to new hires. It's a beautiful idea and almost always a bad investment. Printed books freeze the guide at a moment in time; the guide is out of date the day it prints. If leadership wants a tangible artifact, print a small run of the top-line "personality and voice" pages as a hand-off document, and keep the operational guide digital. Nothing loses value faster than a lavish printed brand guide that describes an old version of the brand.
Common failures and how to avoid them
After looking at a lot of brand guides, a small set of failures recur. Naming them upfront helps you avoid them, but the more useful move is knowing what to look for when reviewing an existing guide.
Failure 1 — no logo behavior specification
The guide shows the logo and lists misuses, but doesn't specify behavior across surfaces. What color is the logo on the sales deck's title slide? On the product's login screen? On the mobile app icon? Missing behavior specs produce inconsistent execution, because every designer solves the problem differently.
Failure 2 — color without Pantone matches
Covered above; recurs so often it's worth naming again. Hex-only palettes produce printed pieces that vary vendor to vendor. Print buyers don't complain formally; they just leave the brand.
Failure 3 — voice pillars without do-and-don't sentences
Also covered. Pillars are decoration without pairs; pairs without pillars lack a frame. Both are needed.
Failure 4 — a system that only works with the current design team
Some guides encode so much implicit knowledge that they only work when the current design team is available to interpret them. A vendor can't produce good output; a new hire can't produce good output. The test: send the guide to an unfamiliar designer with a small brief and see what comes back. If it needs three rounds of correction, the guide is under-specified.
Failure 5 — no operational-surface guidance
Marketing surfaces are documented; the invoice looks like it was designed by a different company. Include the operational surfaces from day one, or accept that the brand will only cohere in marketing contexts.
Failure 6 — governance skipped
No owner, no cadence, no intake process. The guide is stale by the second birthday. Publishing an owner-of-record and a review calendar is a five-page section that changes the guide's half-life from two years to five.
Failure 7 — a beautiful book nobody uses
Coffee-table quality, gorgeous photography, useless for actual production work. If the guide's pages fill a wall but nobody references it when shipping, it's a monograph, not a guide. Rewrite the top-line summary into a working reference and let the beautiful pages live as a marketing artifact separately.
Failure 8 — the everything-and-the-kitchen-sink guide
Every possible detail included, table of contents four pages long, three hundred pages total. Nobody reads it. The best guides are shorter than they need to be, with clear pointers to reference documents for depth. Length correlates negatively with usage above about eighty pages.
Failure 9 — no accessibility guidance
Accessibility handled as an afterthought or an appendix. The brand's inaccessible surfaces speak louder than the accessibility appendix does. Publish accessibility as an identity principle, not a compliance checkbox.
Failure 10 — the guide about the guide
Pages of introduction, meta-commentary, and self-congratulation about the brand's journey. Cut it. The reader needs to know how to use the logo, not how the brand team feels about the brand.
How to launch (and relaunch) a guide
Writing the guide is roughly half the work. Launching it well is the other half. A guide that ships to a shared drive and doesn't get referenced is worth less than a shorter guide that gets deliberately rolled out.
The internal launch
Announce the guide with a leadership sponsor. A short email or all-hands note from someone with authority — usually the CEO or CMO — explaining why the guide exists, what changed, and where to find it. Silent launches produce silent adoption. A one-paragraph endorsement from leadership is worth more than the guide's cover design.
The vendor rollout
Every vendor and agency working with the brand receives the guide, plus a short cover note explaining what changed. Attach the asset kit. Set an expectation about when the new guide takes effect. Vendors that get a smooth rollout produce on-brand work in the first project after the change; vendors that get a chaotic rollout produce three months of correction cycles.
The training moment
Run a thirty-to-sixty-minute training for the marketing, sales, and product teams. Walk through the guide, answer questions, note edge cases that come up. Record the training and post it. New hires get the recording as part of onboarding. This one investment reduces the ongoing question load on the brand team by half or more.
The two-week feedback window
Right after launch, open a two-week window for feedback. Not a permanent open-door — a specific window with a clear close date. Users flag confusing sections, missing patterns, and misalignments with their team's actual work. The brand team collects the feedback, resolves what can be resolved quickly, and defers the rest to the next scheduled update.
The 90-day check
Ninety days after launch, run a check on adoption. Sample recent work — sales decks, marketing collateral, product screens, external-vendor output — and score against the guide. Where the drift is high, either the guide is missing something or the training didn't stick; investigate. The 90-day check is the difference between a guide that launched and a guide that landed.
Relaunches — usually smaller than expected
Relaunches are more common than full rewrites. A well-maintained guide might go through a small relaunch every year — updated examples, refined voice pairs, new patterns for a product that launched. Full rewrites are rare and usually tied to a broader brand evolution. Communicate small relaunches quietly (email, updated version banner); communicate full rewrites like a fresh launch.
The sunset moment
Sometimes a guide is beyond patching and needs a full replacement. Signs: multiple sections describe brand elements the company no longer uses; the guide reads like a different company; adherence has collapsed. Rather than patching, plan a full rewrite and treat it like a brand-evolution project — audit first, decide, then write.
Composite case studies — five guides in the field
The following patterns come from about thirty brand-guide reviews Inkgility has supported over the last three years. Industries and shapes are grounded in real work. Names and specific figures are composited to protect confidentiality.
Alderwood Coffee — the guide that got shorter
Alderwood, a small specialty roaster, had a sixty-page guide that nobody used. The rewrite cut it to twenty-eight pages, kept only the sections that answered real questions (voice, packaging behavior, wholesale surfaces), and added a two-page operational-surface section. Twelve months later, packaging vendors were producing consistent output on the first try, and the operational surface (line sheets, invoices, tasting notes) had visibly caught up to the marketing surface.
Northline Realty — the guide that added print behavior
Northline, a mid-size real-estate brokerage, had a beautiful digital-first guide that treated print as an afterthought. Yard signs, direct mail, and open-house collateral were drifting badly. The revision added a print-behavior section with defaults, stock recommendations, and vendor-facing specs. Eighteen months later, the field agents' print output was noticeably more consistent, and complaints from the design team about "why does the yard sign look different from the flyer" dropped to near zero.
Halcyon Studio — the guide that added governance
Halcyon, a design studio, had a technically excellent guide with no governance model. When the founding designer left, the guide immediately started to drift. The rewrite added a named owner (rotating annually), a quarterly review cadence, and a lightweight intake process. Two years later, the studio's brand output was more consistent than before the founding designer left, and the guide had gone through six documented updates without a rewrite.
Driftwood Legal — the guide that added voice
Driftwood, a mid-size law firm, had a strong visual guide and no voice guide. The result: partners' bios, marketing copy, and client memos all read like different companies. The addition of a five-page voice section with do-and-don't pairs, delivered as part of a firmwide writing workshop, brought voice consistency up sharply within a quarter and reduced the marketing team's revision load on partner-drafted content substantially.
Meridian Dental — the guide that got a governance owner
Meridian, a three-office dental practice, didn't have a formal brand guide at all — the founder made every branded decision. The engagement produced a lightweight fifteen-page guide plus a governance model where the practice manager owned brand decisions with monthly founder review. Twelve months later, the founder was making about a fifth as many brand-adjacent calls as before, and patient-facing surfaces were noticeably more consistent across offices.
Common threads
Every one of these engagements produced a shorter, better-governed guide than the one that came before. The impulse to "make the guide comprehensive" produces guides nobody reads. The disciplined move — decide what the guide is for, write only what answers real questions, and put governance around it — produces guides that get used. In our observation, the most successful brand guides are usually the second version at the company, written after the team knows what the first version was missing.
The other pattern: operational-surface guidance is the single most-consistently-omitted section, and adding it produces the largest visible change per page written. If you have to pick one section to add to an existing guide, add operational-surface guidance first. Marketing surfaces mostly follow the visual system by osmosis. Operational surfaces almost never do without explicit rules.
