Brand governance without the brand police

Every company past thirty people has a shadow brand team. You've met them: the sales lead with a personal fork of the deck, the recruiter running a careers page in eight unapproved fonts, the regional office whose logo has acquired a drop shadow. The standard response is to hire somebody to say no — a brand police force with a ticket queue and a growing list of enemies. It works for about four months. Then the tickets pile up, the police become the bottleneck, and sales goes back to their fork because the fork ships faster.

Governance fails when it's enforcement. It works when it's architecture: decision rights everyone can recite, assets that are easier to use than to avoid, and an exception process that treats rule-breakers as data. Here's the model we install after brand programmes — the one that survives contact with a growing company.

Start with decision rights, not rules

Most brand guidelines document rules: clear space, minimum sizes, forbidden gradients. Rules without decision rights are just complaints with page numbers. The question governance must answer first is who may decide what, and the answer should fit on an index card.

We use three lanes:

Green — anyone may do this. Using the colour palette, picking from the type scale, choosing photos from the library, writing in the brand voice. Green decisions need assets and examples, not permission.
Amber — the brand steward reviews, asynchronously. New page templates, co-branded materials, event booths, anything external that doesn't clone an existing pattern. Target turnaround: two working days, written feedback, no meetings.
Red — the creative director decides, and it's rare. Logo changes, new marks, naming anything, partnerships that share the masthead, a campaign that repositions. Red should happen a handful of times a year.

The index card test: if a new hire can't classify a decision into a lane after reading one card, your governance is a vibe, not a system. Write the lanes down, publish them where the assets live, and make the steward's name and response-time promise public. Accountability in both directions is what makes people use the queue instead of the fork.

Make the right thing the easy thing

Here is the uncomfortable data point from every governance project we've run: almost nobody breaks brand rules maliciously. They break them because the compliant path is slower than the non-compliant one. The deck fork exists because the official template was a 200-slide museum nobody could edit. The rogue logo has a drop shadow because the only available file was a PNG with a white background on a dark slide.

So the first deliverable of governance isn't a ruleset — it's a usable asset system:

Templates people actually present from. One deck template with twenty well-designed slide types beats five templates with two hundred. Ship it in the tool they use (yes, that tool), with the tokens baked in.
Logo files for reality. Every colour mode, every background, SVG and PNG, named so the right choice is obvious: brassfern-mark-ink-on-paper.svg, not logo_final_FINAL_v3.svg. The filename is governance.
A living home. A linkable, searchable guidelines site, not a PDF — we've written at length about why living guidelines win, and governance is where the argument pays off: you can't have decision rights pointing at a document that's eleven months stale.
Copy blocks. Approved boilerplate, bios, product descriptions and the voice chart examples, copy-paste ready. Half of all brand violations are people rewriting the company description from memory.

Measure the system by its adoption, not its violations. If 80% of external material flows through the assets without a ticket, governance is working — even if the other 20% is chaos. Fix the chaos by improving the assets it routes around.

The exception process is the product

Bold rule: every exception request is a design review of your system, not of the requester. When someone asks "can I do X that the guidelines forbid?", they're handing you free research. Three outcomes are possible:

The rule is right; the requester lacked an asset. Fix the assets. This is most cases: the social team needed a motion lockup that didn't exist, so they made one.
The rule is right; the requester's case is genuinely novel. Grant a scoped exception, in writing, with an expiry date. "This campaign may use the alternate lockup until March; the system absorbs it or retires it then."
The rule is wrong. The world changed — a new platform, a new product line, a new audience — and the guidelines haven't. Update the system and credit the exception that exposed it. This is how living guidelines stay alive.

Publish the exceptions log internally. Nothing builds trust in governance like seeing that "no" comes with a reason, a fix and sometimes a rule change. Nothing destroys it like a graveyard of unanswered tickets.

The steward is a gardener, not a guard

The brand steward role — one person, part-time at most companies — makes or breaks the whole model. Hire or appoint for judgement and generosity, not seniority. The steward's job description in one line: keep the green lane wide and the amber lane fast.

Concretely, we set the role up with four weekly habits:

Office hours, thirty minutes, drop-in. Most amber decisions die in office hours with a template suggestion. Meetings prevent meetings.
A ticket review with a clock. Every amber request answered in two days or escalated with an explanation — silence is the fork generator.
A monthly asset release. New templates, new lockups, retired files. Small and regular, like a product changelog. Governance with a release cadence feels maintained; maintained systems get used.
A quarterly walkabout. Review a sample of what's actually shipping — decks, posts, event materials — not to punish but to spot where reality and the system have drifted. Drift is information.

Note what's absent: approval power over green-lane work, and any mandate to "ensure consistency". Consistency is the outcome of a good system plus easy assets. Making a person responsible for it makes a person miserable.

Governance for the parts of the brand you can't template

Colours and logos are the easy half. The hard half of governance is voice, photography and motion — the expressive layers where templates barely reach. Here the enforcement instinct is most dangerous, because it's where the brand's life is. Our answer is to govern with ranges rather than rules: publish the acceptable range (voice: "confident to playful, never snark; photography: natural light, real spaces, no staged handshakes") and a strip of examples showing the edges. Examples teach judgement; rules teach compliance. You want judgement — the alternative is a brand that ossifies while the company moves. When we rebuilt Copperline Mutual's verbal identity, the entire governance layer for their tone was six "we say / we don't say" spectra pinned where their writers work. Two years on, it's still the most-used page in their guidelines.

What this costs and when to build it

A working governance layer — lanes, asset system, steward habits, exception log — is a few weeks of setup plus a steward's quarter-time ongoing. Build it when the shadow brand team appears: the first forked deck, the first off-brand region, usually somewhere between thirty and sixty people. Earlier than that it's theatre; later than that it's archaeology.

And if you're reading this from inside a company that already has brand police: the way out isn't better policing. It's a better green lane. Give them the assets, give them the lanes, and watch the ticket queue drain itself.

Key takeaways

Governance is architecture, not enforcement: three decision lanes on an index card beat forty pages of rules.
People break rules because the compliant path is slower. Fix the assets before you blame the users.
Treat every exception as a design review of the system; publish the log with reasons and expiries.
The brand steward is a part-time gardener with office hours and a release cadence — never a bottleneck with a badge.
Govern voice and photography with ranges and edge examples, not rules. You want judgement, not compliance.

FAQ

How is this different from just having good guidelines? Guidelines describe the brand; governance describes the decisions. Guidelines answer "what does our blue look like?" Governance answers "who decides whether this partnership page can use it, and by when?" You need both, and the guidelines themselves should be living pages the governance links into.

What if leadership won't fund a steward? Then governance is a shared responsibility with explicit rotation, or — more honestly — you don't have governance, you have hope. The quarter-time steward costs less than one decent off-brand reprint or one wrong-logo campaign. Run the forked-deck audit and the arithmetic usually settles itself.

How do we handle agencies and freelancers inside this model? Put them in the green lane with the asset system on day one of onboarding, and require amber for anything net-new. External partners cause drift not from malice but from missing the index card — so make the card page one of every brief.

Our brand is changing soon; should we wait? No — install governance now and it becomes the machine that rolls the change out. Rebrands fail at the seams between old and new, and the rollout middle is exactly where a lanes-and-steward system earns its keep.