A brand audit checklist is the specific list of what to compare against the guide: the written rules, the logo files in circulation, the colors and type in recent work, and increasingly, what AI tools have produced on the brand's behalf. It is not the guide itself, and it is not a daily habit. It is a point-in-time check that catches what ongoing brand compliance missed between reviews.
Most audits start the same way: someone opens a blank spreadsheet, writes "check everything" at the top, and stalls two sections in because nobody defined what everything means. A sales deck is still running last year's logo mark. Nobody can say whether the hex value in the template matches the one in the guide, because nobody has looked at both side by side since the rebrand. What follows is a checklist built around where brand work actually drifts, plus where to put what you find so the audit still matters in six months.
Key Takeaways
- A brand audit checklist is a point-in-time comparison between the guide and the work, not the guide itself and not the day-to-day compliance work tracks.
- It covers four areas: the written rules, the visual assets in circulation, the work already shipped, and now, AI-generated drafts.
- Checking color means comparing against the hex values in the guide, not a description of them, and testing contrast against the WCAG 2.1 minimum wherever text sits on a brand color.
- AI-generated work needs its own line on the checklist, since one person can now produce a dozen drafts before anyone reviews a single one.
- Findings belong in one log tied to the guide, not a spreadsheet that gets opened once and forgotten.
What a brand audit checklist actually checks
A brand audit measures whether the guide, the files, and the work still agree, at one point in time. That differs from the guide, which states the rules, and from brand compliance, the ongoing state of work matching those rules day to day. An audit is how a team finds out compliance has drifted; compliance is what the audit is checking for.
The checklist exists so the audit does not depend on whoever happens to run it that quarter remembering what matters. Four areas cover most of what goes wrong: the rules as written, the assets in circulation, the work already made, and the newest addition, anything an AI tool produced without anyone checking it against the brand first.
The written rules: guidelines and voice
Start with the guide itself, before looking at anything it is supposed to govern.
- Is the guide the current version, or is a prior draft still sitting on a shared drive with more recent edits?
- Does the banned-words list still match how the team actually writes, or has the tone shifted since the last update?
- Do the last three campaigns follow the voice as written, or has one channel quietly developed its own?
A rule is only useful if it is specific enough to check. "Sound confident" fails an audit, because two reviewers will disagree on whether a given sentence passes. A rule like the one below does not:
Never in a headline: "industry leading," "next generation," a percentage claim.
Headlines: sentence case, no exclamation marks.
If the guide reads more like the first kind of rule than the second, rewriting it belongs on the audit's findings list, not just the checklist itself.
The visual assets: logos, colors, and type across what shipped
Pull five to ten pieces of recent work: a deck, a landing page, a few social posts, an email. Check each against the guide:
- Is the correct logo lockup in use, including on dark backgrounds, where a light mark is often swapped for the full-color version by habit rather than rule?
- Do the colors match the guide's hex values, not just "close enough" by eye?
- Where text sits on a brand color, does the pairing clear the WCAG 2.1 contrast minimum, or did a designer pick a combination that looks fine on a bright monitor and fails for everyone else?
- Is the typeface the one named in the guide, or a close system-font substitute because the real one was not installed?
This step goes faster when the guide shows the real thing rather than describing it: colors as swatches, logos at presentation quality, type in the actual typeface. A description invites someone to eyeball a comparison; the real thing does not.
The newest line on the checklist: AI-generated work
A few years ago, this section of the audit did not exist. Now it is often where the drift is largest, because AI made producing brand work fast and cheap without making any of it automatically on-brand. A marketer can generate a dozen headline options before lunch, and none of them has seen the guide unless someone built that connection on purpose.
Add these lines:
- When Claude, ChatGPT, or Cursor is asked to write or design something for the brand, does it have access to the brand's rules at all, or is it working from a generic prompt and the model's own instincts?
- If it does have access, is that access current, or was it set up once and never updated when the guide changed?
TrueKeep answers this by writing the brand's decisions into a design brief and serving it through an MCP connection. Model Context Protocol is the open standard behind that connection, and it is how Claude, ChatGPT, or Cursor read the brief before they draft, once someone connects them through the AI connection guide. A final pass then checks the output against the brand and fixes what is off before it ships. An audit should confirm that connection exists and still points at the current brief, not assume it from a setup meeting six months back.
Where to log what the audit finds
A spreadsheet started for one audit rarely survives to the next one. Findings are more useful living next to the guide they are measuring, in the same place the team already works from, because that is where the next person doing the audit will look first.
Three kinds of findings need three different homes:
- A one-off exception (this specific deck used the wrong lockup, and it is already fixed) goes in a short exception log, dated, so the same question does not get re-debated later.
- A pattern (every social post this quarter ran a slightly different blue) belongs in the guide itself, as an update, not just a note that it happened.
- A missing rule (nobody had written down what to do when the typeface is not installed) belongs in the design brief, so the next AI-generated draft gets it right without anyone catching it in review first.
This is closer to the job brand governance does day to day than to a one-time cleanup. An audit without a governance model behind it just produces the same findings every time, because nothing closes the loop between the audit and the guide.
How often to run it, and what it costs to skip
Quarterly works well for most small teams. It catches drift before it compounds, without turning the audit into a second job. A company mid-rebrand or in a fast hiring stretch should run it more often for the first two quarters, then settle back to that pace.
The tradeoff runs in both directions. Audit every month, and a small team spends more time checking the brand than making anything with it. Audit once a year, and by the time anyone looks, the sales deck, the onboarding emails, and the AI-generated social posts have all drifted in slightly different directions, and the fix touches more than the guide. Pick a cadence the team will actually keep, and put it on the calendar instead of leaving it to whoever remembers.
FAQ
What is a brand audit checklist?
A brand audit checklist is the specific list of items to compare against the brand guide at one point in time: the written rules, the logo files and colors in circulation, recent work, and what AI tools have produced on the brand's behalf. It exists so an audit does not depend on whoever runs it remembering what to check.
How is a brand audit different from brand compliance?
An audit is a snapshot; compliance is the ongoing state the audit measures. A team can run a clean audit in March and still drift by June if nothing keeps the guide and the work in sync between audits, which is the job brand compliance is built to do.
How often should a team run a brand audit?
Quarterly works for most small teams. A company mid-rebrand or hiring quickly should run it more often for a couple of quarters, then return to a quarterly cadence once the rate of new work settles down.
What should a brand audit checklist include for AI-generated content?
A line confirming that any AI tool used for brand work, Claude, ChatGPT, or Cursor, has current access to the brand's rules, usually through a design brief served over an MCP connection, and that this access was updated when the guide last changed, not set up once and left alone.
Who should run a brand audit?
One named owner, usually the same person responsible for keeping the guide current day to day. A committee can review the findings, but the audit itself goes faster and gets finished when one person owns running it start to finish.
Where to start
A brand audit checklist works when it checks four things in order: the written rules, the visual assets actually in circulation, the work already shipped, and now, what AI tools have produced without a human reviewing every line. Running it once a quarter catches drift while it is still a quick fix rather than a rewrite.
Log what you find next to the guide it is measuring, not in a spreadsheet nobody will reopen. If the audit turns up rules nobody wrote down or an AI connection nobody set up, get started and put the guide and the design brief in the one place both will actually get used.