A skill that makes Claude plan a rebrand or refresh the way a senior brand strategist would, before anyone designs anything. It writes the problem as one sentence with a number, maps who to interview and writes their questions, then judges every asset the brand owns on evidence: keep, refine, drop or add. It places the work on a six-step ladder from hold to new name and writes a one-page decision memo. Then it counts every touchpoint, builds the rollout timeline backwards from your fixed dates, writes what staff, partners and customers are told, and ends with a risk register and the measures to read for 90 days.
Inside are five reference files. One holds question sets for eight stakeholder groups, with survey sample sizes. One covers the fame and uniqueness test for brand assets, the verdict rules and the decision matrix. One lists every touchpoint, the rollout models and the US rules that shape a rollout (USPTO filings and fees, Google site moves, Gmail sender rules, app store names, GS1 barcode triggers), each sourced and checked on 17 October 2026. One covers the launch story, 21 common risks and the go or no-go check, and the last is a full plan for a fictional granola brand. It works one stage at a time, never invents survey results, quotes or lead times, and never calls a name clear or safe.
It plans and does not draw. The signed memo goes to a designer or to your identity skills. It cannot see the trademark register, your analytics, your stock or your suppliers' quotes, so you paste those, and the lead times in its files stay planning ranges until you replace them with real quotes. Legal questions go to your attorney, and the rules it quotes cover the US. With code execution on, it computes colour contrast, counts themes in a review export and returns the scope, timeline and risk register as a spreadsheet. Bring your guidelines, pack photos, reviews and trademark numbers, and it will ask far less.
Claude skill · version 1.0
Rebrand planner
Checking your access.
In the download · 7 files
- SKILL.mdThe method Claude follows: brief, listen, equity, decision, scope, rollout, launch story and risks, one stage at a time, with an output template for each.
- README.mdHow to install it in claude.ai or Claude Code, how to start, and the licence.
- references/stakeholders-and-questions.mdThe six stakeholder roles, how many people to talk to, question sets for eight groups, mining reviews and tickets, a recognition survey with sample sizes and margins of error, and the synthesis template.
- references/equity-and-decision.mdWhat counts as a brand asset, the fame and uniqueness test, proxies, verdict rules, the production and WCAG 2.2 contrast check, the six-step ladder, the decision matrix, three sourced reversals and the decision memo.
- references/scope-and-rollout.mdThe touchpoint inventory, tiers, hidden items, design deliverables, rollout models, the critical path, and the US trademark, tax, domain, email, app store, listing and barcode rules, sourced and checked on 17 October 2026.
- references/launch-story-and-risks.mdWhat customers want to know, the one-sentence reason, who hears it when, the core story and its versions, rules on "new" claims, renames, 21 common risks, scoring, the go or no-go check, rollback triggers and 90-day measures.
Install it and begin
01Install it in Claude CodeTerminal mkdir -p ~/.claude/skillsunzip ~/Downloads/rebrand-planner.zip -d ~/.claude/skills/# Start it with /rebrand-planner. If this is your first skill, type /reload-skills once.On claude.ai there is nothing to unzip: upload the zip as it is. Every step is on Install and use a Claude skill.
02Your first message to ClaudeText Use the rebrand planner skill. We are [BRAND NAME], selling [WHAT YOU SELL] in [COUNTRY AND CHANNELS]. A rebrand is on the table because [WHY NOW, IN ONE LINE]. [WHO DECIDES] signs it off, and our fixed date is [DATE OR NONE]. Start with the brief and tell me what is missing.Attach your current brand guidelines, photos of every pack or sign, and any reviews or support tickets you can export. Trademark registration numbers and a rough split of the budget between design and rollout help the later stages.
Common mistakes
Where rebrands go wrong, and what the skill does instead
Instead ofOpening a mood board because the founder says the brand feels dated.
Do thisWrite the problem as one sentence with a number you can check. If the only reason a customer would hear is "a fresher look", plan a refresh at most.
Instead ofDropping the mascot, colour or pack shape customers use to find you, because the team is tired of it.
Do thisTest recognition with the name hidden, or count how often reviews mention the asset. Short of a rename, keep the two strongest assets recognisable, even if both are redrawn.
Instead ofRevealing a new name, then asking a lawyer about it.
Do thisHave an attorney search and file before anything goes public, and buy the domain and handles before the name leaves the room.
Instead ofBudgeting for the logo and forgetting the hundreds of places it lives.
Do thisCount every touchpoint (SKUs by sizes by languages, signs, templates) and tier it: right on launch day, within 90 days, or at natural replacement. The rollout usually costs more than the design.
Instead ofDropping a word from the brand name on pack without asking about barcodes.
Do thisA change to the primary brand on pack needs new GTINs at unit, case and pallet level, and new item set-up with every retailer. Check with GS1 US and each buyer before the artwork is final.
Instead ofAnnouncing "a bold new chapter" and calling an unchanged product "new".
Do thisTell customers what stays the same first: the recipe, the price, their account. Say "new look" when only the pack changed.
Instead ofLetting staff and retail buyers find out on Instagram.
Do thisPartners hear months ahead under confidentiality, the whole team one to two weeks before, and loyal customers on the morning of the reveal.
Instead ofLetting the old domain lapse after a rename.
Do thisKeep permanent redirects for at least a year, as Google advises, renew the old domain for years ahead, and hold the old social handles as placeholders.
What to do it with
- ClaudeFree tier, paid above it
A language model that writes, edits, reads long documents and runs tools, including agents that work inside a codebase.
Bad at It is agreeable by default, so it will improve a bad idea instead of telling you it is bad. Ask it to argue against you and the quality of the answer changes.
- Claude CodeSubscription
An agent that reads a whole repository, edits files, runs the tests and explains what it changed.
Bad at It is fast enough to make a large mistake quickly. Work on a branch, read the diff, and never let it push. Watch the starting mode: on Pro, Max and Team plans it opens in auto mode, which Anthropic's own list says may read a .env file and send the key in it to the API that key belongs to. Switch it off.
- Wayback MachineFree
Shows what a brand's site looked like at any point in its history.
Bad at Captures are patchy and heavy pages often archived broken.
Why it matters
Left to itself, a request to rebrand turns into logo options. Most rebrands that go wrong fail somewhere else: the team drops the one asset customers used to find them, forgets half the places the brand lives, files the name after the reveal, or tells customers nothing about what stayed the same. This skill makes Claude decide how far to go on evidence, protect what people recognise, and plan the rollout and the launch story before design starts.
- Type
- technique
- Period
- ongoing