Set up your client project: market rules, approved claims, and the do-not-translate list
Run one project per client and the second market costs a fraction of the first. The things that decide what can run are stable per brand, so add them once and every market review inherits them. The brand lexicon in particular only gets written down once, and it is the item that stops a trademarked line being helpfully translated into something nobody owns.
What to add
Market rules and prior clearances
What local counsel has already ruled on, which packaging and deposit rules apply, and any platform or broadcaster policy the brand has run into before. Past clearances stop the same question going back to legal twice.
Approved claims and their evidence
The claims the brand is allowed to make, the comparator each one rests on, and where the substantiation lives. A claim without its evidence is a hold, and the register will say so.
Brand lexicon and do-not-translate list
Product names, trademarked lines, loyalty terms and the words that stay in English on purpose. Without it, a reward currency or a tagline gets translated and the equity goes with it.
Local reviewers and sign-off owners
Who clears copy, who clears claims, and who clears promotion mechanics in each market. The register routes to roles, and the roles need names behind them.
Guide Juma with project info
Each knowledge item takes a short description. One line is enough, and it tells Juma when to reach for that item instead of guessing.
- Market rules: "Local rules and past clearances per market, including packaging, promotions and platform policy."
- Approved claims: "Claims we are allowed to make, the comparator behind each one, and where the evidence sits."
- Brand lexicon: "Names, taglines and terms that never get translated, with the reason each one is protected."
- Sign-off owners: "Who clears copy, claims and promotion mechanics in each market, by role and by name."
Find out what breaks before you translate it
Frequently Asked Questions
What does the localization review include?
Two files. A branded PDF covering the market take, the lines that cannot travel literally, the counts behind each verdict with the asset IDs listed so they can be checked, and the routing for everything that needs clearing. Alongside it, an editable CSV that is the authoritative register.
Each register row carries the asset, its channel and placement, the original wording verbatim, the source URL and the date it was observed, the verdict, what fails and why, the local copy, a literal English back-translation, what the meaning change was, the rule family, the owner role, a status, a decision date and an evidence link.
How is this different from translating the campaign?
Translation moves the words. This moves the intent and tells you where that was not possible. A line can be translated perfectly and still fail, because the joke's register shifts, the cultural reference is local to the first market, or the claim is allowed in one place and not another.
The audit runs before the rewrite for that reason. Each line gets a verdict and a stated reason first, so the version that comes back is traceable to a decision rather than to a translator's instinct.
Can we review copy in a language nobody on the team speaks?
That is what the back-translation column is for. Every local line comes with a literal English rendering, and a separate column stating what changed in meaning, or none.
Read the back-translation and you can see a pun collapse, a joke invert, or a claim get stronger than the original allowed. Human review on every output, and the back-translation is what makes that review real rather than a formality.
Does this handle the legal side?
It prepares the legal side; it does not replace it. Nutrition, environmental, comparative and promotional lines are pulled out of the creative queue, labelled with the rule family they fall under, and held when the evidence or the comparator is missing.
Local counsel still makes the call. What the register changes is that nothing arrives on their desk unlabelled, and no claim ships because it was mistaken for a joke. Where a rewrite would introduce a new claim, the review says so rather than presenting it as a safe fix.
Does it work for markets outside Europe, and for B2B?
Yes. The structure holds for any market move: the verdict vocabulary, the back-translation, the claim separation and the register are market-agnostic, and what changes is the rule family each row is routed against.
B2B works the same way with a different risk profile. The jokes matter less, and the regulated claims, the role and seniority conventions, and the contractual or procurement language matter more. Name the market and the audience in the first message and the audit weights itself accordingly.