Operations
Why an unconfigured multilingual campaign earns a hold verdict at launch review
Review an unconfigured multilingual campaign against evidence, language, access, permissions and recovery gates, with a transparent desk-only HOLD verdict.
What to take away
- The hold verdict is a documentary finding. The review saw a concept, not a delivered campaign, so release readiness could not be established.
- Four evidence gates appliedsource and language, audience and permissions, rendering and account, claims and corrections.
- CAP records, RM6302 and GOV.UK accessibility guidance set the standard but cannot fill a missing campaign file.
- A new review needs a versioned launch pack with named owners, dates, versions and an openable location.
- Name the reviewers by gate before resubmittinglanguage, privacy and PECR, advertising, accessibility.
The gap between a concept and a campaign
The reviewing body was the Bilingualbrand.co.uk editorial team. Desk research closed on 6 September 2026, and no independent reviewer has been named.
The desk worked from the D040 operations brief, an internal task list that sets what the review checks and who signs it off. It also read public records from the Government Commercial Agency, the ICO, the Committee of Advertising Practice, NCSC and GOV.UK.
The concept named no versioned source, no substantiation file, no language variety and no service type. There was no enrolment, no account access, no translation, no comprehension test, no audience upload, no post, no advertisement and no recovery drill.
So the verdict is hold, not fail. A fail would judge work that ran. A hold records that nothing submitted could be judged. A similar desk review of an agency ended in the same place, as why a multilingual social agency review stays on hold explains.
Evidence gates a campaign must clear
| Gate | Evidence the review needs | Why the concept failed it |
|---|---|---|
| Source and language | Approved source, rights record, named translation or localisation instruction | No source claim, no rights, no language decision |
| Audience and permissions | Defined population, recruitment record, marketing route, preference process | No audience record, and preference cannot be inferred from a name or profile |
| Rendering and account | Accessible final render, client-owned account, access matrix, exit route | No asset, no platform, no account, no recovery evidence |
| Claims and corrections | Substantiation file, correction and withdrawal drill result | No claim test, no drill result |
CAP's rules on misleading advertising require documentary evidence behind objective claims and clear qualifications. They set the standard for the source gate. They cannot fill a substantiation file.
RM6302 is the Crown Commercial Service agreement UK public bodies use to buy translation and interpreting. It shows how language work is procured, and supplies no campaign evidence.
A claim travelling from English into two or three languages needs its own audit trail. The advertising checks from source claim to correction follow that trail line by line.
Example: a versioned launch pack for a bilingual campaign
Take an English and Polish campaign for a UK retailer. A pack that could pass a new review holds eight files in one shared folder, each with an owner and a date:
A versioned launch pack
- Source substantiation file, version 3, with a 12 month expiry.
- Rights note covering the claim and the image assets in both languages.
- Polish target asset approved by a native-level reviewer and a subject specialist.
- Accessibility check on the final render, covering captions, alt text and contrast.
- Audience and permission record, held separate from research consent.
- Platform source page and account-access matrix for the client-owned account.
- Publishing preview with the scheduled date and the reviewer signature.
- Correction and withdrawal drill result, plus the recovery steps taken.
Reviewers who belong on the sign-off list
Multilingual social work needs fewer sign-offs than the earlier list implied. Name the translation lead and a native-level language reviewer. Add a subject specialist for any regulated claim, then a privacy and PECR reviewer.
An advertising and consumer reviewer covers claim wording, and an accessibility reviewer covers the final render. The client account owner signs the access matrix and the exit route.
The ICO's direct marketing guidance covers planning, fair collection and respect for preferences. A named privacy reviewer must assess the actual data and channel before marketing begins.
Public records help size a field, and the method for researching multilingual competitors from public records shows where that reading should stop. NCSC's social media safety guidance recommends two-step verification and a recovery plan, taken from the real account.
Checks to complete before resubmission
- Approve the source and confirm rights in writing.
- Name the language instruction and the target-asset reviewer.
- Record audience and permission decisions apart from research consent.
- Supply accessible final-render evidence.
- Identify the client-owned account, its owners and the exit route.
- Run the correction, withdrawal and recovery drills, then file the results.
Common questions
Why does an unconfigured campaign get a hold rather than a fail?
Nothing was submitted that could be tested. There was no approved source claim, target-language asset, audience record, platform or account, so release readiness could not be established.
What must be in the launch pack for a new review?
Source substantiation, rights, an approved target asset, language decisions, accessible render evidence, an audience and permission record, a platform source page, an account-access matrix, a publishing preview and a monitoring plan, plus the drill results.
Which reviewers must be named?
A translation lead, a native-level language reviewer, a subject specialist where claims are regulated, a privacy and PECR reviewer, an advertising and consumer reviewer, an accessibility reviewer and the client account owner. A named independent reviewer is still to be assigned.
Does the hold block the campaign for good?
No. The hold lifts when the artefacts exist and a new review reproduces the synthetic test, then inspects the scheduled and live routes. Until then, the draft hold and the operational hold remain the supportable conclusions.
