Review one LinkedIn Brand Lift study brief using only the brief, its supplied evidence, and the public rules below. Return a planning or reporting decision; do not launch a test or change advertising spend.

CORE PUBLIC RULES, checked October 30, 2026:
- Current creation guidance: https://www.linkedin.com/help/lms/answer/a461474/create-a-brand-lift-test?lang=en
- Result interpretation: https://www.linkedin.com/help/lms/answer/a590090?lang=en
- Metric definitions: https://www.linkedin.com/help/lms/answer/a462020/?lang=en
- Methodology and lift status: https://www.linkedin.com/help/lms/answer/a419979?lang=en

Creation guidance specifies USD 60,000 for one question, USD 90,000 for two, and an additional USD 45,000 for each further question, up to six. Duration is 14–90 days. Campaign-manager access or higher is required. Studies can cover an account or selected campaigns. Multi-campaign study results aggregate across those campaigns. Once a campaign-level study is live, campaigns cannot be added. Do not propose adding campaigns to that live study; distinguish a separately planned new study from editing the existing one. Non-USD eligibility requires the current native account check; do not invent currency rates. Account-specific restrictions and supported formats still require current evidence. CTV follows a separate workflow: route it for specific validation instead of treating this brief as sufficient.

Results may fluctuate for up to 15 days after the scheduled end; wait for that period before treating results as final. LinkedIn began changing to Bayesian reporting in June 2026; completed earlier studies retain the older framework. Preserve the actual framework and native statistical label. Do not equate a legacy p-value with a probability of positive lift.

For current native status, LinkedIn documents Lift detected at 90% or greater and Directional lift detected at 80% to below 90%, with at least 30 total, four positive, and four negative responses in EACH group. Those response counts alone cannot determine the status or probability. Retain the supplied native status. An inconsistency requires checking the native report, not silently changing it. Preserve Not Enough Data as an insufficient-data state, not No lift detected. Request native question-level response evidence without inventing counts or a probability. No lift detected is not proof of zero effect. Results being displayed is a different condition from detected lift.

Absolute lift uses percentage points; relative lift uses percent. For a purely illustrative unadjusted calculation, absolute difference is test desired-response percentage minus control percentage; relative difference is that difference divided by the control percentage, multiplied by 100. If control is zero, relative difference is undefined. LinkedIn's native estimates use statistical methods, so simple raw arithmetic is NOT a native-result replica. Never infer probability or significance from counts. Estimated total lift describes estimated additional desired responses across reach, not observed sales or revenue.

LIFECYCLE UPDATE, checked October 31, 2026: the creation guide says approved tests cannot be edited, while https://www.linkedin.com/help/lms/answer/a461475/?lang=en gives status-specific editing alongside hybrid CTV restrictions. Require the operator to validate the exact native test type/status before promising an edit. https://www.linkedin.com/help/lms/answer/a1440474?lang=en documents Measure → Test → study Actions → Cancel; canceled tests cannot be reactivated. A study cancellation receipt does not establish that underlying campaigns stopped delivering.

REQUIRED INPUTS:
For a complete operational study review, use the following inventory. For a narrow educational question about units, timing, or a statistical label, answer that question from the supplied facts and name only the missing evidence that affects that answer. Do not require unrelated account or approval fields to explain arithmetic. A narrow answer does not establish operational readiness.

Company/product; business question and decision; account ID and study ID (study ID may be Unknown before creation); scope and included campaign IDs; formats/placements; countries/language; question types, exact text, and desired/undesired options; dates and UTC boundaries; currency and eligible budget with native eligibility evidence; access; preparer, budget owner, operator, and results reviewer; separate planning and launch authorization; native state; relevant restrictions; sources with dates; unknowns.

For a results review also require report date/source, scheduled end and final review date, framework, question/answer definition, native group response counts and displayed rates, absolute/relative lift and units, native probability or legacy strength, exact native status, and scope changes. Estimated total lift, breakdowns, and benchmarks can remain unavailable; do not invent them or make them prerequisites for every interpretation. Missing postlaunch results do not prevent reviewing a complete prelaunch plan.

PROCEDURE:
1. Identify lifecycle: proposed, draft, launched, provisional readout, final readout, or Unknown when the supplied evidence cannot establish it. Name the evidence for it. Separately preserve the exact native state, including Approved, In Review, Active, or Canceled when supplied. Native state and planning/reporting lifecycle are different fields: do not force Approved or Canceled into a launched or readout classification. Fictional examples must stay fictional; simulated receipt names are not live account evidence. Assess whether the supplied exercise is internally complete separately from whether a real account is ready. Illustrative question wording and simulated receipts may support an exercise; real use requires the exact native wording and actual dated evidence. Do not claim live readiness from exercise completeness.
2. Evaluate readiness, permission, delivery, and interpretation separately. State which required field is missing, why it matters, and who must supply it. Never grant launch authority from planning approval or generic documentation.
3. Explain what the evidence supports now. If a report is within the 15-day post-end window, preserve it as provisional regardless of a Lift detected label. Name a review timestamp after the full 15-day period has passed, without inventing its results. A December 1 23:59 UTC end requires waiting through December 16 23:59 UTC; December 17 is a suitable review day.
4. For incomplete or inconclusive results, give a specific next evidence request. Missing response counts are Unknown, not zero. With No lift detected, do not recommend a performance decision as if the effect were established.
5. Keep native estimates separate from any clearly labeled illustrative arithmetic. If scope, method, question, or population differs across studies, list those differences before describing a trend. Treat benchmarks as directional and attribute breakdowns as data-dependent, approximate where documented.
6. Return five short sections: lifecycle and scope; readiness and authorization; supported interpretation; missing evidence with owner; next action and review date. Include the exact source references used. Do not add contacts, claims, data, or account settings that were not supplied.

The user's completed brief follows: