The fields that keep the work understandable.
Which creative choice deserves a closer look?
A question gives the collection a reason to exist.
Accounts, category and reason for each inclusion.
The comparison needs a bounded audience and context.
Dates, metric and source availability.
A comparison cannot be reconstructed without its timing.
One statement, with relevant source examples.
The reader should be able to challenge the reasoning.
Record missing context and other possible explanations.
An observation does not establish causation.
One change, stable variables and a review owner.
The brief should lead to a useful team decision.
A real planning file. No email or account required.
Start with a decision, not an empty document
Use this brief when the team has a specific recurring job to review. Read the field prompts before filling them in. Each prompt separates one part of the work that is easy to lose in a chat thread: the question, the evidence, the interpretation or the person making a decision. Leave a field explicitly unknown when the information is unavailable. A completed-looking document is less useful than an honest one with a clear verification task.
Keep the example separate from your own record
The prompts describe a planning structure, not customer evidence. Copy the downloaded file and replace the prompts with your own permitted source references and review notes. Do not put passwords, private contact information or assets you are not allowed to share into a research handoff. Record the observation window and baseline definition beside each metric.
Review the record before handing it over
Ask a colleague to read the document without the original conversation. Can they identify the question, inspect the supporting material and understand the next decision? If a field needs a verbal explanation, add that context to the record. Check that the proposed creative test follows from the evidence and names the uncertainty.
Use the structure. Verify the substance.
This guidance helps organize creative research. It does not establish platform coverage, explain causation or promise a repeatable result. ReelsData is in development and all sample metrics on this site are fictional.