In development · Interactive examples use fictional data
Illustration of the planned research workflow
Reference accountsSource-linked observationsA brief with boundaries
A clearer starting point for the next team decision.
Begin with the question
Frame a report around the decision the team needs to make: which opening to test, which format to explore or where the current content mix is repetitive. Keep the reference set and observation dates visible.
Show what supports the conclusion
Connect every interpretation to source examples and the available observations. Include counterexamples and coverage limitations where they affect the conclusion. Avoid comparing unlike platform metrics as though they measure exactly the same behavior.
End with a bounded next step
Propose an experiment with a clear owner and a review date. The planned report should help the creative team write the next brief. Report formats, sharing permissions and exports are proposed capabilities, not a live service today.
Keep the limits visible
Know which source, observation window and comparison support the finding. The examples on this site are fictional; live coverage and permitted use must be established before a pilot.
A useful starting point
Begin with the question
Frame a report around the decision the team needs to make: which opening to test, which format to explore or where the current content mix is repetitive. Keep the reference set and observation dates visible.
CREATIVE RESEARCH / SAMPLE BRIEFOne finding. The evidence behind it.
WEEKLY DIRECTION / FICTIONAL EXAMPLE
Start with the first action.
A process-led opening gives the team one specific creative detail to investigate.
ProcessDemonstrationStory
Prepared example / not live analysis or customer reporting
Keep the context
Show what supports the conclusion
Connect every interpretation to source examples and the available observations. Include counterexamples and coverage limitations where they affect the conclusion. Avoid comparing unlike platform metrics as though they measure exactly the same behavior.
Propose an experiment with a clear owner and a review date. The planned report should help the creative team write the next brief. Report formats, sharing permissions and exports are proposed capabilities, not a live service today.
Starting pointThree example observations / process-led opening hypothesis
Keep the source and scope attached to the work.
✓
Useful handoffOne finding, supporting examples and a proposed creative experiment
The next owner needs enough context to understand the deliverable.
✓
Review decisionKeep the observation separate from the interpretation in the client handoff.
A named owner should be able to accept, revise or reject the direction.
The details matter
Keep the interpretation close to the evidence.
Give each finding a job
A weekly report should answer a client question, show the few examples that support the answer and identify a useful next discussion. A long list of posts is not automatically a finding. Group the evidence around a message the team can evaluate, while retaining links, observation dates and the comparison used to interpret it.
End with a proposed experiment
Translate the research into a bounded creative test: for example, compare a process-led opening with a product-led opening while keeping the audience and message stable. State what remains uncertain. The client should be able to accept, revise or reject the proposal without confusing the fictional demonstration on this site with a real performance report.
A worked planning record
A finding someone else can follow.
A question, the supporting context and a decision that remains open to review.
Creative researchIllustrative data
REFERENCE SET / SAMPLE WEEK
The weekly creative direction.
4.7×highest relative observation
Fictional accounts and metrics Observation ≠ explanation
A few things to make clear.
Is this a live analytics service?+
Not yet. ReelsData is in development. All charts and account records on this website are fictional examples. Data coverage and refresh schedules will be established before a pilot is offered.
Which platforms will it cover?+
Coverage has not been committed. Each platform needs a viable data source, clear field definitions and a tested refresh policy. A pilot will name exactly what is covered.
Does an outlier predict a viral video?+
No. An outlier describes unusual observed performance against a stated baseline. It does not establish what caused the result or predict the outcome of a new post.
What should we agree before a pilot?+
Source access, permitted use, reference accounts, observation frequency, history and baseline methodology. All demonstrations here use fictional metrics.
