A better starting point
for your next idea.

Help me review a small content experiment without mistaking correlation for causation.
Goal: {{goal}}
Posts and available metrics: {{metrics}}
What changed between posts: {{changes}}
Audience or timing context: {{context}}
Summarize observations, plausible explanations and limits. Pick one variable to test next week and define a success measure tied to the goal. Include a simple experiment log with hypothesis, change, measurement and stop rule. Do not invent benchmark numbers, claim statistical significance from a tiny sample or optimize views when the goal is useful engagement. If data is missing, propose measurement first.

Create a concise research brief using only the supplied material.
Research question: {{question}}
Source excerpts with URLs and publication dates: {{sources}}
Decision context: {{decision context}}
Required freshness: {{freshness}}
Return the answer, evidence table, conflicting findings, confidence limits and next verification steps. Attach a source to each factual claim and distinguish reported facts from your inference. Do not invent citations or assume a source says more than the excerpt provides. If the material is too old or incomplete, mark the relevant question unanswered. End with a recommendation only if the evidence supports one.