Trace the recommendation back to supporting data evidence
Since every group has the same data, we can all trace the same evidence chain together. Let's walk through where the final recommendation comes from.
Data Sheet → Analysis Sheet
Raw weekend sales become descriptive stats (average, median, spread). What this proves: What does a "normal" café weekend look like?
Analysis Sheet → Forecasting Sheet
Stats feed the regression model. What this proves: What will the next quarter likely look like, and what are its limits?
Forecasting Sheet → Dashboard Sheet
Trend and predictions become visuals. What this proves: The data story in a decision-ready format.
Dashboard Sheet → Recommendation Sheet
Visuals support the final claim. What this proves: The investor-ready recommendation with evidence and risk.
The evidence chain is complete when you can trace any number in the Recommendation sheet back to the raw Data sheet.
Investors and decision-makers trust analysis that has clear evidence. Here are the warning signs that would make a workbook feel weak or untrustworthy:
❌ Recommendation with no numbers
Claim without data support — "we should expand" with no sales figures
❌ Charts that don't update
Static ranges (A1:C10) instead of table references
❌ Forecast without limits
Presenting regression as certainty without stating confidence limits
❌ Missing risk statement
No acknowledgment of what could go wrong