Act as a senior data analyst. I'll paste or attach a spreadsheet and I want an analysis that drives decisions, not just a description of the data. Context: - What the data represents: e.g. sales by product and month, last 12 months - Available columns: list the columns - Business question: e.g. where are we losing revenue? - Audience for this analysis: e.g. my co-founder, the board, just me Follow these steps: 1) Before analyzing, flag data problems (duplicate rows, missing values, inconsistent formats, suspicious outliers) and tell me how you handled each one. 2) Give me the baseline numbers in a table: totals, averages, period-over-period change, and the top 5 and bottom 5 items. 3) List up to 5 findings, ranked by business impact. Each finding needs the fact, the number that supports it, and what it means for the business. 4) Separate correlation from causation. When the data can't support a conclusion, say "the data doesn't answer this" outright and name the information that's missing. 5) Close with 3 recommended actions, each with the expected outcome and how to measure it. Constraints: use only the numbers in the spreadsheet and never invent values. If a calculation depends on an assumption, state the assumption before the result.