Week 1 — Questions, Data & Spreadsheets · ~120 min
Weekly project: analysis brief & data dictionary
ProjectFree previewThe brief
You are the new analyst at a small retail business. The owner has handed you two files — the practice customers and orders data — and said, "Tell me what this can tell me." Before you analyse anything, produce an analysis brief that makes the work answerable and honest.
Deliverables (one workbook)
- source sheets —
orders_sourceandcustomers_source, untouched. - dictionary — every column, with type, unit, allowed values, and notes (from the lab).
- brief sheet, containing:
- Stakeholder and decision — who you're working for and what they will decide.
- Three questions, each written with population, period, grain and metric.
- For each question: answerable? (Yes/No) and why.
- The chosen question — the one you will answer in Weeks 2–3.
- Two limitations of the data.
- Checks — a small table with formulas proving: 8 order rows, 5 customers, total revenue 160, date range 2026-01-05 to 2026-03-21.
- work sheet — the
Orderstable with revenue and month columns.
Example of a strong question
The owner wants to decide which region to visit first for a customer-feedback trip. Across all 8 order lines (Jan–Mar 2026, one row per order line), which region produced the most revenue? Answerable: Yes — region and revenue are both available.
Example of an honest "No"
Which category is the most profitable? Answerable: No — the data has prices but no costs. We'd need unit cost per category.
Acceptance checklist
- Source sheets are untouched (reviewer compares with the CSVs)
- All 10 columns appear in the dictionary; identifiers are marked "do not sum"
- Three questions, each naming stakeholder, decision, population, period, grain, metric
- At least one question is honestly not answerable, with what's missing
- Check formulas return 8, 5, 160 and the correct date range — as formulas, not typed numbers
- Revenue column uses one formula filled down, not hand-typed values
- AI-use note included in the submission
Submit a share link to the workbook. In the notes, include your AI-use note ("No AI used" is fine).
Sign in to submit your work.