Search courses, chapters, or pages...
Pick out the specific thing that happened in a messy business story, such as an order placed, a refund issued, or a ticket closed. You’ll practice separating the event from opinions, explanations, and later consequences.
Use what you learned in the previous lesson to solve real-world problems.
Decide what one record should capture when an event happens: who or what was involved, when it happened, and what basic facts were observed. You’ll see why mixing several events into one record makes reports hard to trust.
Check what you understood with a short quiz.
Recognize the business objects that events are about, such as customers, products, stores, employees, vendors, or accounts. You’ll separate an entity’s descriptive traits from the events that happen to it.
Sort fields into numbers you calculate with and labels you use for grouping. You’ll classify examples like revenue, quantity, region, product category, and customer segment as measures or dimensions.
State exactly what one record represents before anyone starts counting. You’ll compare grains like one order, one order line, one customer per month, and one store per day to see how each changes the meaning of a metric.
Decide whether a measure can be summed freely, summed only across some dimensions, or should not be summed at all. You’ll avoid common mistakes like adding percentages, averaging averages, or treating balances like sales.
Use dimensions to break a measure into useful comparison groups without changing what is being measured. You’ll reason through category choices like region, channel, product family, and customer type.
Choose the timestamp that matches the business meaning of a report, such as order date, ship date, payment date, cancellation date, or ticket close date. You’ll see how the same activity can answer different questions depending on which time field is used.
Match reporting periods to the way the business actually runs, including days, weeks, months, fiscal calendars, time zones, and cutoff times. You’ll spot why “sales this week” can change if the period boundary is unclear.
Distinguish records that describe something happening from records that describe a state at a point in time. You’ll compare transactions like purchases with snapshots like inventory on hand or account balance at day end.
Trace business processes that unfold across several milestones, such as opened, assigned, resolved, shipped, delivered, or canceled. You’ll use milestone timestamps to reason about cycle time, wait time, and status without double-counting the same case.
Convert messy labels, codes, units, and free-text entries into values a report can group consistently. You’ll practice spotting problems like “NY” versus “New York,” dollars versus cents, and product names that mean the same thing.
Treat blank, unknown, not applicable, and zero as different meanings instead of one vague problem. You’ll decide how each case should appear in a report so the data does not hide uncertainty or create false totals.
Recognize records that repeat the same event, reverse an earlier event, or update a value after the fact. You’ll reason through how duplicates, refunds, corrections, and late-arriving records can change a reported number.
Review this chapter with practice based on your mistakes.