Classic Models
Customers, sales reps, offices, orders, order lines and payments for a wholesaler of scale model cars, 2003 to 2005. Clean keys make it a good first model.
Start here
Lessons and reference
Look things up
Start here
SQL
Power Query
Data modeling
Power BI and DAX
Finance playbooks
Run queries
Build and check
Recognizable business data, all synthetic and read-only. Pick a topic that matches your work, then open one to see its tables, sample questions, and the practice files built from it. Most databases exist on both SQL Server and MariaDB.
Pick the goal closest to yours. Each link opens the database that fits it best.
Clean keys, familiar business data and few tables. Start here if SELECT, JOIN and GROUP BY are new.
Customers, sales reps, offices, orders, order lines and payments for a wholesaler of scale model cars, 2003 to 2005. Clean keys make it a good first model.
A small trading company that buys from suppliers and sells to customers: orders, order details, products, shippers, employees and purchase orders.
A small demo company database, available on MariaDB. It is the default database for the legacy API examples.
Where revenue comes from: orders, order lines, products and customers. Good for margin, mix and top-N questions.
A compact order-entry model with customers, orders, order detail, products, suppliers and employees.
The classic Microsoft sample: people, employees, products, purchasing and sales for a bicycle manufacturer, in a normalized transactional design.
Films, actors, categories, customers, rentals and payments for a DVD rental chain. Great for many-to-many relationships and date arithmetic.
The ledgers behind the books: invoices, bills, payments, journal entries and receivables aging. Start here if you close, reconcile or report.
Accounting data shaped like QuickBooks Online: invoices and their line items, bills, payments, deposits, journal entries and items.
Accounting data shaped like QuickBooks Desktop: customers, vendors, invoices, bills, payments, accounts and journal entries.
A real accounting system layout with hundreds of tables for accounts receivable, accounts payable, general ledger, inventory and orders. Explore with care: use the schema explorer.
A CPA firm's billing, realization and WIP, and a nonprofit's donors, grants and fund accounting.
A small-business CPA firm: clients, engagements, tax returns, time entries, invoices, payments, retainers, a balanced general ledger and document tracking. About 1,000 clients, six years of history, a client map and a July-June fiscal calendar.
A regional nonprofit: donors, gifts, pledges, campaigns and events, grants and proposals, fund accounting with a general ledger, payroll by fund, volunteers, program sites and a mobile food market. Donor map and a July-June fiscal calendar.
Municipal and utility operations: 311 requests, dispatch, parcels and budgets, or work orders, assets and purchasing.
Big Bear Lake is a small mountain resort city in California, and this is municipal-style operating data for it, the way a city manager, finance director or public works superintendent would see it: 311 requests, snow plow operations, fire and sheriff dispatch, building permits and code enforcement, a city-run water utility, the lodging tax on short-term rentals, parcels and owners, fund accounting with budgets, and council agendas and votes. The records are invented; the place and the kind of work are real. July-June fiscal year, with coordinates on almost everything.
A water and electric utility's field operations in six regions: assets and their condition, 25,000 work orders with labor and materials, vendors with purchase orders and invoices, customer accounts with a year of monthly bills, and service interruptions. Lookup tables are unique and every relationship is a declared key.
Stars, snowflakes and big fact tables: the shapes you build a semantic model on in Power BI or Excel.
A retail sales star schema with stores, channels, products, promotions, 30,000 customers, budgets, 18 currencies, monthly inventory and three years of sales. The best database for DAX and time intelligence.
A retail data warehouse built to teach modeling: a star schema with snowflaked product and store dimensions, customers that change over time (type 2), a price history, a month-end inventory snapshot, an order pipeline with a date for each stage, a factless promotion table, four currencies with daily exchange rates, and budgets at a different grain.
A manufacturer's order history: 25 million order lines to wholesalers, retail chains and retailers, plus a three-level bill of materials and a customer hierarchy. On SQL Server it is a compressed columnstore table; on MariaDB the same rows are computed on the fly, so nothing large sits on disk.
Deliberately untidy tables, one technique each: promoted headers, unpivots, JSON, fuzzy merges and geocoding.
A Power Query playground: a joinable retail backbone plus deliberately messy tables, one per technique, from promoting headers and unpivoting to JSON, fuzzy merges, geocoding and custom functions.
Time-series readings from a connected home, and countries and cities with coordinates for maps and pure SQL.
One connected family home: 83 devices in 12 rooms, 13 months of sensor readings, solar, battery and EV energy, automations, alerts, security cameras, weather, utility bills, sleep, trips and a floor plan. October-September fiscal year.
Countries, cities and spoken languages with population, area and life expectancy. No dates, so it is ideal for pure SQL and for hierarchies and maps.
No databases in this topic.
Follow a learning path, or try the Power BI connection guide to pull any of these into Power BI Desktop.