Two fact tables, five conformed dimensions and eight written measure definitions, one of them semi-additive, form this AC444 Unit 4 data model, built before any view. Searches like "ac 444 unit 4 assignment example", "ac444 unit 4 sample" and "ac444 unit 4 example" land here.
What a finished AC444 Unit 4 data model exercise looks like
A diagram page and four pages of specification. At the center of the diagram sits FACT_SERVICE_VISIT, one row per completed visit, about 84,600 rows for 2025, carrying billed amount, labor minutes, labor cost and materials cost. Beside it stands FACT_ACCOUNT_MONTH, a periodic snapshot holding one row per account per month-end with a status flag. Both connect to DIM_DATE, DIM_BRANCH and DIM_SERVICE_LINE; the visit fact also joins DIM_CUSTOMER and DIM_TECHNICIAN. Each table specification lists keys, attributes and the grain stated in one sentence. The measures page defines eight calculations in plain language and formula, revenue, direct labor, materials, gross margin, margin percent, route hours, active accounts and churn rate, and marks active accounts as semi-additive: summed across branches, never across months.
How a AC444 Unit 4 example is structured
Grain is declared before anything else, following Kimball's method, because every later choice depends on what one row means. The visit grain is chosen over the invoice grain since commercial accounts are billed monthly for several visits, and margin per route hour needs visit-level labor. The snapshot fact exists because a count of active accounts cannot be derived reliably from visits: an account with no visit in a month is still active. Dimensions are conformed across both facts so views can place revenue beside account counts without mismatched branch or date definitions. Measures are written out rather than left to the tool's defaults, and each carries its additivity. Deliberate exclusions get a short section of their own, sales tax and deferred revenue among them, which notes that the reconciliation unit will account for the latter when dashboard figures meet the ledger.
Grain stated first
One completed visit per row in the main fact table, argued against invoice-level grain because monthly commercial billing hides visit labor.
A snapshot for active accounts
Month-end rows per account exist because an account without a visit that month is still a customer the model must count.
Five dimensions both facts share
Date, branch and service line connect to both fact tables with identical keys, so revenue and account counts line up in any view.
Eight measures in words and formulas
Revenue through churn rate are defined in plain language beside their calculations, with additivity marked for each measure.
Summed across branches, never months
Active accounts add up across locations at one month-end but not across months, and the definition says so explicitly.
Left out on purpose
Sales tax and deferred termite renewal revenue stay outside the model, with a note on where each will be reconciled.
Where marks go in AC444 Unit 4
An undeclared grain undermines a model before anything else, since every measure's meaning depends on what one row represents and a grader cannot check a sum without knowing. Mixing grains in one fact table, visits alongside monthly account counts, is a frequent structural error and tends to produce inflated totals later. Semi-additive measures handled as ordinary sums draw deductions once a view adds twelve month-end counts into a meaningless annual figure. Dimensions defined differently in two facts, such as branch codes that disagree, break every comparison built on them. Measures left undefined, trusting the tool's default aggregation, read as unfinished. Rubrics in this unit also reward a stated reason for each table, so a diagram without narrative, however correct, often earns less than its structure deserves.
Get a AC444 Unit 4 example written to your instructions
Share the source tables or extract your Unit 4 exercise supplies, the rubric and whichever BI product the course requires, since measure syntax varies between products. The schema, grain statements and measure definitions come back within 24-48h, all drawn from your own columns, and a first custom sample costs nothing.
AC444 Unit 4 questions, answered
What does semi-additive mean in practice?
A measure that can be summed along some dimensions but not others. Active accounts can be added across the six branches at one month-end to get a company total, but adding January through December produces a number that describes nothing. The sample defines the measure to take the last month-end value in any period, and the definition states that rule in words.
Why two fact tables instead of one?
Because the two sets of facts have different grains. Visits happen at a moment; account status is a condition at month-end. Forcing both into one table either repeats account rows for every visit or loses accounts with no visits that month. Separate facts sharing conformed dimensions keep both correct and still let a view show them side by side.
Does the exercise require a specific tool?
Sections differ. Some want the model built in a BI application with relationships and measures defined there, while others accept a diagram and written specification. The sample provides both, and measure formulas can be written in the expression language of whichever tool your section uses once you name it.