Every lesson in the Business Analytics slide course, in full text: 5 decks, 381 slides.
The Data Analyst's Toolkit — First Session (Sports Business)The kickoff session for a student building a portfolio of end-to-end sports-business-operations analytics projects for internship interviews. It introduces the five-tool stack - SQL, Python, Power BI, Excel, and GitHub - giving the honest pros and cons of each, showing where each one fits in a project, and explaining how they hand off to one another to form a single pipeline that extracts, analyzes, presents, and publishes. Across 31 slides there are five tool sections, each with a strengths-and-limits breakdown and a tool-choice trap, two runnable examples (a ticket-revenue SQL query and an attendance EDA in pandas), an SVG pipeline diagram, the end-to-end recipe, a full worked sports project carried through all five tools, and two checks. No prior knowledge is assumed.
Monte Carlo Simulation — Four Corners CaseA master's-level walkthrough of the Four Corners financial-planning case in R, in 43 slides. It builds the deterministic compounding engine, treats Goal Seek as root-finding, sets up the two input distributions (uniform salary growth and normal portfolio growth), builds the Monte Carlo engine, and reads out the risk with P(success), expected shortfall, and VaR and CVaR, before extending to the 25-year horizon and a reusable template. Every decision is justified: why the half-year earnings convention is used, why the simulation mean lands below the deterministic $772,722, and why running more iterations does not raise the 1% success rate. The deck includes four traps, four checks, and five figures rendered from the actual model.
Decision Analysis — Oceanview Property PurchaseA master's-level walkthrough of the Oceanview Development property-purchase case (Camm et al., 4th ed., p. 780), in 34 slides. It builds the decision tree and folds it back to the no-research decision, which is to bid, at an expected value of $50,000. It then revises the referendum odds with Bayes' theorem from the survey's reliability and derives the conditional strategy: bid if approval is predicted, for an expected value of $229,268, and don't bid if rejection is predicted, where the expected value is -$74,576. Finally it values the $15,000 study, with an EVSI of $44,000 against an EVPI of $70,000, an efficiency of 62.9%. Five traps and four checks are each tied to a real decision-analysis misconception, the decision tree is drawn in SVG, and every number is reproduced from the R script and the answer key.
Power BI on Process Data: From a Historian Export to a Live ReportA live-build companion for turning water-testing and waste-stream data into a Power BI reporting tool. It reshapes a wide historian export into one row per reading, puts the source behind a parameter so the go-live is a single value change, flags the four kinds of bad process reading rather than deleting them, builds a star schema with a marked date table, writes measures for the mean, a rolling average and three-sigma control limits that draw flat, then reads the resulting control chart and dose-against-pH scatter to answer the chemical dosing question, closing on the gateway and the go-live checklist.
Jira Burndown in Tableau: Planned, Actual & Projected on One ChartA live-build companion deck for a Tableau tutoring session. It pivots three wide Jira date columns into a single long axis of Milestone Type and Milestone Date, then builds the FIXED-LOD running-percentage calculation that stops a self-normalizing Actual series from claiming 100% completion before it is true. It covers TODAY()-boundary and dashed-Projected styling, including the fallback for versions before 2023.2, and finishes with a data-hygiene audit of a real export - a column shift, three typo dates, and an orphan Parent Epic - that students must catch before trusting a single number.
Want this taught 1-on-1? Alexander tutors Business Analytics — $55/session, free consultation.