Interactive Textbook Module Syllabus
Defining the Outcome
Every technique for validating a model assumes the outcome is already correctly defined — which makes the definition the one decision no diagnostic can check. Learn why a label is a choice rather than a fact, what changes between contractual and non-contractual settings, how censoring and involuntary events distort a target, why a successful intervention corrupts its own label, and why the quantity worth predicting is often the effect of acting rather than the outcome itself.
Full Interactive Chapter (Account Required)
The complete chapter reading, embedded live mini-widgets, inline Check-Your-Understanding questions, and progress tracking are available to enrolled students and verified instructors.
Chapter Section Outline (14 Sections)
Topics and managerial frameworks covered in Defining the Outcome:
-
1.
The Question Nobody Asked
-
2.
A Label Is a Decision, Not a Fact
-
3.
Contractual and Non-Contractual Worlds
-
4.
Choosing a Horizon
-
5.
The Prevalence You Create
-
6.
Censoring: Customers Whose Story Isn't Over
-
7.
The Grey Zone
-
8.
When the Model Changes Its Own Outcome
-
9.
Predicting the Wrong Thing Entirely
-
10.
Proxy Outcomes
-
11.
Write It Down
-
12.
Case Application — Three Definitions, Three Models
-
13.
Reference Card
-
14.
What This Sets Up
Students manipulate these interactive visual simulations directly inside the chapter reading before answering gated check-in questions:
How Long Is Your Horizon?
Sliding the horizon moves prevalence (statistically convenient) and warning time (operationally useful) in the SAME direction, while the amount of usable training data moves the opposite way. Students discover there is no setting that wins on all three, and...
Whose Outcome Do You Actually Know?
Draws twelve customers on a calendar that ends where the data stops. Each customer's clock starts at their own sign-up, and the label asks "churned within H days of signing up?
Who Should You Actually Contact?
The chapter's hardest idea (Ascarza's retention futility result) made arithmetic: contact customers ranked by churn RISK and you sweep up lost causes; contact by treatment EFFECT and you reach persuadables. The widget computes net value under both strategie...
This chapter connects directly to these full MktPraxis analytical apps (with pre-loaded business case scenarios, CSV upload, and reproducible Python export):
Predict binary outcomes like Churn vs. Retain.
Pre-loaded Business Scenarios: Influencer Posts: What Makes a Story Drive a Sale? · B2B Email Sequences: Predicting Demo Bookings · Digital Promo A/B Test: High Incentive vs Standard Offer · Subscription Churn: Predicting Customer Attrition · Event Marketing: Predicting Registration Conversion · RFM-Based Purchase Prediction: Who Buys Next? · SignalPath MarTech: Which Leads Actually Close? · Will They Come Back? Predicting Season-Membership Renewal
Compare binary classification algorithms side-by-side on training and holdout data. Select from Logistic Regression, Decision Tree, k-NN, Naive Bayes, and Linear SVM.
Pre-loaded Business Scenarios: CloudMetrics SaaS — Customer Churn Prediction · NexaFlow — B2B Lead Conversion Scoring · PixelReach Ads — Click Prediction for Programmatic Bidding
Model customer retention and conditional spend to forecast CLV by segment. Identify high-value at-risk customers, evaluate marketing levers, and benchmark acquisition spend via CLV/CAC analysis.
Interactively explore how the classification threshold affects profit vs. cost trade-offs in binary ML models. Find the optimal T* that maximises net profit.
Pre-loaded Business Scenarios: FreshBox — DTC Snack Subscription · StreamFlow — Streaming Service · Kiteline Analytics — Enterprise-Tier B2B Analytics SaaS
See how long customers stay, which sign-up cohorts hold on best, and how honest your churn rate really is. Cohort tables and Kaplan-Meier curves that count the customers who have not left yet.