Chapter 24 Quiz: AI for Marketing and Customer Experience

Chapter 24 Quiz: AI for Marketing and Customer Experience

Multiple Choice

Question 1. Which of the following best describes the key difference between data-driven marketing (Era 2) and AI-augmented marketing (Era 3)?

(a) AI-augmented marketing uses more data than data-driven marketing.
(b) Data-driven marketing relies on human-created rules, while AI-augmented marketing uses models that learn and adapt autonomously.
(c) AI-augmented marketing eliminates the need for human marketers.
(d) Data-driven marketing is less effective than AI-augmented marketing in all circumstances.

Question 2. At which level of the personalization maturity model does the system anticipate customer needs before the customer expresses them?

(a) Level 2: Rule-Based Personalization
(b) Level 3: Model-Based Personalization
(c) Level 4: Real-Time Adaptive Personalization
(d) Level 5: Predictive and Proactive Personalization

Question 3. A marketing team sends the same promotional email to all customers who have not made a purchase in 30 days. What level of personalization is this?

(a) Level 1: Segment-Based
(b) Level 2: Rule-Based
(c) Level 3: Model-Based
(d) Level 4: Real-Time Adaptive

Question 4. Which layer of the conversational AI architecture determines when a chatbot should transfer a conversation to a human agent?

(a) Natural Language Understanding (NLU)
(b) Dialog Management
(c) Knowledge Retrieval
(d) Escalation Logic

Question 5. A customer sees a display ad on Monday, clicks a search ad on Wednesday, receives an email on Friday, and purchases on Saturday. Under last-touch attribution, which channel receives 100 percent of the credit?

(a) Display ad
(b) Search ad
(c) Email
(d) All three receive equal credit

Question 6. Data-driven attribution uses algorithmic methods to calculate each touchpoint's contribution. Which mathematical concept, also used in model explainability (Chapter 26), underpins Google's data-driven attribution model?

(a) Bayes' theorem
(b) Shapley values
(c) Principal component analysis
(d) Gradient descent

Question 7. Which of the following is NOT a component of the real-time bidding (RTB) ecosystem described in the chapter?

(a) Ad exchanges
(b) Demand-side platforms (DSPs)
(c) Customer data platforms (CDPs)
(d) Publisher ad servers

Question 8. According to the chapter, what is the most significant structural shift facing digital advertising?

(a) The rise of generative AI for ad creative
(b) The decline of television advertising
(c) The restriction and elimination of third-party cookies
(d) The increasing cost of programmatic ad inventory

Question 9. NK designed a three-tier opt-in personalization system for AthenaPlus. What percentage of pilot members opted into Tier 3 (Proactive), the highest personalization level?

(a) 12 percent
(b) 23 percent
(c) 41 percent
(d) 67 percent

Question 10. Which of the following best describes the "creepy line" in marketing personalization?

(a) The legal boundary established by GDPR for data collection
(b) The boundary between personalization that customers welcome and personalization that makes them uncomfortable
(c) The technical limit of what AI can infer from behavioral data
(d) The point at which personalization becomes unprofitable

Question 11. According to the chapter, which factor most strongly determines whether personalization feels helpful rather than invasive?

(a) The sophistication of the algorithm
(b) The amount of data collected
(c) Whether the customer has control and perceives a clear value exchange
(d) The price of the product being recommended

Question 12. What is the gold standard for measuring the causal impact of a marketing AI intervention?

(a) Before/after comparison
(b) Correlation analysis
(c) Incrementality testing with holdout groups
(d) Customer surveys

Question 13. NK's AthenaPlus personalization engine integrates outputs from models built in previous chapters. Which of the following combinations correctly represents the three scoring models she used?

(a) Demand forecasting (Ch. 8), recommendation engine (Ch. 10), sentiment analysis (Ch. 14)
(b) Customer segmentation (Ch. 9), churn risk scoring (Ch. 7), product affinity scoring (Ch. 10)
(c) Regression analysis (Ch. 8), topic modeling (Ch. 14), prompt engineering (Ch. 19)
(d) Neural networks (Ch. 13), time series forecasting (Ch. 16), computer vision (Ch. 15)

Question 14. Athena's approach to dynamic pricing, as described by Ravi Mehta, adjusts prices based on which factors?

(a) Individual customer willingness to pay and browsing duration
(b) Product, time, and inventory level — with the same price for all customers at any given moment
(c) Customer segment, loyalty tier, and purchase history
(d) Competitor pricing, customer income level, and geographic location

Question 15. In NK's AthenaPlus pilot results, which metric did NK describe as the one she was "most proud of"?

(a) Email open rate increase of 100 percent
(b) Revenue per loyalty member increase of 17 percent
(c) Opt-out rate decline of 57 percent
(d) Click-through rate increase of 171 percent

Question 16. The chapter describes the "Why was this recommended?" feature in NK's system. What percentage of Tier 3 members clicked this feature at least once during the pilot?

(a) 8 percent
(b) 15 percent
(c) 23 percent
(d) 41 percent

Question 17. Which of the following is an example of contextual targeting (as opposed to behavioral targeting)?

(a) Showing a running shoe ad to a user who recently searched for "marathon training plans"
(b) Showing a running shoe ad on a running blog, regardless of who is reading it
(c) Showing a running shoe ad to users who purchased running shoes from a competitor
(d) Showing a running shoe ad to users in a "fitness enthusiast" audience segment

Question 18. According to the chapter, what is the primary risk of AI-generated marketing content?

(a) It is always factually inaccurate.
(b) It is too expensive to produce at scale.
(c) It tends to be mediocre, brand-inconsistent, or factually unreliable without proper quality control.
(d) Customers can always detect AI-generated content and respond negatively.

Question 19. Ravi Mehta's final question to NK — "Where's the line between helpful personalization and manipulation?" — foreshadows which part of the textbook?

(a) Part 4: Prompt Engineering and AI Tools
(b) Part 5: AI Ethics, Bias, and Governance
(c) Part 6: AI Strategy and Organizational Transformation
(d) Part 7: The Future of AI in Business

Question 20. Which of the following statements about CLV models is correct?

(a) Traditional CLV models are more accurate than ML-based CLV models because they use fewer assumptions.
(b) ML-based CLV models can capture non-linear patterns and interaction effects that parametric models miss.
(c) CLV models should never be used for service level differentiation because doing so is inherently unfair.
(d) Deep learning CLV models analyze demographic data but cannot process temporal purchasing patterns.

Short Answer

Question 21. NK received two marketing emails on the same morning — one that felt helpful and one that felt invasive. In two to three sentences, explain the design difference between the two emails and what it reveals about the relationship between AI capability and customer experience design.

Question 22. Explain why the chapter describes marketing AI as a "synthesis" of nearly every technique from Parts 1-3 of the textbook. Identify at least four specific techniques from previous chapters and their marketing applications.

Question 23. The chapter states that "attribution modeling is not just a technical exercise — it is a political exercise." In three to four sentences, explain what this means and why data-driven attribution often encounters organizational resistance.

Question 24. Describe the difference between inference and observation in the context of the creepy line. Why do customers respond differently to these two types of personalization, even when both use the same underlying data?

Question 25. NK's loyalty program pilot showed that opt-out rates declined 57 percent even as personalization intensity increased. In two to three sentences, explain what mechanism NK attributes this to and why it matters for the broader debate about AI-powered marketing.

Answer key is provided in Appendix B: Answers to Selected Exercises.