CTO Data Strategy & Analytics 5 — Questions and Answers
Question 1: A CTO must present the ROI of a proposed data platform investment to the board. Which framework is most appropriate?
- Total Cost of Ownership (TCO) analysis alone
- Value driver mapping linking data capabilities to revenue, cost, risk, and agility outcomes (Correct answer)
- Number of petabytes that will be stored on the new platform
- Benchmark comparison of query speeds versus competitors
Correct answer: Value driver mapping linking data capabilities to revenue, cost, risk, and agility outcomes
Boards respond to business value framing; mapping data capabilities to specific drivers like revenue growth, cost reduction, and risk mitigation demonstrates strategic ROI.
Question 2: In a multi-cloud data strategy, what is the primary risk a CTO must plan for?
- Vendor lock-in to a single cloud provider's proprietary data formats
- Excessive redundancy causing data inconsistency across clouds (Correct answer)
- Too many open-source tools reducing vendor support options
- Higher latency for all analytics queries regardless of workload
Correct answer: Excessive redundancy causing data inconsistency across clouds
Multi-cloud deployments risk data fragmentation and synchronization inconsistencies when the same datasets exist across providers without a coherent replication strategy.
Question 3: Which data anonymization technique replaces a sensitive identifier with a consistent token, preserving referential integrity across datasets?
- Data masking with random values
- Tokenization (Correct answer)
- Data aggregation
- Differential privacy
Correct answer: Tokenization
Tokenization substitutes a sensitive value with a non-sensitive token that remains consistent, allowing joins across datasets without exposing the original identifier.
Question 4: A CTO is establishing a 'golden record' strategy for customer data. What problem does this solve?
- Slow query performance on large customer tables
- Duplicate and conflicting customer records across multiple source systems (Correct answer)
- Lack of encryption for customer personally identifiable information
- High storage costs from retaining inactive customer profiles
Correct answer: Duplicate and conflicting customer records across multiple source systems
A golden record (master data management) creates a single authoritative customer profile by deduplicating and reconciling data from multiple source systems.
Question 5: When building a real-time analytics pipeline, which architecture pattern provides the lowest latency for updating business dashboards?
- Nightly batch ETL with incremental loads
- Micro-batch processing every 15 minutes
- Event streaming with Apache Kafka and stream processing (e.g., Flink or Spark Streaming) (Correct answer)
- Daily full table scans with delta detection
Correct answer: Event streaming with Apache Kafka and stream processing (e.g., Flink or Spark Streaming)
Event streaming with Kafka and a stream processor delivers sub-second latency by processing events continuously as they are produced rather than in batches.
Question 6: A CTO is asked to ensure analytics results are explainable to regulators. Which type of ML model best satisfies this requirement?
- Deep neural network with 50+ layers for maximum accuracy
- Gradient boosting ensemble trained on all available features
- Logistic regression or decision tree with documented feature importance (Correct answer)
- Unsupervised clustering model with automatic feature selection
Correct answer: Logistic regression or decision tree with documented feature importance
Interpretable models like logistic regression and decision trees provide clear, documentable reasoning for each prediction, satisfying regulatory explainability requirements.
Question 7: Which statement best describes the difference between descriptive, predictive, and prescriptive analytics?
- Descriptive uses AI, predictive uses SQL, prescriptive uses spreadsheets
- Descriptive answers 'what happened', predictive answers 'what will happen', prescriptive answers 'what should we do' (Correct answer)
- Descriptive is real-time, predictive is historical, prescriptive is future-oriented
- They are synonymous terms used interchangeably in modern analytics platforms
Correct answer: Descriptive answers 'what happened', predictive answers 'what will happen', prescriptive answers 'what should we do'
The three analytics types form a maturity progression: understanding past events, forecasting future outcomes, and recommending optimal actions.
CTO Data Strategy & Analytics 5
A CTO must present the ROI of a proposed data platform investment to the board.
Which framework is most appropriate?