Product Sense via Machine Learning
Product and Metrics - Case Studies
-
How to improve a given metric
-
Marketplace - Focusing on supply or demand?
-
Actionable insights vs spurious relationships
-
Behavioral vs demographic variables
-
How to create key growth metrics
-
How to come up with new feature ideas?
-
How to estimate user LifeTime Value?
-
User Similarity
-
Design metrics for customer service
-
Product Development - Should we test this new feature?
-
Metrics - Understanding what drives them
-
Metrics - How to find bugs
-
How to predict user behavior?
-
Metrics - How to create an effective metric
-
Metrics - Self-selection bias
-
Metrics - Trade-off between short term vs long term effects
-
How to improve revenue from ads
-
Metrics - How to create an effective metric for gaming apps
Unbalanced Classes
Missing data
Fraud and Machine Learning - Case Studies
A/B Testing - Practice
A/B testing - Case Studies
New Product Case Studies - Miscellaneous
-
INSIGHTS: Should users be asked for their CC info when starting a free trial?
-
INSIGHTS: How much would you charge for a new product?
-
INSIGHTS: What was the data-driven hypothesis behind testing FB stories?
-
INSIGHTS: Trade-off between revenue and user experience metrics
-
INSIGHTS: What did you do when you got unexpected results?
-
INSIGHTS: How would you place ads on a page?
-
INSIGHTS: Would you test a feature allowing users to easily switch account?
-
INSIGHTS: Would you merge two similar products or run them separately?
-
INSIGHTS: Categorical variables in regressions. Better one-hot-encoding or multiple regressions?
-
INSIGHTS: How would you make Airbnb booking process smoother?
-
INSIGHTS: Can a bug actually improve your metric? What to do if that happens?
-
INSIGHTS: How to improve a marketplace?
-
A/B TESTING: Can you do a t-test on a dummy variable like conversion (0/1/)?
-
A/B TESTING: How would you test the success of a new ad campaign?
-
A/B TESTING: Define test statistical significance in layman’s terms. Why 0.05?
-
A/B TESTING: What would you do if test p-value is slightly above signficance level?
-
A/B TESTING: What to do if after a test some metrics are up and some down?
-
A/B TESTING: How to test different prices?
-
A/B TESTING: When would you design an A/B test as a sequence of multiple tests?
-
A/B TESTING: How can you halve the width of a t-test confidence interval?
-
A/B TESTING: How to check if a metric is actually up after a successful test?
-
A/B TESTING: How to avoid the A/B test issue of over-optimizing for the current user base?
-
A/B TESTING: When would you increase a test significance level?
-
METRICS: Conversion rate down, but absolute number of conversions up. Is it good or bad?
-
METRICS: How choosing an average-based metric vs a percentile one would impact product development?
-
METRICS: When to run a test on multiple metrics?
-
METRICS: When is accuracy a bad metric for a model?
-
METRICS: What kind of analysis is required after a logging-related bug?
-
METRICS: What kind of analysis is required after a product-related bug?
-
METRICS: In most metrics, you expect the average to be larger or smaller than the median?
Projects with Solutions
Collection of tech company blog posts/case studies
Data Challenges with Solutions
-
Introduction
-
INSIGHTS: Workplace Diversity Analysis
-
Solution: Workplace Diversity Analysis
-
INSIGHTS: Funnel Analysis
-
Solution: Funnel Analysis
-
INSIGHTS: Subscription Retention Rate
-
Solution: Subscription Retention Rate
-
INSIGHTS: Sessionize user activity
-
Solution: Sessionize user activity
-
INSIGHTS: Video Sharing Analysis
-
Solution: Video Sharing Analysis
-
METRICS: Ads Analysis
-
Solution: Ads Analysis
-
METRICS: Hotel Search Data
-
Solution: Hotel Search Data
-
A/B Testing: User Referral Program
-
Solution: User Referral Program
-
A/B TESTING: Pricing Test
-
Solution: Pricing Test
-
ML: Applying for a loan
-
Solution: Applying for a loan
-
ML: Optimization of Employee Shuttle Stops
-
Solution: Optimization of Employee Shuttle Stops
-
ML: Clustering Grocery Items
-
Solution: Clustering Grocery Items
-
ML: Credit Card Transactions
-
Solution: Credit Card Transactions
-
ML: Song Recommendation
-
Solution: Song Recommendation
Metrics - SQL Exercises
-
New Coding Course
-
Metrics: Time delta between consecutive events
-
SOLUTION: Time delta between consecutive events
-
Metrics: Segment by mobile, web, and cross-device users
-
SOLUTION: Segment by mobile, web, and cross-device users
-
Metrics: Identify power users based on their history
-
SOLUTION: Identify power users based on their history
-
Metrics: Estimate total and running values
-
SOLUTION: Estimate total and running values
-
Metrics: Summary stats - Custom implementation of the median vs average
-
SOLUTION: Summary stats - Custom implementation of the median vs average
-
Metrics : Rank users within groups
-
SOLUTION: Rank users within groups
