Data Analysis & Decision Making Flashcards
7 cards from real GMC practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 7 Data Analysis & Decision Making flashcards as text
A growth team wants to measure the incremental impact of email campaigns beyond organic conversions. Which methodology is most appropriate?
Answer: Holdout group testing
Holdout group testing withholds the email from a control group, allowing the team to measure lift that is truly incremental to baseline organic behavior.
In regression analysis for growth modeling, what does a negative coefficient for a variable indicate?
Answer: As the variable increases, the outcome metric decreases
A negative coefficient means the variable has an inverse relationship with the outcome — as it rises, the predicted metric falls.
Which of the following best describes 'activation' in the AARRR (Pirate Metrics) framework?
Answer: A user experiencing their first meaningful value from the product
Activation is the moment a user first experiences the core value of the product, often called the 'aha moment.'
A growth analyst builds a predictive churn model. The model has 92% accuracy but the team finds it unhelpful. What is the most likely issue?
Answer: The dataset is imbalanced — most users don't churn — so accuracy is misleading
With highly imbalanced classes (e.g., 5% churn), a model predicting 'no churn' for everyone achieves ~95% accuracy while being useless for identifying churners.
What is the primary purpose of using UTM parameters in growth marketing campaigns?
Answer: To track campaign source, medium, and name in analytics platforms
UTM parameters append tracking data to URLs so analytics tools can attribute traffic and conversions to specific campaigns, sources, and mediums.
A SaaS company has an MRR of $500K and a churn rate of 5% per month. What is the approximate average customer lifetime in months?
Answer: 20 months
Average customer lifetime = 1 / churn rate = 1 / 0.05 = 20 months.
When reporting on growth experiment results to stakeholders, which approach is most credible?
Answer: Present the primary metric, guardrail metrics, confidence intervals, and sample size
Presenting primary metrics, guardrails, confidence intervals, and sample size gives stakeholders the full context needed to make informed decisions.