Growth Methods for Feature Launch | Dataford Interview Questions

Your question is Growth Methods for Feature Launch. Take a moment with it on the right.

Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).

Business Context

StreamHub, a short-video app, launched a new creator referral feature and wants to know which statistical methods are most useful for evaluating product growth. Rather than answer conceptually, use the data below to quantify growth using several common methods.

Problem Statement

Assess whether the referral feature improved user growth and identify which statistical methods are most informative for this decision.

Given Data

Weekly new activated users were tracked for 8 weeks before launch and 8 weeks after launch.

Period	Week Index	New Activated Users
Pre-launch	1	12050
Pre-launch	2	12180
Pre-launch	3	12240
Pre-launch	4	12310
Pre-launch	5	12420
Pre-launch	6	12510
Pre-launch	7	12640
Pre-launch	8	12720
Post-launch	9	13080
Post-launch	10	13240
Post-launch	11	13310
Post-launch	12	13490
Post-launch	13	13620
Post-launch	14	13710
Post-launch	15	13880
Post-launch	16	14020

Assume a simple linear trend model where launch effect is measured as a level shift after week 8.

Requirements

Compute the pre-launch and post-launch mean weekly activations.
Test whether the post-launch mean is significantly higher using a two-sample t-test at $\alpha = 0.05$ .
Construct a 95% confidence interval for the difference in means.
Estimate the launch effect using a simple regression with an indicator for post-launch period.
Briefly explain which statistical methods are most useful here and why.

Assumptions

Weekly observations are treated as approximately independent.
Variance is stable across the two periods.
No other major product or marketing changes occurred during these 16 weeks.
The goal is directional product-growth assessment, not full causal attribution.