How do I approach Machine Learning interview questions?

Machine Learning questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master machine learning interviews.

What difficulty level is this interview question?

This is a medium difficulty Machine Learning question, commonly asked during Onsite rounds at Squarespace.

What role is this question designed for?

This question is commonly asked for Data Scientist candidates at Squarespace during technical interviews.

Predict Seller Intent From Subscription Data | Squarespace Interview Question

Q: Predict Seller Intent From Subscription Data

This question evaluates a data scientist's competency in defining target labels from temporal data, handling timestamp edge cases and missingness, engineering features from categorical, free-text and user-agent fields, and building and interpreting an interpretable binary classification model.

You are given a take-home dataset, seller_intent_take_home_dataset.csv, containing about 5,000 new subscription records from a website-building platform. The business goal is to understand which factors are associated with a new user becoming an active seller and to build an interpretable prediction model for seller intent.

The dataset has the following columns:

Column	Type	Description
`subscription_id`	string	Unique subscription identifier.
`subscription_start`	timestamp	When the subscription officially started.
`first_sale`	timestamp, nullable	When the user made their first sale, if any.
`days_in_trial`	integer	Number of days the user spent in trial.
`subscription_plan`	categorical	Subscription plan selected.
`subscription_period`	categorical	Billing period, such as monthly or annual.
`discount_amount`	numeric	Discount amount applied to the subscription.
`country`	categorical	User country.
`site_topic`	categorical/free text, nullable	User-provided site topic from onboarding.
`site_need`	categorical/free text, nullable	User-provided reason or need for creating the site.
`user_agent`	string	Raw browser user-agent string.

Important data issues:

There is no existing is_seller label. You must define the target variable using first_sale and subscription_start .
Some rows have inconsistent timestamps, including cases where first_sale occurs before subscription_start .
site_topic and site_need have substantial missingness, roughly 28% to 32%, and many long-tail categories.
user_agent is a raw string, so useful features such as device type, operating system, and browser must be extracted.
The final notebook or report must be concise enough to present clearly and should prioritize data reasoning, feature engineering, and interpretability over complex hyperparameter tuning.

Deliverables:

Define the target variable and explicitly state how you handle timestamp edge cases.
Perform exploratory data analysis focused on data quality, missingness, class balance, and feature distributions.
Engineer useful features from categorical fields, timestamps, discounts, trial length, and user_agent .
Train an interpretable model, such as logistic regression, and evaluate it using appropriate classification and ranking metrics.
Explain the main drivers of predicted seller intent, while distinguishing predictive associations from causal claims.
Prepare a short presentation explaining your assumptions, modeling choices, results, limitations, and recommended next steps.

The dataset has the following columns:

Column	Type	Description
`subscription_id`	string	Unique subscription identifier.
`subscription_start`	timestamp	When the subscription officially started.
`first_sale`	timestamp, nullable	When the user made their first sale, if any.
`days_in_trial`	integer	Number of days the user spent in trial.
`subscription_plan`	categorical	Subscription plan selected.
`subscription_period`	categorical	Billing period, such as monthly or annual.
`discount_amount`	numeric	Discount amount applied to the subscription.
`country`	categorical	User country.
`site_topic`	categorical/free text, nullable	User-provided site topic from onboarding.
`site_need`	categorical/free text, nullable	User-provided reason or need for creating the site.
`user_agent`	string	Raw browser user-agent string.

Important data issues:

There is no existing is_seller label. You must define the target variable using first_sale and subscription_start .
Some rows have inconsistent timestamps, including cases where first_sale occurs before subscription_start .
site_topic and site_need have substantial missingness, roughly 28% to 32%, and many long-tail categories.
user_agent is a raw string, so useful features such as device type, operating system, and browser must be extracted.
The final notebook or report must be concise enough to present clearly and should prioritize data reasoning, feature engineering, and interpretability over complex hyperparameter tuning.

Deliverables:

Define the target variable and explicitly state how you handle timestamp edge cases.
Perform exploratory data analysis focused on data quality, missingness, class balance, and feature distributions.
Engineer useful features from categorical fields, timestamps, discounts, trial length, and user_agent .
Train an interpretable model, such as logistic regression, and evaluate it using appropriate classification and ranking metrics.
Explain the main drivers of predicted seller intent, while distinguishing predictive associations from causal claims.
Prepare a short presentation explaining your assumptions, modeling choices, results, limitations, and recommended next steps.

Predict Seller Intent From Subscription Data

Quick Overview

Solution

Submit Your Answer

Predict Seller Intent From Subscription Data

Quick Overview

Solution

Submit Your Answer