TikTok Interview Questions — Page 1

TikTok·Data Scientist·Technical Phone Screen

Jan 2023

TikTok data scientist interview with a SQL cohort analysis question. Pretty standard product analytics setup but the specifics tripped me up a bit.

Given a table of creator posts with user IDs and post dates, write SQL to assign each user to a cohort based on their earliest post date. For every cohort and posting date combination, return the total number of posts and the count of distinct active users, ordered by cohort start date and posting date.

“My first instinct was to just group by post_date and call it a day, which would've completely missed the cohort part.”

View Post

TikTok·Data Scientist·Technical Phone Screen

Jun 2025

TikTok data science interview with a single meaty causal inference question about a notification feature rollout. No randomization was possible, so the whole thing was about justifying observational methods under real-world constraints. Brutally specific.

Your team launched a notification feature without randomization due to marketing constraints. Design a causal study to estimate its effect on 7-day retention, comparing propensity score matching and difference-in-differences across multiple dimensions: treatment and control definitions, covariates for conditional ignorability, common support, pre-trend checks, staggered rollout, clustering strategy, sensitivity analysis, and how you'd communicate results to leadership.

“This was a lot.”

View Post

TikTok·Data Scientist·Technical Phone Screen

Jun 2025

TikTok data scientist interview with a meaty metric investigation case. Single question but it covered basically every angle: data validation, segmentation, causal inference, and decision-making. Left feeling like I either nailed it or completely missed what they were actually looking for.

Your checkout conversion rate jumped from 3.2% to 4.5% on a specific date and held elevated for three days. Walk through how you'd investigate this end-to-end, covering data validation, segmentation to catch Simpson's paradox, hypothesis ranking with rough impact estimates, separating product-causal from mix-driven effects, deciding on next actions with guardrails, and what you'd produce in the first 60 minutes.

“This is a monster of a question.”

View Post

TikTok·Data Scientist·Technical Phone Screen

Jun 2025

TikTok data scientist interview with a deep metric investigation case. One question, but it was a beast. The kind of problem where you think you know where to start and then realize halfway through that you forgot three entire layers of the analysis.

The Ads Manager dashboard shows a 15% relative drop in ad-creation completion rate on a specific day compared to the prior 7-day baseline. Walk through how you would triage and investigate this: verify the drop is real, localize where it's coming from, determine root cause, quantify impact, and decide whether to roll back or continue.

“This question is basically five questions duct-taped together and they want all of them.”

View Post

TikTok·Data Scientist·Technical Phone Screen

Aug 2025

TikTok data scientist interview focused entirely on experiment design for a messaging feature, specifically a Quick Reply A/B test. Five questions back to back, all deeply technical, no behavioral fluff. Left feeling like I needed to have read three textbooks I hadn't opened in a year.

Define the primary north-star metric for a Quick Reply feature launch, including the exact formula, unit of analysis, numerator, denominator, inclusion criteria, and a 24-hour attribution window from tap to reply send.
Propose at least two guardrail metrics that protect long-term product health during the experiment, and specify the formula, measurement unit, and acceptable movement thresholds for each.
Overall reply send rate increases but average conversation length drops and complaint rate rises among ads-acquired users. How do you segment and interpret these conflicting signals to avoid Simpson's paradox and decide whether to ship, hold, or iterate?
Design a metric to detect empty engagement such as accidental taps or replies deleted before sending, and explain how you would implement it using existing telemetry and factor it into the launch decision.
Lay out a pre-analysis plan for this experiment: how would you pre-register metrics, define stopping rules, and prevent metric fishing? Also explain how you would validate that 'exposed' users actually saw the Quick Reply entry point rather than just being eligible for it.

“This is where I spent the most time and still feel shaky about my answer.”

View Post

TikTok·Data Scientist·Technical Phone Screen

Sep 2025

SQL window function question at TikTok for a Data Scientist role. Pretty focused technical screen, just the one problem but it had enough layers to keep you busy for a while.

Given an impressions table with ad_id, user_id, and a timestamp column, write a single SQL query that returns, for each (user_id, ad_id) pair with at least two impressions, the most recent timestamp, the second most recent timestamp, and the difference between them in seconds. Use ROW_NUMBER ordered by timestamp descending to break ties.

“The core mechanics clicked fast for me.”

View Post

TikTok·Data Scientist·Technical Phone Screen

Sep 2025

TikTok Data Scientist SQL round, all five questions were analytics-flavored with a strict no-window-functions constraint which honestly tripped me up more than I expected. The schema was straightforward but the edge cases were where they were really testing you.

Using only ANSI SQL with no window functions, compute daily active users for each day in a 7-day window, counting distinct users who had a session event on each day.
For users who signed up within a 7-day window, calculate conversion rate by channel, where a converter is any user with at least one purchase within 7 days of their own signup date. Do not double-count users with multiple purchases.
Without window functions, compute day-1 retention for each cohort day: among users with a session on day D, what fraction also had a session on day D+1? Return cohort date, retained users, cohort size, and retention rate.
For all users, return each user's first purchase date and their total lifetime revenue, using only aggregates and GROUP BY. No window functions.
Identify the single top channel by total purchase revenue for a given 7-day period, breaking ties by lexicographically smallest channel name. Return one row with channel and total revenue.

“Straightforward GROUP BY on event_date with a WHERE filter and COUNT(DISTINCT user_id).”

View Post

TikTok·Product Manager·Onsite - Product Sense / Strategy

Apr 2026

TikTok PM interview with a single product sense question about age verification. Pretty short on details but the question itself is a meaty one that took me a while to unpack properly.

How would you prevent underage users from lying about their age when signing up for the platform?

“My first instinct was to jump straight to technical solutions like ID verification or credit card checks, but that ignores like 90% of the actual problem space.”

View Post

TikTok·Product Manager·Onsite - Product Sense / Strategy

Apr 2026

PM interview at TikTok with a product design question focused on trust and safety in e-commerce. One question, pretty open-ended, left me wishing I'd gone deeper on the technical side of things.

How would you design a product to verify the authenticity of items being sold on TikTok Shop?

“I went straight to user trust and seller verification flows, which felt right, but I think I underweighted the supply chain angle.”

View Post

TikTok·Product Manager·Onsite - Product Sense / Strategy

Apr 2026

PM interview at TikTok with a product sense question about improving CSAT for Costco. Pretty standard case-style round but the cross-industry angle (TikTok asking about a membership retailer) threw me a bit.

How would you improve the customer satisfaction score for Costco, a membership-based retail store?

“I spent the first minute just trying to figure out why TikTok was asking about Costco.”

View Post