TikTok Data Scientist Interview Process

The interview process at TikTok usually kicks off with the application. You either send an online application or receive outreach from recruiters, often through platforms like LinkedIn. This is where TikTok learns about you - your background, skills, and interest in the job. You might see updates on the application site (such as "written test" or "resume screening,”) telling you if you need to do a test or if they're reviewing your resume.

If your application looks good, you'll have a chat with someone from HR. They'll ask about your experience and why you're interested in the job. This helps TikTok decide if you're a good fit for the job and if you're genuinely interested in working there. So, make sure you have a clear reason for why you’re applying to TikTok, as well as why you think you’d succeed there as a data scientist. Also, feel free to ask any questions you have.

At the start of the process, candidates tackle technical tests to gauge their foundational skills. There are two technical rounds in general for data scientists:

Tech Screen I

In the first round of technical screening, you will face questions on SQL (Structured Query Language) and LeetCode-style/HackerRank problems or a live technical interview—all of which are essentially coding challenges of medium difficulty. Here, the focus is on fundamental concepts like data structures, algorithms, basic statistics, and ML. You will face at least one SQL question requiring the use of window functions; you'll typically solve them using SQLPad. 

• Make sure you have a solid grasp of the built-in functions available in SQL and know when and how to use them effectively.
• Ask for feedback or suggestions and be open to discussing different approaches to solving the problem.

Example Questions

• Query the dates when the following customers became satisfied customers (customer_id in 1, 2, 3, 4, 5, 6, 7, 8, 9, 10). Skip customers who never became "happy customers."
• Write a query to retrieve the shortest movie from each category. The order of the results is not important. In case of ties, return only one. Include columns: film_id, title, length, category, and row_num.
• Create a query to determine the percentage of revenue for each film with an ID less than or equal to ten within its category. Calculate the percentage using the formula: (Revenue of film_id x) * 100.0 / Revenue of all movies in the same category. Return columns: film_id, category name, and percentage.
• Write a query to find the average customer spend per month. The average customer spend is calculated as the total customer spend divided by the unique number of customers for that month. Utilise EXTRACT(YEAR from ts_field) and EXTRACT(MONTH from ts_field) to extract year and month from a timestamp column. The order of the results is not significant.
• Query to identify customers who placed more than one order on the same day. Include columns: customer_id, order_date, and the count of orders placed on that day.
• Write a query to fetch films released in the last three months. Include columns: film_id, title, release_date, and category.
• Create a query to calculate the total revenue generated by each category. Return columns: category name and total revenue.
• Query to identify customers whose total spend is above the average spend of all customers. Include columns: customer_id, total_spend, and average_spend_across_all_customers.

Tech Screen II

Moving on to the second screen, they dive deeper into your skills and experience. This typically means a review of your resume and a follow-up technical discussion with an engineer or data scientist from the TikTok team. They're looking for examples of any hands-on experience you have or specific projects or challenges you've tackled that align with the role's requirements—make sure you've got a couple up your sleeve. You can also expect some Maths, A/B testing questions. 

If you need to practise the technical screens, there are several TikTok Data Scientists on Prepfully who can provide great advice and guidance on how to tackle these interviews. Book a session with them directly here.

Example Questions

1. Use SQL window function to get rank.
2. What is bivariate? Most of the questions were resume-based.
3. What is the difference between Random Forest and XGBoost?
4. Tell me two social media apps that you use daily, and tell me their differences.

There are typically three rounds in the onsite interview process for TikTok Data Science roles:

• Product Sense
• Stats & A/B Testing
• (Optional) Specialised Rounds 
• Take-Home Product Analytics Test
• Behavioral/Culture Fit & Department Manager Round

Product Sense

In the product sense interview, they basically want to see if you can build out a strategy from scratch, or critique and improve an existing product. You might get questions like: What would you do if you were the PM or CEO of X company?

They are typically straightforward product questions, but they'll be going into a lot of details, so, be prepared for that. They want to get insight into your creativity and instincts regarding products.

Example Questions

• Tell me about a product you think is awesome and why. 
• Average spending per advertiser dropped from 30% last month. How would you investigate? What data will you use?
• How to determine the frequency of emails we send to our users per week?

Stats & A/B Testing

In the Stats & A/B Testing round, you'll tackle statistical concepts and their use in A/B testing scenarios—they want to see if you can apply them to real-world situations, especially in the area of product development to make smart calls about product features.

Example Questions

• Give me an example where the p-value is smaller than 0.05, but you still decided not to launch the new feature.
• Type I error vs Type II error, what is the p-value, and what is a confidence interval?
• Let's say you're the data scientist for your company's marketing/advertising division. The marketing executive wants to test multiple new channels, including:

Youtube ads

Google search ads

Facebook ads

Direct mail campaigns

Given these new marketing channels, how would you design an a/b test to utilise the marketing budget in the most efficient way possible?

Behavioural and Fit Assessment

This round aims to delve into your past experiences, achievements, and challenges—anything that will give insights into your behaviour, thought process, decision-making mindset, and problem-solving skills.

Questions like "Tell me about a time when you had to work under pressure," or "Describe a situation where you had to resolve a conflict within a team” are bound to pop up. Another good question is: "Describe an experience that you consider most valuable.”

As an example answer, think of a specific situation from your past work or studies that taught you something important. It could be overcoming a big challenge, leading a successful project, or learning from a mistake. Explain why it was valuable and what you learned from it. 

Finally, you will speak with your prospective department manager who will be looking to further assess your fit within the team dynamics. This is typically a concluding interview where they inquire about your long-term goals, your leadership style, collaboration approach, etc. The manager will also be looking to assess your knowledge of TikTok’s mission and how you can contribute to it.

Example Questions

• Why TikTok?
• Why do you want to leave your current job?
• What is your favourite app? How do you like it compared to TikTok?
• What attracts you to the DS role?
• Talk about your most valuable/challenging/rewarding experience.

Specialised Rounds — Take-Home Product Analytics Test

In the later stages, DS candidates at Tiktok typically get a homework assignment called a product analytics test—where you have to come up with well-thought-out solutions to a real-world business scenario. You'll receive a short problem statement to work on, and asked to present your case in less than 150 words.

For example, the problem statement could go something like:

“As a data scientist, evaluate the effectiveness of the 'Recommendation Engine' feature on the TikTok platform. Identify one primary metric to measure its success and suggest additional metrics for a comprehensive evaluation."

In this scenario, you'd need to analyse the impact of the recommendation engine on user engagement or retention, for instance. The primary metric could be something like "click-through rate" or "time spent on platform." Additional metrics might include "conversion rate," "user satisfaction scores," or "number of repeat visits." Your task would be to propose these metrics and outline how they collectively provide a holistic view of the feature's performance.

Example Case Prompts:

• Propose experimentation and rollout plan, keep in mind the speed of innovation, statistical rigour, and potential risk mitigation in case of unforeseen issues.
• Let's say that you work at TikTok. The goal for the company next quarter is to increase the daily active users metric (DAU).
• Executive A believes the best way to increase DAU is to improve the recommendation algorithm in TikTok's "For You" feature (TikTok news feed).
• Executive B believes the best way to increase DAU is to acquire more new users.
• Executive C believes the best way to increase DAU is to improve the creator tools.
• The engineering team must prioritise one feature at a time.
• How do you figure out which executive is right?
• What data points and metrics would help validate your choice?
• Considering a 2-dimension matrix that can only be traversed by one adjacent position at a time and never diagonally. Create an algorithm to traverse that matrix from its upper-left corner to its lower-right corner using the shortest possible path most efficiently.
• Two people are each stuck on their island, connected by a ferryman with a lockable box. Each person has their lock and key but can't send the key along with the box. One person wants to send the other a diamond, but it must be locked into the box, or the ferryman will steal it. How do you ship the diamond without the ferryman stealing it?