Skip to contentSkip to footer

Science Interview Questions

Salaries
Interviews

43,845 science interview questions shared by candidates

In order to push existing users to refer their friends, we're running a special reward program, in which the user is given an instant $10 discount for posting a referral message as their Facebook status. The message looks something like this: "Check out this company Jerry.ai --they automatically checks if you’re paying the lowest price for insurance and they will also find the best quote for you. As a friend of mine, you can get $20 off your insurance purchase. Click here to get the $20 gift credit: jerry.ai " The user is given the option to post this message on his/her Facebook account during the purchase. Once they post this message, they instantly get the $10 discount on their purchase. In other words, we don't wait for any of their referred friends to actually signup with us before giving them the discount. We feel that doing this would make the users more likely to post the referral message. Assume that this reward program has been running for a couple of months, and we have some data collected in our database. We want to know if running this program has been a good idea or not, i.e., are we acquiring new customers with it, or are we just losing money by giving out $10 discounts. Assume that you have the following database tables: 'User' and Purchase' Table User id, name, referring_user_id Table Purchase id, user_id, date, total, discounts In the Purchase table, the 'total' field contains the dollar amount of the job. The 'discounts' field consists of the total discounts given for the appointment (including rewards, coupon redemptions, etc.). Therefore, the customer pays: 'total' - 'discounts' as their final bill. Given this data: 1) 'What' information would you derive from it, and 'how' will you derive it (you can give SQL queries, pseudo code, ... whatever you're comfortable with) 2) Using the information from Step 1), how would you make a recommendation on whether this rewards program should be continued or discontinued
avatar

Data Scientist

Interviewed at Jerry

4.1★
Apr 27, 2021

In order to push existing users to refer their friends, we're running a special reward program, in which the user is given an instant $10 discount for posting a referral message as their Facebook status. The message looks something like this: "Check out this company Jerry.ai --they automatically checks if you’re paying the lowest price for insurance and they will also find the best quote for you. As a friend of mine, you can get $20 off your insurance purchase. Click here to get the $20 gift credit: jerry.ai " The user is given the option to post this message on his/her Facebook account during the purchase. Once they post this message, they instantly get the $10 discount on their purchase. In other words, we don't wait for any of their referred friends to actually signup with us before giving them the discount. We feel that doing this would make the users more likely to post the referral message. Assume that this reward program has been running for a couple of months, and we have some data collected in our database. We want to know if running this program has been a good idea or not, i.e., are we acquiring new customers with it, or are we just losing money by giving out $10 discounts. Assume that you have the following database tables: 'User' and Purchase' Table User id, name, referring_user_id Table Purchase id, user_id, date, total, discounts In the Purchase table, the 'total' field contains the dollar amount of the job. The 'discounts' field consists of the total discounts given for the appointment (including rewards, coupon redemptions, etc.). Therefore, the customer pays: 'total' - 'discounts' as their final bill. Given this data: 1) 'What' information would you derive from it, and 'how' will you derive it (you can give SQL queries, pseudo code, ... whatever you're comfortable with) 2) Using the information from Step 1), how would you make a recommendation on whether this rewards program should be continued or discontinued

Coding: one leetcode medium around string, parenthesis, backtracking, etc. ML: different supervised algorithms, pros, cons, comparison. Advanced deep learning, time series models, differences between traditional ML. Latest architectures like BERT, etc.
avatar

Data Scientist

Interviewed at Lucid Motors

3.2★
Nov 14, 2023

Coding: one leetcode medium around string, parenthesis, backtracking, etc. ML: different supervised algorithms, pros, cons, comparison. Advanced deep learning, time series models, differences between traditional ML. Latest architectures like BERT, etc.

Viewing 1311 - 1320 interview questions

Glassdoor has 43,845 interview questions and reports from Science interviews. Prepare for your interview. Get hired. Love your job.