Our team is looking for talented Mobile Game Data Analyst
Being a team member at Fox Cub is a unique opportunity to work with a talented close knit team from around the world from the comfort of your home. We respect people who are honest, direct and have a “go-get-it-done-attitude”. Smart people like you don’t belong in corporations with bureaucratic hierarchies. Fox Cub Games is a company where you will make a difference!
Key Requirements
You’ll help us develop experiments, analyze results, and make recommendations on what we can do to improve the game. You’ll work closely with me and our game designer. My cofounder and I are both Zynga alums, so we take analytics very seriously.
We’ll ask you questions about how the game is doing, and expect you to figure out the best way to gather the data to answer the question. Example question might include:
- “How can we predict which players will become payers?”
- “Is the level curve too hard at levels 85-100?”
- “Was our new sale schedule last week better or worse than the old one?”
- “Which of our machines are the most popular?”
We are looking for someone who is already very familiar with SQL, and ideally also Excel. We use Periscope.IO for our dashboards. Familiar with advanced statistical or machine learning techniques would be awesome, but not required. You do need to be a strong problem solver and clear communicator. You will often be asked complex or vague questions, and will need to think critically to figure out what work (if any) is correct for you to do.
We don’t care where you are in the world, but you must be able to consistently work core hours of 9-5 Pacific Time. We collaborate online with Slack, Teamspeak, and Hangouts.
Strong familiarity with social or mobile games is required. Experience with mobile games would be ideal.
How to Apply
Please complete this short sample test and send your result and your resume to jobs@foxcubgames.com. We look forward to hearing from you!
~~~~~~~~~~~~~
Question 1: Please write a short (<100 word) description of a feature or improvement you would make in your favorite mobile game, and how you would determine whether the feature was successful.
Question 2:
As part of our user acquisition process, we’re trying to identify players who are likely to become payers. The more accurately we can do that, the better we’ll be able to target our ads. We could simply tell the ad network who the payers are, but the number of payers is too few for the ad networks to learn from effectively. It also takes too long for players to convert to payers, and we want to get data back to the ad networks much more quickly.
We’ve been tracking a number of key events that we think could be good predictors of whether someone will be a payer. These are in the “events” table:
- dt (timestamp)
- platform (string)
- user_id (string)
- event_name (string)
Each time any player performs any of the events, a new entry is created. If the event is “level_up” for example, a player could perform that event multiple times, so there would be multiple entries. For this exercise, let’s assume the events that we’re tracking are: “level_up”, “spin”, “out_of_coins”, “collect_free_coins”, “increase_bet”.
We also have a “revenue” table that tracks all of our purchase transactions:
- dt (timestamp)
- platform (string)
- user_id (string)
- usd_spend (float)
and an “install” table to track when players started playing:
- dt (timestamp)
- user_id (string)
- ad_network (string)
How would you go about conducting this analysis? What is the high level structure/plan? What queries would you write? What are likely problems and how do we avoid them? How do we decide which event to use as the best predictor?
You don’t need to write any queries, but may include them to communicate your intention.
We’re looking to understand how you break down a complex analysis task, and how you communicate your thought process.