You found that only 10% of the allotted slots for databases are being used by your customers. What would you do?

My thoughts on answering this type of Google technical interview question:

1. CLARIFY:
   1. Is this database a Google owned one? For example, it is Google Cloud? Yes. 
   2. Would you confirm the problem is we want customers to be using a higher amount of their allotted slots in the database v. we need more customers in general (i.e. acquisition problem)? Yes - the former statement is correct.
   3. Are the customers isolated to a particular region or are they a specific type of customer (ex. new customers only, etc.)? You choose.
   4. Are the customers equally split between Apple v. PC? Yes.
   5. How long has this issue been happening? Is this issue very recent v. long standing? More recent development.
2. GOOGLE CLOUD BACKGROUND: Google's mission is to allow users access to the world's information. Google Cloud Platform offers a suite of cloud computing services that run on same infrastructure as Google search, Gmail, etc. Google Cloud provides users access to data storage, analyics, etc. For example, it offers BigQuery, which is a data warehouse that allows users to run analytics over large amounts of data. Google Cloud's offerings clearly fit within Google's mission to allow access to informatino.
3. SLOT OVERVIEW: A BigQuery slot is a virtual CPU that executes SQL queries. BigQuery automatically calculates how many slots are required by each query, depending on query size and complexity. Users can either pay for a certain number of slots or pay as they use them. PResumably, this issue is only occuring for users who have paid for slots but have not used them.
4. POTENTIAL ISSUES: There are two types of factors we can look for re. why users are not using all of their allotted slots: external and internal factors. 
   1.  External Problems: 
      1. Overestimation of Need: Users are overestimating how many slots they need / purchasing more than necessary from Google. Data sets they're using may not need as many slots.
      2. User Awareness: User is not aware of how many slots they purchased / have accessible. Potentially forgot how many they purchased.
      3. Job Incompletion: Users are not bothering to execute all query requests they need / run the analytics they need.
      4. External Competitor: External competitors have entered space and are taking users who would be using Google Cloud / BigQuery. Users are running requests elsewhere.
   2. Internal Problems: 
      1. Software Issues: User is making query requests but software is not registering it. Were there software updates (bug patches, updates, etc.)?
      2. Query Estimation: BigQuery is supposed to automatically calculate how many slots are needed. Algorithm to assist with calculation could be off / under estimating need. 
      3. Optimization: Alternatively, BigQuery developers found a way to lessen the number of slots needed per database - i.e. optimizing time / efficiency. Understanding of how many slots needed for databases by team is now incorrect / needs to be updated. Metrics on users needs to be updated.
      4. Metrics: Metrics on engagement are off / 10% is not actually an accurate. 
5. HYPOTHESIS: Let's assume interviewer confirms issue is an external problem: overestimation of need. My hypothesis is that users do not adequately understand how many slots are needed to be purchased for the size of their data sets. 
6. ACTION ITEMS / POTENTIAL SOLUTIONS: 
   1. Google could send surveys to users that have not used all their slots asking reasons why: did they overestimate need, how did they estimate their need, etc
   2. Google could document on website / in BigQuery guides recommendations on how many slots users should purchase given their database needs. Could also include examples. 
   3. Google could prompt users that have not used all their slots after a certain time period to use them (ex. send email, mobile notification, etc.). Could suggest ways that users could optimize their slot usage based on their datasets.
   4. Google demonstrates new ways that users can use BigQuery to increase engagement on platform / user.