Facebook Interview Question: I don't think there are any v... | Glassdoor

Interview Question

Product Manager Interview Menlo Park, CA

I don't think there are any very difficult questions. You

  should always ask clarifying questions if something is not clear. I had to ask many clarifying questions.

Interview Answer

5 Answers


Lets talk about the photo thing,

Improvements have to do with metric. Just improving a system is vague. I would ask if there is a metric that you are currently wishing to improve. Time spent on photos or adoption of photos or overall engagement? Most times they will tell you to explore the metrics that you would want to improve.

Overall engagment will include:

1) Time spent on Facebook:

Add photo editing tools. a paint in Facebook (instagram did this very well)

2) Adoption of Facebook phot

Machine intelligence annotations on photos. Build upon the idea of how photos can recognize faces. Photos should also recognize buildings, dogs, landscape e.t.c. This way, photos can be a way to discover the world. An exmaple use case is you are in Japan, you see ruin, you take photo with the discover feature turned on - Facebook annotates it for you. In addition, you can start doing interesting things such as suggesting places where others who posted the photo went. You might even have ads, for those places.

Anonymous on Dec 21, 2014

A natural follow up question is how you would measure if this Annotation feature is worth to pursue.

You have to do A/B testing where one version of the platform has the feature and the other does not have. Measure for example, shares. A photo that was annotated vs the one that was not annotated. Which one led to more shares?

Measure how likely the suggested places can invoke clicks. Say for example you have posted Niagra falls, and facebook annotates it to Niagra and then suggests nearby places of others that went there. How many times people click on the other suggestions. In one version, there is no option to even click. What you are looking for is how they will click.

Anonymous on Dec 21, 2014

Going back to the photo question - Improvement based questions are really focused on design. So it's important to go back to the fundamentals....

Ask clarifying questions - i.e. who, what, why, and how
Identify users, then use cases, needs and goals
Prioritize the goals
Focus on one goal, and list solutions for improvement
Tradeoff the solutions and select the best one as your recommendation

Anonymous on Jan 17, 2017

RE: "Newsfeed engagement has decreased 2% week-over-week. why?" or "You're sitting with a data analyst, and you know there's a mobile problem in our new release. What would you ask him?"

Comprehend the situation and then brainstorm possible causes:

In comprehend the situation, we need to ask some questions:
- How is engagement defined? is it likes/posts, etc.
- Over what period of time, 2 weeks, 1 month, 6 months, etc.
- What platform, mobile, web?
- What markets, USA, Europe, Asia, etc? Where's the root cause?
- What age demographics?
- any particular users affected more so than others

- User changes
---more users on a specific platform over another, i.e. mobile vs web, android vs iOS
---competitors affecting engagement, i.e. snapchat, others
---substitutes, video games, etc.
- Logging Changes
---information not being reported properly
---backend issues
---human programming errors
- Inbound or Channel Changes
---socio economic changes
---geographic changes
- App or Software changes
---new features/change of features causing issues

Anonymous on Jan 17, 2017

"Tell me about a time when you experienced a challenge building a product. What did you do?" or "What's your biggest accomplishment?"

Answer using SAR


Anonymous on Jan 17, 2017

Add Answers or Comments

To comment on this, Sign In or Sign Up.