Really awesome. Some great questions. Did feel at times like Lauren was tried to yield water from a stone with the interviewer but I appreciate her persistence and calm!
Excellent root cause analysis, and I will consider this a great example for debugging any issue And I loved the questions that PM asked the interviewer
That was a great answer. Just a few additions I can think of : Checking : State wise cut New/ Old user logins % users sending requests across New/Old #Requests/ Users for Old/New Male/Female Cut Also Requests have 2 main sources : 1. Search Traffic 2. Recommendations Checking if % conversion across Search and Recommendations has come down. Probably are recommendations got broken.
While the structure is great, I found it a bit odd that there wasn’t a discussion/alignment on 1) definition of ‘friend requests’ and 2) impact on overall user count. Some additional questions I’d ask : -Is friend request here = # of requests submitted or # of requests accepted ? -Did we change anything on the user workflows on how a friend request can be submitted ? (eg friends can be added from their profile or via suggested friends feature)
I don’t believe this is a good answer. She did well in terms of having various buckets for analysis (internal, external, etc.), but the whole purpose is for her to analyze, not ask him (the interviewer) for the answer. Would not rely on this for practice.
This was a very strange interview. The first line of thought should have been if the decrease is specific to the 'friend request' functionality or if the overall user activity has reduced. This would have been a much more cleaner route leading directly to reduced activity in Europe and so on.
The fact that he pointed out this drop occurred only in the US makes me start thinking externally rather than questioning internal issues. If it were a system bug, it would likely affect all geographic locations, unless product versions vary by country. This is something that has to be clarified as well, in my opinion. Is there any A/B test running that targets this specific location, for instance. Also asking about competitors before clarifying if the drop has happened also in other metrics, and most importantly if the time that is spent on Facebook has changed does not make much of the sense. If this has not changed, that means that problem is most probably internal. If she would actually listen to the answers she received instead of asking template questions, they would come to conclusion faster.
Interesting but... annoying haha. If the interviewer is the person with all the info, then as the PM I'd just ask what has been abnormal lately. And then after he relays some info, then you can draw potential conclusions from there. It does show how she was critically thinking, however again if she had the metrics or other information infront of her this would expedite the process. With all that said, it just all sound like one make it up as you go situation.
The interviewer gives the answer at 8:58. The candidate still does not get it . The interviewer then further explains that this is the reason FB friend requests are down . I would consider this a poor interview.
The law in Europe may not have been enough to cause a 10% decline in US friend requests, so she was checking more areas to incase the issue had multiple factors. If the law was passed in US it's a more direct route , as a 10% overnight decline in that case is more than reasonable. People typically send the majority of their requests within the same country, not across continents.
Don't leave your product management career to chance. Sign up for Exponent's PM interview course today: bit.ly/3CYaAln
Very unique interviewer style! Loved Laurens approach of leading with questions and creative pivots. I used to work with her, and she is such a star!
At Facebook, right?
Really awesome. Some great questions.
Did feel at times like Lauren was tried to yield water from a stone with the interviewer but I appreciate her persistence and calm!
Great way to describe it. Could’ve been a better story if narrated a little differently. Good demonstration of pushing through
Excellent root cause analysis, and I will consider this a great example for debugging any issue
And I loved the questions that PM asked the interviewer
It feels more like a fishing expedition than drawing from well structured issue tree. Good attempt nonetheless.
That was a great answer.
Just a few additions I can think of :
Checking :
State wise cut
New/ Old user logins
% users sending requests across New/Old
#Requests/ Users for Old/New
Male/Female Cut
Also Requests have 2 main sources :
1. Search Traffic
2. Recommendations
Checking if % conversion across Search and Recommendations has come down.
Probably are recommendations got broken.
Hey Shivam, glad you liked the video! Thanks for sharing your thoughts as well 💪
perfect answer!
This RCA interview was quite different and difficult to tackle with. Still Lauren did a great work. 👍🏻
Glad you enjoyed it!
While the structure is great, I found it a bit odd that there wasn’t a discussion/alignment on 1) definition of ‘friend requests’ and 2) impact on overall user count.
Some additional questions I’d ask :
-Is friend request here = # of requests submitted or # of requests accepted ?
-Did we change anything on the user workflows on how a friend request can be submitted ? (eg friends can be added from their profile or via suggested friends feature)
Hey Pratyush, those are great questions. Thanks for sharing!
This approach would take a PM a team and a few days to execute. User segmentation on target metric would do this in 5 mins
I don’t believe this is a good answer. She did well in terms of having various buckets for analysis (internal, external, etc.), but the whole purpose is for her to analyze, not ask him (the interviewer) for the answer.
Would not rely on this for practice.
This was a very strange interview. The first line of thought should have been if the decrease is specific to the 'friend request' functionality or if the overall user activity has reduced. This would have been a much more cleaner route leading directly to reduced activity in Europe and so on.
Bring her back on. I need pt 2.
The fact that he pointed out this drop occurred only in the US makes me start thinking externally rather than questioning internal issues. If it were a system bug, it would likely affect all geographic locations, unless product versions vary by country. This is something that has to be clarified as well, in my opinion. Is there any A/B test running that targets this specific location, for instance. Also asking about competitors before clarifying if the drop has happened also in other metrics, and most importantly if the time that is spent on Facebook has changed does not make much of the sense. If this has not changed, that means that problem is most probably internal. If she would actually listen to the answers she received instead of asking template questions, they would come to conclusion faster.
The approach was really smooth
awesome mock interview
Interesting but... annoying haha. If the interviewer is the person with all the info, then as the PM I'd just ask what has been abnormal lately. And then after he relays some info, then you can draw potential conclusions from there. It does show how she was critically thinking, however again if she had the metrics or other information infront of her this would expedite the process. With all that said, it just all sound like one make it up as you go situation.
Is it her job to ask him or her job to tell him where she'd be looking? I'm confus.
Very unstructured answer.
The interviewer gives the answer at 8:58. The candidate still does not get it . The interviewer then further explains that this is the reason FB friend requests are down . I would consider this a poor interview.
The law in Europe may not have been enough to cause a 10% decline in US friend requests, so she was checking more areas to incase the issue had multiple factors.
If the law was passed in US it's a more direct route , as a 10% overnight decline in that case is more than reasonable.
People typically send the majority of their requests within the same country, not across continents.
Loved it