You are such a friendly interviewer compared to those cold and impatience one!!!!
@bibiworm
3 жыл бұрын
Engagement metrics: % of DAU who uses chat, average # of chat sessions per user per day, average number of messages sent per user, average number of messages read per user per day, average time between message and response per user per day, average time spent on chat per user per day, etc
@mrfolk84
3 жыл бұрын
👍
@yujiefu2251
2 жыл бұрын
It helps a lot, thank you
@bibiworm
3 жыл бұрын
At 16:00, she finally talked about the impact of the green dot on the goal of chat function.
@andranik7615
2 жыл бұрын
I can see so many boastful comments here from people nitpicking imperfections of the interview. For those who are here to practice for an upcoming interview, remember, there are no perfect answers in a case study interview, and the interview can diverge to one area or another depending on how to conversation goes and your interviewer. Here, Jay was happy for it to go in such direction organically. Even if the interview has some shortcomings, it's still very useful to see how certain problems are being tackled.
@adityapadala
3 жыл бұрын
Could you restrict the case interviews to 20 mins - 30 mins? In most interviews case question is combined with technical code or background project question.
@iqjayfeng
3 жыл бұрын
Yep we went a little more in-depth with this one
@ozziejin
2 жыл бұрын
the interviewer probably would've cut her off or guided her back on track or lost patience after the first 5 minutes
@vikash252
3 жыл бұрын
can someone please summarize this very crisply!! its all over the place!
@littlemida
3 жыл бұрын
It would be good if we have a time limit for each interview as many companies I interviewed with only take 20-30 mins for case interviews?
@bibiworm
3 жыл бұрын
I am very lost with this question. Defining metrics of course is an important part of it, but I thought the focus was on how to test without ab testing...
@rogerronan2222
3 жыл бұрын
i realize it is kinda off topic but does anyone know a good site to stream newly released series online?
@makaiyosef241
3 жыл бұрын
@Roger Ronan i watch on flixzone. Just search on google for it =)
@one_man_corp
4 ай бұрын
This seems to me is a common pitfall where people didn't clarify enough about what the project's context and goal is before going too far. As interviewer, instead of just letting them fail like this with thoughts all over the place, sometimes you need to interrupt and ask questions such as "take a step back" to help them back on track.
@alanzhu7538
3 жыл бұрын
13:23 I think we want the inactivity streak to be right skew, right? So more users came back in a shorter period of time.
@oliveiralgm
2 жыл бұрын
The number of conversation sessions initiated (wether they're new or not) per user or total and time spent on chat/user would be good metrics to track. If these metrics go up and hav stat sig in users in the phased launch would tell us if the feature was successful.
@ericdiaz1701
3 жыл бұрын
This helped me prepare for an interview! Thank you a ton!
@AndrewMoMoney
3 жыл бұрын
Nice job, Rim! New video by Jay, gotta binge it.
@bibiworm
3 жыл бұрын
How is she going to measure draft to complete percentage at message level? A message is either finished or not finished. Plus, if we are analyzing at user level, which is most likely the case, this metric would be inappropriate, and has to be aggregated to user level.
@jyotigoyal7898
3 жыл бұрын
To track user retention by plotting the inactive curve, wouldn't we want to make the curve right-skewed (bulk data to be on the left side)?
@bibiworm
3 жыл бұрын
No, because x-axis is the number of days inactive. Bulk data to the right.
@jyotigoyal7898
3 жыл бұрын
@@bibiworm Oh right. Thanks, Yi
@swynd
3 жыл бұрын
@jyoti i think you were right to begin with. We want number of days inactive to be as low as possible, so we would want the bulk of the data to be on the left (closer to 0), so a right skew.
@pranabislam7875
2 жыл бұрын
Yeah they made a mistake thinking they'd want it to be left skewed. Right skewed for distribution of inactivity is what you'd probably want if you were LinkedIn
@anuvaagarwal3492
7 ай бұрын
13:39 In the distribution of various periods of inactivity, if the inactivity streak is organised left to right in ascending order, we would ideally want the distribution to be right skewed, isn't it? We want a larger number users to have a low inactivity streak. Kindly clarify 33:00 How would you calculate the likelihood of messaging after the user has seen the green dot?
@bibiworm
3 жыл бұрын
If she could be a bit concise, that’d be better. I mean 6 mins into the interview, she still didn’t get to the goal of LinkedIn or LinkedIn chat.
@datahat642
3 жыл бұрын
The other supplementary ideas took a major portion of the interview over the objective of testing without using A/B testing method
@alexz7032
3 жыл бұрын
Great content!
@bibiworm
3 жыл бұрын
Pre-post is never the way to go...
@esthergao4049
2 жыл бұрын
Hi Yi, could you please suggest some better ways? In my current work, we usually just compare them with similar users (e.g. through PSM). Thanks!
@bibiworm
3 жыл бұрын
33:35, again how do you define conversation with the metric # of messages sent/# of conversations?
@SoumyaDasgupta
3 жыл бұрын
Just one review, this interview isnt very engaging. Maybe Rim answers the question with a proper clarity, so she takes the answers very slow and answer calmly. Thanks both tho.
Пікірлер: 35