I'm Dianna, a product manager who worked at companies like Facebook, Stripe, Google and IBM. You'll get the latest and greatest tips on how to ace product manager interviews and grow your product career.
Hi Dianna, This video is so helpful breaking down how to prioritize metrics! How would you then apply this frame work to prioritizing the pain points or solutions? So let's say they asked For airbnb, what customer complaint would you want to prioritize? Or what would you want to improve first for airbnb - travelers not being able to check in or travelers arriving to a dirty home? Let me know! Appreciate the help and all of your videos!
Most interviews usually last for less than 30 mins where the hiring team member actually talks much more than this and if the person likes chatting then you get less than 25 mins at best. How can someone write so much (The way the person wrote in this video) and cover so many details? Whats happening right now is that a lot of PM's watch these kind of interviews and go on to interview someone and expect people to repeat the same performance within the timeframe which is much shorter than this video. This interview in reality is more than 1hr long. I have attended interviews where the hiring manager was chatting after every 2 mins and asking me to mold my answers to the answer he had in his mind. How can anybody avoid that? These interviews are not honest and lack real time issues faced by the people?
Very insightful video Dianna. For user segmentation, the approach that you have suggested makes sense. However, with this approach, it seems that the user segments ideated can come across as broad and generic and has implications on the ideation of pain points and solutions. How would you suggest we tackle this issue? Is uniqueness of pain points identified and solutions ideated a marking criteria for the interviews? I want to avoid the scenario of coming up with a pain point or solution that is already available in the market.
I was asked the same question in an interview recently with Meta. One follow up question was : If you have limited engineering resources, how would you allocate the resources between small (private) and large group?
@diannayau - I have feedback for your videos. While the answers are good, it's not clear for what levels this level of detail is acceptable. For example if you were interviewing for a Level 7 PM at Meta. Would the above answer still make the interviewer go Wow? How would you do this differently if you were interviewing for a Level 7 PM as opposed to a Level 5 PM? Can you post any mock interview videos specifically focusing on senior level PM roles? Level 6/7
@diannayau - I have feedback for your videos. While the answers are good, it's not clear for what levels this level of detail is acceptable. For example if you were interviewing for a Level 7 PM at Meta. Would the above answer still make the interviewer go Wow? How would you do this differently if you were interviewing for a Level 7 PM as opposed to a Level 5 PM?
Not necessary. I did it in my interview but found that it takes a lot of time to write & speak, and I ran out of time. So I now have to "redo" the Sense interview, where I'm planning on only speaking with manual pen/paper which works better for me.
Great video. Some comments: <1> I would prefer to keep revenue/profit metrics out of scope since they're not product success metrics - you mentioned # of completed rides as NSM, that's a good enough proxy for revenue imo. <2> I would track Driver/Rider ratio at Uber level instead of at the driver-level.
Isn't access most important, since they don't have a card to swipe (apartment spots), or garage-opener (home garages)? If drivers cannot access spots, other pain-points don't matter.
I have a question about the scenarios that she shared relating to failure and conflict. To what degree will the interviewer be thinking in the back of their mind about the potential for these scenarios to come up again? Will this candidate be seen as a risk if the failure or conflict story is large enough that it provides additional cause for concern? Just wondering from a candidate standpoint where the limits are and what we should be comfortable sharing.
🙌¡Thank you Dianna! I've been watching your interview videos, from root cause analysis to product sense and even though I'm not interviewing for Meta they've all been very useful for my interview preparation.
Great video! Before you get into the hypothesis, it might help to also talk aboutvthe difference in the value that feature a (reactions) has va feature b (conments) since you use those implications in the hypotheses
@Dianna Yau How much time did the full interview take? Pain points and solutions are quite detailed - Could have taken well more than an hour. Based on your experience, do you see the detailed notes we see here in the 30-40min interview?
GMV appears to be a better long-term NSM compared to # of items purchased/sold since it aligns with not just buyer & serller goals, but also IG Shops goals, if long-term strategy is to earn a commission based on value of sale. Comments?
Agreed. GMV also aligns better with the goals of the brand selling items and the commissions that creators might be earning. There's some hesitation in general in taking revenue focussed goals but for e-commerce and service marketplace businesses, that's generally the metric that most public companies also report on.
Any suggestion for how to transition from mapping the ecosystem to user segmentation? Assuming the product is a multisided ecosystem, what rationale should you provide on which user group within the ecosystem to focus on?
Hello! Loved your video! Im currently aplying to PM in a digital waller startup focused in GZ, they have three products, minors can open accounts, the facilitator to buying cripto and money savings outside of the economic situation of the country. In this case how would you messure ?
Good video. A few comments: since Meta makes virtually ALL its revenue via advertising as per their annual reports, I feel Ad impressions (for impression-based ads) and actions (for action-based ads) should somehow be part of the metrics tradeoff discussion. For example, if Time spent of FB/day goes down, and we see a corresponding decrease in # of impressions/actions per day, then DON"T ship, even if you see broader participation going up. Even considered by itself, a decrease in # of impressions/actions per day attributed to the experiment would be enough for the DON'T ship decision. What do you think?
Wonderful video! Some comments: Level 0 (top level NSM): It's important to note that NSM is defined within a specific timeframe (week, month etc) so it would be # of nights booked in the last week/month, etc. Level 1 - Define ACTIVE users as those who have made at least 1 confirmed booking in the given timeframe. - It’s important to note that it’s AVERAGE # of nights booked per guest, so you can multiply it by the # of active users to get the NSM - You have not considered that a guest may have made multiple bookings in the time-frame. So a more accurate Level 1 equation would be NSM = (# of active guests) x (avg # bookings per guest) x (avg # of nights per booking) Level 2 - You have dis-aggregated (# of nights booked per guest) into Listings, Views, Confirmed Bookings, and Canceled Bookings, but how do they roll up to nights booked? :) - A more accurate way to setup your Level 2 equation is to use the (avg # bookings per guest) from my alternate Level 1 equation above, and then disaggregate it using the metrics you listed in Level 2.
Good series, thanks...a question. You did mention Zocdoc, which allows you to browse docs by specialty, location, insurance type, etc. and allows you to book appointments. Surely Zocdoc addresses the top pain-points at 12:01 of <1> lack of info to make informed decisions, and <2> need to search multiple platforms, no?
Hi Dianna, why would the supply of stories increase if stories live longer? I understand how the supply would increase because people may engage more with stories and that excites the creator. Is there any other reason?
Also - there are some free online tools to transcribe videos. So as Austin mentioned towards the end - you can now fairly easily create the transcript and then go and doodle with it in ChatGPT or other similar tools :-)
Great Video! One correction i guess is that Waze is mentioned as a competitor to Google Maps and this is true to an extent, but i think it was imperative to mention that Waze got acquired by Google so while it may seem like a separate brand, I believe that Google enjoys the data and of course the engagement of Waze users as well... My 2 cents
@@TheBigCheezeIt Ahh got it! My experience has been to try and aim for 35-40 min and the interviewer will typically prompt you to move on or speed things up. For example, if you are short for time, they will typically ask to you to speed up through certain parts or prompt you take make a decision. That has been my personal experience. Has it been different for you?
I love your videos - they are supppper helpful did you think about creating content for actual work of product manager - how to deal with requirement gathering, roadmap building, ect... that would be great :)