Collaborate on Job Estimate Template Word for Product Management with Ease Using airSlate SignNow

Job estimate template word for Product Management

how many drivers does uber  need to serve the bay area   hey everyone welcome back to another exponent  product management mock interview my name is kevin   and on today's show we have selena we're going  to be doing an estimation question but before we   dive into that selena do you mind just telling  the audience a little bit about who you are   yeah for sure and thanks for having me um i'm  selena i'm a product manager at google and i've   been there about two years now and prior to  that i was an apm at yahoo cool thanks for   coming on today's show we're going to be doing an  estimation question like i said and this is what   i'd like to ask i'd like for you to tell me how  many drivers does uber need to serve the bay area   how many drivers does uber need to serve the  bay area okay i'm typing but nothing is writing   okay there we go there's a little bit of a lag  awesome so i am gonna assume that when we say uh   the amount of drivers that uber needs um they're  talking about not necessarily how many people are   like you know on staff in general just more like  given some peak traffic time how many uh drivers   are needed to cover that demand yeah we can do  that that sounds good awesome cool so then um   let's say uh let's say we're just gonna visualize  like you know it's a friday afternoon at like 5   30 pre-covered conditions like people are either  coming back from work or going out to like dinner   with their friends and so at a really high  level the way i think we can approach this   is um we want to know how many you know writers  are being demanded per you know this peak traffic   hour um sorry rides are being demanded uh per  peak traffic hour and then we also are going to   want to calculate how many of these rides per  driver um can be done within an hour as well   um and so if we take the demand and divide it  by uh the supply we'll come out with the number   of drivers in the end so basically rides per  hour divided by um rides uh per driver per hour   will get us the number of drivers awesome so if  we tackle the left hand part of this first um   in order to find how many rides are being demanded  let's sort of take a sense of like how many people   we're working with in general so i know that there  are about seven million people in the bay area and   you know if we sort of think about the  types of people who would be picking uber   um someone who's under 18 can't really take uber  because it's part of just their policy and say if   you're above you know 60 you might be too old  to be familiar with the technology um and so   let's say we have like a 80-year life  expectancy for people um and if we assume   a uniform distribution of ages uh cutting out  under 18s and over 60s that's going to be about   3.5 million people who could potentially use uber  alrighty and so now that we have uh this number um   i also want to call out that the amount of rides  would differ depending on where you are in the   bay area um so i don't know if you're familiar  but you know an sf most people don't have cars   they walk they depend on public transportation  um and in the south bay things are much more   spread apart so like most people actually do have  a car um and so you know the demand would be less   there so um i do want to break it out by like sf  versus everywhere everywhere else um and so i know   that sf has a population of uh one million and so  if we just subtract out like for everywhere else   that's gonna be like 2.5 million people um  does this make sense so far yeah yeah this   is a good segmentation on the geography and for  for the people who who might not be familiar um   sf is more city-like whereas everywhere  else is a little bit more suburb-like   yeah exactly um cool so now if we want to sort  of take these two separate geographies um you   know not everyone out of these 1 million and 2.5  million cohorts are going to be ubering again like   some people might walk drive you know shuttle  if you're working in tech um and so i think   one way to sort of approximate this is you know  given uh most people in sf do not have a car um   and you're thinking about the average friday night  uh we know there's three people per household you   know maybe you have roommates maybe you're a small  family um and so if that's true that's going to be   a and you say one person per household has plans  that are not within walking distance or on like   a bart line um so that would give us 33-ish  percent of the 1 million um which is about 333. let's just make it 350 to make it a  little bit easier on us in the future um   and so that would be how many rides per hour are  coming from us and so given what i said earlier   about the south bay just having like way more cars  i would i would um guesstimate that the demand is   probably 10x less um like literally everyone has  a car here so if we just sort of um cut that down   to 10 we would get point five k um or sorry yeah  three uh we would get three point three percent   um which let's make that five percent for  math um and then five percent of 2.5 million   is 125 k i believe and so if we total  up 350 k plus 125k that will give us   the total number of rides in the bay area  in that given hour and that's about 475k   alrighty okay so that's the first part of the  equation now we kind of have to think about   okay out of all these rides um how many  rides can a single driver handle per hour   and so um let's see i think you know when  i think about the average trip in an hour   um i'm thinking it might be you know 15 minutes if  we if we decide to include uberpool because uber   has a benefit of like you don't have to go back  to the start you can just pick up where you left   off and start taking the next person to wherever  they're going um so keeping that in mind um i   think a 15-minute trip is reasonable so it would  be you know four trips per hour and you know given   the popularity of pool let's say um on average  like you know you have the single riders but you   also have you know pools with like you know three  to four people so maybe like two people per trip and of course we can revisit this based on  like how many people we think are actually   using pool versus other types of uber but for  now like let's go here um so that's gonna be   um eight rides per driver  per hour okay so now if we go   back to the original equation we wanna take  these two numbers and divide and so we get 435k divided by eight and that is  about 60 62 k let's just say 60k   um okay so now i've gotten to some final answer i  kind of want to check to see if this is close um   one way i might think about is like out of a seven  million person population like how many do i think   are actually uber drivers and 60k is a little  less than one percent of the total population of   the area it actually makes a lot of sense um like  you know given tech is a really big industry here   i don't imagine um you know more than one percent  of the would you know have the dispensable income   to like spend on ooze and so it makes sense that  you know less than one percent of the population   uh would be you know drivers for people that can  afford extra rides um so i'll pause there in case   you have any questions on how we got here yeah  this this totally makes sense one question that   i have you don't have to go through the math for  this but you made the assumption here that we're   that customers are taking uberpool what if this  is during covet times but with the same demand as   non-cova times and everyone wants to  take uberx so uh one customer per car   where would you change some numbers if  you had to go back and make those changes   yeah good question so assuming you know we're in  covit and they've not allowed pools um this part   would go away it would basically just be like a  very serial process you would have to pick up one   person at a time but four would remain the same  because i'm assuming the distance that someone   is going is still about the same okay um and if  you i think in here you you made some pretty good   assumptions where most of these based around your  intuition and if so what sort of additional data   might you take if you had access to additional  data to make your estimation more accurate   yeah definitely i think um a lot of my more like  intuitive guesses are around like how many trips   per hour and how many people would be in said  trip uh if i had data around like the split   of usage between like uberx uber like is it called  luxury uberpool um we would be able to have a more   accurate estimate of how many people provide  um and then if i had data around like how far   you know different mooney and bart stops are  and what the split of geography in terms of   um public transportation coverage um i could  potentially do a better job with the number   of trips per hour like for example maybe it  takes longer to get places in the south bay   and so we would want to split like maybe  it's like two rides per hour in the south bay   but in sf maybe it's actually much shorter  rides and maybe it's actually like five or six   um so i would want to layer that in as well  cool um yeah i i think that makes sense i   think your gut check makes sense is there  anything else that you might want to mention   um no i think um overall uh yeah i think  the gut check makes sense um i think   uh given that given that we're thinking about like  the total number of drivers it was important that   we think about peak only because from like a user  experience perspective like that's all that really   matters like we don't care about the lazy you know  sunday where no one is going anywhere um the peak   is what really matters here totally great so i'll  take off my interviewer hat i've written down some   feedback here that i'd love to share with you and  at the end i'd also love any tips that you might   have for the audience who might be preparing  for an estimation type interview so i think i   think you did a really good job in the beginning  clarifying what the question means i think with   this question it could mean any number of things  you wanted to focus on peak time which is fine um   i think there are some other approaches you could  have taken one was like what you mentioned take uh   looking at the aggregate number of drivers  needed that that would have been a whole another   uh route um if you went that path you also had  some pretty good assumptions that you say to them   up front such as assuming that the people with 0  to 80 life expectancy have uniform distribution   i think you also did a good job breaking down  the geography so uh taking the bay area and   not treating it as some uniform land mass  but really thinking about sf and non-sf and   yeah you also did a great job of showing your  soft skills in terms of not just answering the   question but also taking the time to check in with  me making sure we're on the same page and making   sure that uh you you check to see if this was an  overestimate or an underestimate at the very end   so um all in all great answer and yeah i'd love  to hear if you have any tips for the audience   who might be preparing for their own estimation  question interviews yeah awesome thank you for   that feedback i think i agree with the general  feedback you gave i think it's important um   after like every big chunk of your answer to  just pause and give your interviewer a chance   to interject um because they probably do want to  ask questions at some point they don't want to be   silent the whole time and another one is yes just  being really clear where you're making assumptions   and where you don't have the data and just telling  your interviewer like hey like if at the end we   aren't really happy with the outcome like let's  go back and visit like this particular estimation   cool thanks selena and that's it for today's video  for the audience good luck with your upcoming   pm interview thanks so much for watching don't  forget to hit the like and subscribe buttons below   to let us know that this video is valuable for  you and of course check out hundreds more videos   just like this at try exponent dot com thanks for  watching and good luck on your upcoming interview