Profitability consulting case interview: Self driving Uber (w/ Bain and McKinsey Consultants)

[Music] Tessa how are you doing today good thanks how are you I am fantastic where in the world are you coming from where are we where are we talking right now

I'm actually based in Malta right now so kind of a fun location what about you uh Malta feels like a madeup place that maybe you'd only hear in Master and Commander or some one of those like

Patrick what's his name books about sailing so it's good to hear that it's a real place and people still are there that's exci I'm in Minnesota it's far less exciting than Malta it's basically

like the anti- Malta in many ways well as a Canadian we welcome our fellow um we we think of you as one of us so very pleased that you're joining me

from Minnesota that's awesome um so we have a great case lined up today um just a little bit of background on the case um uh this is going to be kind of like

in between what is like a very interviewer uh like lead case and an interviewee lead case but I am going to ask you to kind of like Take the Lead maybe more than some of the typical

cases we we do where I'm just driving the whole time does that make sense great sounds great all right you ready to start sure I am let's do it all right

uh so this case is about Uber and weo whmo being Google's self-driving car unit um Uber's management is evaluating the profitability impact of switching to

fully self-driving where every every car is 100% automated and drives itself they've partnered with Google's whmo self-driving unit on the driving technology front and continue to run all

the operations including the user-facing application incar experience payment and customer support they've hired your firm to help them evaluate the profitability

impact of switching to this model the fully self-driving model in a single City that's like the scope of our analysis which is just Austin Texas so

so before we launch in um just so you kind of like grock the idea right like Google owns the cars Google owns like all the technology all the sensors that makes them self-driving they just

provide these cars to Uber and then Uber basically operates it as a self-driving unit if that makes sense yep makes a lot of sense all right uh so any questions before we get started yeah so I'd love

to hear actually a little bit more about that so it sounds like and I'm familiar with the general Uber model where Uber runs the platform and Uber really runs the um the sort of Technology but then

the the drivers need to take care of their own like cleaning and their own insurance I believe um and really just making sure paying for their own gas is that is the split of responsibilities

the same in in the weo model would weo be ultimately responsible for all of the things that the drivers are currently responsible for or is there anything else that would sort of change in that model yeah uh so Whow is respons

responsible for like operating the car right making it actually drive down the streets maintaining the car so that's like repairs if there's an accident or if like the self-driving Hardware or one

of the sensors gets knocked whatever um and then like continuing to develop and refine the actual Tech underneath it right like the software that drives the car Uber is responsible for maintaining

the user facing application right like where you like book The Uber app uh the incar experience so basically like what screens are available can you in the car

like adjust your ride um uh customer support and the cleaning uh and then also you can think of uh Uber is responsible for like the fuel in this scenario okay so it sounds like the

cleaning and the fuel are the incremental and then there's a couple of more Tech pieces that they might be adding like into the car um in addition for sure okay sorry I think I interrupted you in the middle of your description of the case but no no no

that I mean I think I think you basically grock it and so my like kind of overall question right is like how would you approach this okay great um I think sort of my next question for you

is can you tell me a little bit more about um exactly um what the Uber situation is like in this city today so you know how many rides are currently

being done what's the average cost per ride um I think getting a little bit of a better sense of like is Uber super popular in this city today or not quite um would be would be great to understand

yeah uh I mean it's Austin right tech for City uh so Uber does uh 10,000 rides a day in Austin uh and the average like Fair per

ride is $20 and I also give you like a little bit of a clue on kind of the economics as they sit today um So currently in the Uber model the drivers

take 75% of the revenue from them from a ride uh Uber takes the remainder so it sounds like uber then makes about $5 on

average per ride in each of those 10,000 rides that they're doing revenue revenue yes perfect um one last question for you I'm wondering if you have any ideas or

tips about the maturity of the wio technology um I'm not I've actually never ridden in one um but I've seen a lot of videos people have like posted Instagram stories about it um I've seen

some of those videos on Tik Tok where they're all in a parking lot like kind of grid loocking each other so would love to hear a little bit about um is this technology that's proven is it

ready to go um any concerns from that perspective yeah I mean you know the robots get a bad WP I would say uh you know I've been in plenty of like human

operated Uber rides where I'm like oh this person like should be replaced by a computer immediately this is crazy um so I would say like you can think about the

technology as um being better than people right so in terms of like number of accidents from a safety perspective right like uh uh you know maybe uh the

computer's ability to discuss you know the Mets game or you know their favorite Kinks album is a little bit different but certainly from a driving perspective uh the technology is mature and better

than a person right if you've ever been in one of these it's actually like typically they're pretty cautious right the computer is pretty cautious um and so yeah like there are going to be

accidents right like crazy things happen but you can think about having fewer accidents with the robot driving than with a person driving perfect okay that's helpful um do you mind if I take

a minute to sort of structure my thoughts here no let's do it okay are you ready to go I'm ready to go perfect um so considering that this

is a profitability case and really what we're doing ultimately is identifying what the differential in profitability will be between the current state of

uber operating in Austin and this potential future state of uber plus weo operating in Austin um the way that I see the main structure for this case is

to look at it through a lens of profitability um so if we think about overarching profit equals Revenue Minus cost what we want to understand is what the revenue is for each of those two

options and what the cost is going to be for each of those two options when we look at the revenue bucket it seems like right now in the just regular Uber model

um it's the number of rides per day multiplied by the um sort of amount of take-home Revenue that Uber gets which you've already given me the hint is is

$5 per ride um and for the weo model what we'll want to look at is um one do we think we see an impact to that number of rides um or two do we see an impact

to that take-home $ five dollar will weo take more than a than a driver who's driving in a car will it take less or will it take it about the same um so that's going to be the revenue

comparison then when we look at cost um we already identified sort of during the prompt a couple of differences in some of the costs um so right now Uber

um currently takes care of uh sort of the technology behind the platform they probably have I'm sure they have like a head office that they have people staff out of so they have some headcount and um probably some fancy perks for those

people so there's going to be some cost related to that um but overall right now they don't really deal with any of the cleaning they don't deal with any of the insurance um it seems like in the weo

model we'll need to definitely add on that cleaning cost um we might have some additional sgna I'm not sure if there would be anything else in terms of like um

customer support that would need to be had um and then the other main cost that I see especially during the transition period is maybe some type of marketing costs just in terms of like letting

people know that there's going to be this major switch happening to what they've come to know and love as the Uber model thus far it's really about comparing that the incremental revenue and the incremental cost and then

ultimately understanding what the profitability impact is going to be got it and I think separate to that um the only other bucket that I want to consider is really the risks to this

model um and I think that in this case in particular there are a lot of potential risks um this is a complete business model shift it's something that

maybe some segment of customers are super comfortable with and excited about um but there might be other customers who are the opposite of excited about it um and especially if there was to be

some type of accident or something that goes wrong even if it is on average safer than um a a person driving it's something that needs to be considered in terms of the brand reputation um even

though it would just be a pilot in Austin it's how does that impact the brand overall um and then the other thing that I'd be really concerned about is that Uber globally right now relies

on its Fleet of drivers how do the drivers feel about this um does this impact would would would a pilot in Austin actually impact the global Fleet of drivers and um make it

so that drivers would prefer to work at one of Uber's competitors like bolt or who might sort of set up their own taxi system in a particular city or a particular region um so all of these

just to consider some of the risks with with making this holistic shift um but rather than focus on those negatives I'd love to start with really the profitability um comparison and we can

kind of go into the risks after that yeah so before we go all the way down kind of like that route I'd love to you know one of the things you said is um

you know maybe like most customers are totally comfortable getting into a robot car right but like maybe some of them aren't right like how would you actually like factor that into like a basic

profitability Kil like how are you going to handle that yeah so I think we have the base case which is the current number of rides that occur um I think what we would really need to understand

is like a customer perception survey um in terms of the number of customers who are comfortable versus not comfortable um and I think the best way to do that would really be to go out there and talk

to customers whether it be through survey focus groups um maybe on the app as people are on the Uber app like ordering or tipping after the fact there could just be a quick like $2 off your

next ride if you feel out this survey of how you feel about um driverless cars um obviously talk to people isn't quite as good as actually seeing what their purchase patterns are going to be which

is why you would launch a pilot like this in one particular city rather than doing it throughout the global Fleet um but the other way that I think you might be able to test it is instead of doing a

full All or Nothing pilot um whether you could start to integrate some driverless cars into the City and see what segment of people are using them versus not um I think there's a couple of different ways

you could go about doing that for right now what I'm thinking is that we would want to operate under the base case of the same number of rides if you want an Uber if you want to get from point A to point B if you're on a night out and you

just need to get back home you don't really care who's going to be driving you um and I think some of these um potential sensitivity analysis could be conducted as a future future State great

awesome all right so let's start with current state right and I think my first question is let's calculate the profitability of uber operating in Austin with the standard model great

okay so the way that I think we should look at this is in terms of um per ride the revenue and the cost and then what we can do is just either multiply that

by the 10,000 rides or what I'm actually inclined to do because I don't love math and like to keep the numbers quite small is just also compare that with the per ride profitability of um the future

State and sort of look at those two and then we can really think about what the Major Impact is going to be on 10,000 rides a year does that sound okay y great what uh what information do you need yeah so I think in terms of Revenue

we're actually pretty Set uh we have the $20 we have the 75% which goes to the driver so we know that the revenue per um ride is going to be $5 on the cost side um I'd love to sort of talk through

what some of those costs are going to be and whether you have any of that information so I think on the cost side um I'd first start with um support and I think that the Uber support that could be if anyone has any issues with the

ride maybe the driver drops them off at the wrong spot or doesn't come pick them up um so this is going to be sort of an outsourced team for Uber overall um but I'm wondering if we can just sort of

estimate on a per ride basis if I was to think maybe half the rides need a support and that support costs kind of $1 worth of time um maybe that's about 50 cents per ride then does that sound

about right I think that's reasonable great okay so support 50 cents um then on marketing let's say um you know I'd say a similar amount uh maybe a little

bit more for marketing per ride do you have any information on what marketing might cost per ride I mean let's let's think about uh I know I know you hate math but watch me force you to do it

right so let's say let's just assume marketing was the same right uh let's assume it was like 50 uh 50 cents a ride um what would that imply for like annual

marketing spend in Austin and does that make sense yeah so 50 cents per ride times 10,000 rides per year is about $5,000 um 10,000 rides per day

oh per day sorry so that's $5,000 per day um we can say um let's just say 350 days per year to make it simple 350 Time

5 um that's um that's 175,000 per year got

it7 yeah it doesn't so it's 5,000 5,000 rides a day right yeah times 100 would be how

much F 5,000 time 100 is 500,000 100,000 three yeah about that so that how's that number feel 1.5 million

yeah that feels that feels actually that feels a little bit High to me um I think there are a lot of people who are using Uber today but I think we can go with it I think it's it's a reasonable enough assumption that we can go with it so

marketing 50 cents and support 50 cents um then if we go to um let's say that there's some sense of like research and development product development that

needs to be done um and let's say that's about um I don't know $11.50 per ride this is current state is that right this is current state yeah so this is just regular product development continuing

to make sure that the app has everything that it needs to continue to function new payment methods integrated um what do you think developers developers are expensive let's go I like it okay great

um and then I think there's just some other General um sales and administration that that would need to occur so there's probably legal there's government Affairs maybe lobbying that

needs to happen um to a certain extent and so for that again let's just go back to that sort of marketing and support cost per ride and let's say about 50

cents per ride okay um so in total that would be $3 per ride that sound okay in terms of costs cents for three and then

150 for um the research and development so we're making $5 in um revenue and we're uh spending $3 in cost so that's

$2 in profit um for the um current state got it and that's per ride times 10,000 rides per day and then times um

350 or 365 days per year got it but we're gonna work the difference on per ride basis is that right yeah that's good thinking got it okay good um okay

so given that then we got to switch over yeah um uh what like questions do you have assumptions do you need to make in order to figure out kind of like the revenue and profit per ride in the new

world yeah so I'd love to talk through in the exact same structure that we just did so if we focus on Revenue first um and I think as we've mentioned base case scenario let's just assume that

everything stays the same um same number of rides per day and um same uh $20 per ride but I think the main thing that might change here is actually the amount

um which weo takes home rather than the amount that the driver takes home my initial assumption is especially because we know that some more of the costs are going to be taken on by Uber um that

that that Revenue split is actually going to vary um and my estimation there would be maybe it's a 50-50 split maybe this is an equal partnership Uber takes home half weo takes home half how does

that sound y I think the I think that's right the the robots are going to get paid less than the people which again is another their cause for robot rights

they really need toize it's not great for the robot exactly agreed yeah um so that would mean that the take-home revenue for Uber per ride is actually

$10 per ride rather than the $5 um but let's take a look at the costs and see um and and just sort of thinking about that $2 in profit that we saw in the

base case scenario the cost would need to be $8 or less in order to match or actually exceed the profitability per ride okay yeah that's what we're sort of targeting here 8 or less let's see if we

can get there um and then in terms of the cost I think we should follow the same structure for the four costs that we broke down before but we should think about which ones are the same and which ones might have an

incremental and then we have a couple of new costs in addition to that but let's start with sort of the base cost that we had before um so the first sort of Base

cost uh would be um in terms of the support um I think that this would actually need to increase in the driverless model because people might have more questions especially at the beginning um more questions and you

don't have someone in the car to actually converse with um I also think that people are kind of people Pleasers you don't want to complain about like the nice man driving you from point A to point B but like that horrible robot who

didn't quite drop you off in the right spot um so we assumed 50 cents what if we assumed $1 for this uh for support going

forward I'm really worried about these robots I think you should I know okay keep going perfect um then in the marketing um I think that the marketing

should also increase I think there's going to need to be new messaging new Communications probably a new like look and feel slightly you would also mentioned that Uber would be responsible

for some type of communications within the Uber um within the ride so it's like who's making those assets there needs to be some type of increase here um so again I would increase that from the 50

cents to the $1 for marketing yep uh then when we look at the R&D overhead um so we had assumed $1.50 for this but what I actually think about

this is that most of the R&D that would be occurring in this new shift would actually be taken on by the partner and that's one of the great Parts about partnering with such an Innovative

company like Google and weo um so for this we had assumed 150 in the current scenario but I would actually decrease that to maybe $1 um and say Uber takes

on less of an R&D role in in this case what do you think about that yeah I think that's reasonable um and the Assumption there is

that Uber was previously spending some of its R&D expense on like the vehicle side yeah on on on just Innovations in

General on something to kind of keep up with the market okay whereas Now by partnering weo is what's keeping up with the market that's really the advancing the technology piece okay yep and they I

I'm also guessing that there R&D teams who need to work closely together to say what's going to work for both of these companies to make the partnership whole yeah all right makes sense cool um and

then in terms of other sgna um my biggest concern here is on the regulatory piece which we were talking about and that lobbying two reasons for that one I'm not sure um that Austin would have all of the permits to have a

fleet of driver loose Vehicles Drive around all the time and two um from like that labor side of things and from lobbying to enable Uber to kind of get through I know there's been some

challenges that Uber has had in the past so rather than 50 cents again I'm going to assume $1 and just keep it all consistent all of the current costs are going to be $1 um which totals up to $4

for the current current costs so that is a little bit higher than the costs for the current state um and we still have a couple more costs to

add on so I think in terms of the um the new cost to add on there's sort of a cleaning cost which we can assume maybe would be about 50 cents per ride um

there would be maybe a fueling and or charging cost I think I think weo is electric vehicles but um maybe it's a little bit of hybrid so we'll say either it's fueling or charging maybe you know

the answer to that actually do you know if they're electric I think there's some like weird Jaguar uh I think they're electric they're very quiet so okay okay yeah so

maybe maybe we'll say 150 for cleaning charging it's it's a little bit difficult for me to know because we don't know the distance and um not I don't think I've been to Texas

in like a decade so I don't know what this um electricity costs and fuel costs are there right now um so let's say 150 per ride on fuel charging

yep um and then the technology fee uh which would be maybe um any Ty of fee that we're paying for the actual Tech within those vehicles um actually I think for that if it's okay with you we

would we would have just included that in the marketing cost and we'll sort of we had upped that marketing cost to include the in vehicle technology fee so I think we'll say zero for that so that

would be an additional $2 the two plus four means that the total cost would actually be $6 um making the profit for

the um Uber plus weo option $4 that sound okay sounds great does sound great that's uh that's

doubling the profit that we were making in the current state yep uh and then I'm gonna make you do it again so what would that mean from on an annual basis oh oh

boy oh boy here we go so um on an annual basis so we would say that we have

10,000 um rides per day and that's times $4 in um profit for ride let's just do the Delta because

we'll talk about the it's the Delta so the $2 y so that would be $20,000 per day yep um and then let's do

it your way so we did 20,000 times um 100 yep and that would be 2 million and then we'd want to multiply that by about

3.5 um so let's say like about seven 7.5 million great um and yeah so how do we what do we think about that

number so I would say um it sounds really great to be sort of doubling the um additional Revenue I would say that 7.5 million additional revenue for Uber

seems quite small I don't think profit though right profit but but I don't think that would move the needle on Uber International I think what we need to think about though is if this pilot was

to go well and the driverless fleet was to be adopted everywhere and so everywhere the profit was to double that's where we'd see the impact so what we have to look at this is the absolute

number isn't really that big um but it's the the potential that this uh model has on the overall profitability of the company that's really exciting that and

the ability for the company to really stay ahead of the trend I think if Uber doesn't adopt this you're going to see Bolt adopting it you're going to see Lyft adopting it so it's about being on The Cutting Edge and really trying

something new that all to say I I still am concerned about the risks so I hope that we can we can maybe take some time to think about what some of the risks are and how we would further evaluate

those yeah let's do it what are the risks um I think that the biggest risk is in terms of um is in terms of displacing the drivers and in terms of keeping drivers on side with the Uber

brand globally um my biggest concern with that is I I don't know how many wh cars are manufactured per year but I would imagine it's not nearly enough to

keep up with Uber demand globally so it's going to take a global transition there's going to be a period of maybe a decade maybe even longer than that um and so you need to keep those drivers on

side in all of those cities that there are still drivers driving Ubers um and so I think that's that's the biggest risk I think the way in which you would you would start to deal with that is to

have really clear communication with the driver really focus on a transition plan plan out what city is going to come next help the drivers really plan for their future

um and that would be how I would sort of mitigate that first risk shall I move on to what I think the second risk is yeah um I think the second risk is the regulatory risk and

we talked about this a bit in terms of increasing that cost um but that really is completely unknown to me in terms of how does the City of Austin actually

feel about having all these driverless cars driving around I know you've talked about the advanced technology um but it it is an unknown and it's something that

um some people might not be super excited about yeah and and what do you think like that uh like you know I told you the bar is

they do better than human drivers right at least that's kind of like the the bar that we're inter internally evaluating um do you think that's the same bar Regulators will

use that's a that's a really interesting question I think so I think that comparing against the incumbent is the is the only bar that that would really

make sense to me um what what I think the difference might come in is in terms of what data set you're using and what quantity of um rides you're basing that

safety rating off of um is it that that's true in a very controlled environment is that that's true in the one city that the the cars are currently in or is that something that you can really get behind and I think this is

just um it's probably just a growing pain but it's definitely something that that you need to consider okay good um other risks yeah the other one I would the other main one I would say is in

terms of customers and whether C the customer adoption piece that we talked about I'm a little bit less worried about this um I think one because they've seen how excited people are to drive in Whos so very

anecdotally I don't I don't think it's that much of a risk um I think that Uber is more of a tech forward crowd maybe people who are a little bit less Tech forward are going to be riding in a taxi

anyway um so it's probably more geared towards people who are Tech forward anyway and then the third reason is I think this is just the the way the world is going so even if you have some

percentage of people who lag behind one year two years later you'll catch up to that in Spades um so I I think that that's a it's a pretty minor risk but we've already talked about some of the

ways that you could test that before deploying the solution full scale okay great um one other thing I think it's important for us to consider

is uh the risk of the robot Uprising uh it is we probably should put it in there at least at some point because I've seen a lot of sci-fi and it's coming they're coming for us I'll

leave you to write that in the report I'm not I'm not as familiar with the robot up Rising yeah maybe some kind of robot Insurance we'd have to buy um okay

so net net uh I think we've kind of like laid out the you know the short-term profitability games the risks what's our recommendation here my recommendation is

to proceed with the um fullscale weo uh pilot in Austin Texas I think that um we can see some really really exciting uh

profitability gains as a result of this potential partnership and I think that it's really core to Uber's brand to continue to be on The Cutting Edge um before actually going with the

deployment I would recommend that we double check on the Regulatory and on the customer um adoption pieces which I think are the two major risks that we can learn a little bit more about before

actually going ahead with it um and on the third I would say in terms of the main Next Step as we proceed with the part parip I would want to make sure

that um all Uber drivers have a very clear messaging globally around what the driverless Fleet means for them and for their future um career with the company

awesome well thank you so much thank you hey everyone it's Kenton gavest here the founder of Rocket blocks thank you so much for watching this mock interview video I hope you found it informative and helpful as you are getting ready to

prepare for your own interviews and hopefully Landing that job offer you are excited about we have a ton of great content on the rocket blocks Channel coming out on a weekly basis mock interviews mini lessons and chats with

our rocket blocks expert so if you haven't subscribed yet already please do so there is a big red button below and if you hit that you'll get all of our content as soon as it comes out thanks

for watching and have a great day