0:00
Everyone's always promoting dogfooding.
Search this video
0:00
It feels though that you take this to a whole other level.
0:05
Either once or twice a month I will set aside half a day and then I'll go out and I'll drive and deliver.
0:09
My team and I, we design these amazing features.
0:09
They look so good, and then you get in the car and you have a phone which is sitting three feet away from you.
0:17
You're driving at 45 miles per hour. The world just changes.
0:21
This thing that was looking so great in an office setting, now maybe makes no sense.
0:26
When I was asking people about you and what you're amazing at, a motto came up again and again. It was ship, ship, ship.
0:31
If you are going to go dog food and experience this pain and you're going to document it, what's next? You have to ship this.
0:36
You don't ship documents, you don't ship brainstorming meetings.
0:39
What you ship is code in your product.
0:43
What are some of the most impactful pieces of advice that you share with early career product people?
0:49
What makes a great product manager is not five amazing strategic ideas.
0:49
It's the hows and micro decisions that you made.
0:55
Where should I put the button?
0:55
Should I put the screen there or not?
1:00
What should the copy say?
1:00
Go to a job where you can ship multiple products as fast as possible.
1:09
Today my guest is Sachin Kansal.
1:09
Sachin is chief product officer at Uber where he has been for over eight years and where he leads product management, design, and product operations.
1:17
Prior to Uber, Sachin was chief product officer at Flywheel, a VP of product at Lookout.
1:21
He also spent the early part of his career at Palm where he was director of product management focused on Palm's mobile operating system, webOS and their mobile apps.
1:31
In our conversation, we go deep into his passion for dogfooding and now he makes this a big part of the product culture at Uber.
1:35
He's personally done hundreds and hundreds of drives as an Uber driver and also as an Uber Eats delivery person.
1:41
I've had a lot of people come on this podcast talk about dogfooding.
1:45
I've never met anyone that takes it to the extreme that Sachin does.
1:45
Much of what the product team works on at Uber comes from what he and his team discover from these experiences.
1:55
We also talk about why it's important to always have a ship, ship, ship mentality, PM career advice for early career PMs, how Uber is planning for a future with increasingly autonomous cars and what changes over time, what he's learned from Uber's shift to efficiency and profitability.
2:11
Also, a bunch of really good tips for how to avoid annoying your Uber driver and Uber Eats delivery person.
2:16
This episode is full of wisdom for anyone who wants to build better products, teams, and cultures.
2:20
If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube.
2:25
Also, if you become an annual subscriber of my newsletter, you get a year free of a bunch of world-class products including Linear, Superhuman, Notion, Perplexity, Granola, and more.
2:35
Check it out at lennysnewsletter. com and click bundle.
2:40
With that, I bring you Sachin Kansal.
2:40
Many of you're building AI products, which is why I am very excited to chat with Brandon Foo, founder and CEO of Paragon. Hey Brandon. Hey Lenny. Thanks for having me.
2:54
So integrations have become a big deal for AI products. Why is that?
2:59
Integrations are mission-critical for AI for two reasons.
2:59
First, AI products need context from their customer's business data such as Google Drive files, Slack messages or CRM records.
3:07
Second, for AI products to automate work on behalf of users, AI agents need to be able to take action across these different third-party tools.
3:17
So where does Paragon fit into all of this?
3:19
Well, these integrations are a pain to build and that's why Paragon provides an embedded platform that enables engineers to ship these product integrations in just days instead of months across every use case from RAG data ingestion to agentic actions.
3:34
And I know from firsthand experience that maintenance is even harder than just building it for the first time. Exactly.
3:38
We believe product teams should focus engineering efforts and competitive advantages, not integrations.
3:43
That's why companies like you.
3:43
com, AI21 AI21 and hundreds of others use Paragon to accelerate their integration strategy.
3:52
If you want to avoid wasting months of engineering on integrations that your customers need, check out Paragon at useparagon. com/lenny. Last year, 1.
3:56
3% of the global GDP flowed through Stripe. That's over $1.
4:05
4 trillion and driving that huge number are the millions of businesses growing more rapidly with Stripe.
4:12
For industry leaders like Forbes, Atlassian, OpenAI, and Toyota, Stripe isn't just financial software.
4:18
It's a powerful partner that simplifies how they move money, making it as seamless and borderless as the internet itself.
4:29
For example, Hertz boosted its online payment authorization rates by 4% after migrating to Stripe, and imagine seeing a 23% lift in revenue like Forbes did just six months after switching to Stripe for subscription management.
4:40
Stripe has been leveraging AI for the last decade to make its product better at growing revenue for all businesses, from smarter checkouts to fraud prevention and beyond.
4:48
Join the ranks of over half of the Fortune 100 companies that trust Stripe to drive change. Learn more at stripe. com.
5:04
Sachin, thank you so much for being here and welcome to the podcast.
5:07
It is great to be here, Lenny, thank you for having me.
5:10
When we were exploring what to chat about and where to spend our time, there's a bunch of themes that came up both in our conversations and also when I was talking to people that work with you at Uber and other places.
5:18
And the theme that came up over and over is dogfooding.
5:22
And dogfooding is something that comes up a bunch on this podcast.
5:22
Everyone's always promoting dogfooding.
5:27
It feels though that you take this to a whole other level and you do this better than anyone I've ever spoken to.
5:32
You're basically world-class at dogfooding.
5:38
There's a lot I want to talk about here, but before we get into it, just give us a sense of what dogfooding looks like in your world.
5:43
What is the scale that you go to, to dogfood Uber?
5:49
I mean call it dogfooding, call it just obsession about the products that we ask our end users to use and we should be using it ourselves.
5:56
So in my world, what it looks like.
5:56
There are a few apps in my world.
6:01
There's the Rider app that everyone's familiar with.
6:01
There's the Uber Eats app and let's not forget there's a driver and a courier app as well.
6:08
So dogfooding looks like the following for me in general, on an average, I end up taking maybe five to 10 trips or Uber rides every week, which just happens because I'm an Uber rider.
6:21
Now, when I'm traveling, that's probably even more.
6:28
I end up placing maybe three orders, three Uber Eats orders every week.
6:28
Sometimes that's too much.
6:34
And again, when I'm traveling, that could increase.
6:34
If I'm traveling with family, it can be a lot.
6:39
It could be literally all our meals.
6:39
So those are probably more explainable, more understandable.
6:44
The volume goes up and down.
6:44
But then, how do you test the driver app?
6:44
So I actually go out either once or twice a month.
6:50
I will set aside half a day or a day and then I'll go out and I'll drive and deliver here in the Bay Area.
6:57
So I've been doing that for many, many years now.
7:02
I will say that's probably one of the best parts of my job.
7:02
I really, really enjoy it doing trips as well as doing deliveries, and maybe every time I go out, I'll do about 10 to 12 rides and deliveries.
7:16
So that's what dogfooding looks like at the outset. That's the fun part.
7:23
The painful part then is, "Okay, what did I learn?"
7:23
Because it's very easy for me to do this and then just completely forget about it, not document it, use it anecdotally in some conversation but not be able to go into a lot of detail.
7:35
So what I do after each of these instances, whether it's using the Rider app or whether it is driving, I take a lot of screenshots and then I come back and I write documents or I add to an existing document.
7:47
So I have a driver document, I have a Rider app document, I have an Uber Eats document and I'll come back and I'll just write everything that I learned that is probably not ideal.
8:00
I'll add a lot of screenshots to that doc.
8:05
I will tag people because I know who works on what and then I'll just send people my thoughts on what we think we can improve.
8:12
And after that, we have a process internally which where we go through all of those issues, we prioritize them, not everything that I thought is a great idea is necessarily a great idea.
8:25
So we go through all of that and then we fix issues.
8:30
And I'm not the only one dogfooding in the company, but I do follow through on the things that I reported and make sure that things that should be fixed get fixed and as quickly as possible.
8:41
So that's the the end-to-end journey of dogfooding for me in my work.
8:47
How many trips have you taken?
8:47
How many rides have you driven?
8:47
How many food delivery trips have you done total?
8:52
What are the numbers here just over your time at Uber?
8:57
I think from a driving and delivering perspective, in the seven to 800 range now and I don't know as a passenger how many trips I have done. I can't count that.
9:08
Same for Uber Eats, but the driving and delivering is many hundred trips. Wow.
9:14
Yeah, I think on riding on Uber and getting food delivered, that's not so hard.
9:14
To your point, the real hard part is driving and delivering people's foods and taking them places.
9:21
Also on the docs, I think you're also underselling yourself when your colleagues or Ricky told me that you write 40-page reports on what's broken with screenshots and suggested fixes.
9:30
And then like you said, you follow up on all these fixes.
9:36
You don't just put them out there and then hope people do something with it.
9:41
You actually stay on top of them and make people feel accountable.
9:45
Lenny, when I find those issues in our app, I develop a bit of impatience because if I faced it in my 10 trips that day, we have eight million couriers and drivers in the world.
9:52
So imagine the number of times that thing may have happened to our customers, which are drivers and couriers.
10:06
So I develop a bit of an impatience that we have to fix it, we can't.
10:06
Now, some things are small, some things are big.
10:10
I develop different levels of impatience for different things depending on the level of severity, and then if I don't follow up or my team doesn't follow up, and if you don't fix it, then it's a moot point.
10:20
Then I'm just doing it for fun, which is not bad.
10:25
It's great for me to be doing it for fun, but I want to do it for impact and I also want to do it for setting an example with my team as well.
10:30
Now again, I'm not the only one dogfooding.
10:36
We have hundreds to thousands of people at Uber.
10:36
We want it to be a part of the culture, but we don't want the culture to be that every individual product manager should be going out and dogfooding their own feature.
10:46
We want the culture to be that every leader is doing that as well.
10:46
And that's why I don't want to be shy about or want to shy away from writing these documents because I want everyone to.
11:00
And now there are many, many documents written by many, many people at Uber and honestly it's making our products better.
11:06
And I would say even more important than the specific features that we fix or we add, it's the culture of we are not just going to accept certain problems in our products, we are actually going to fix them.
11:19
That's really, really important.
11:25
So let's talk about just why this is so important to you.
11:28
I want to talk about how you operationalize this in the culture of it at Uber, but just as a CPO, I don't know of many that go draw people off hundreds of times deliver food to people's houses hundreds of times.
11:38
I imagine you have many things you got to do day to day.
11:38
There's hiring, there's drama, fires, all these things.
11:42
Why do you prioritize this so highly?
11:47
Here's my framework of how we think about end users.
11:47
Of course, end user feedback is important.
11:53
Everyone would agree with that.
11:53
Every product manager, every product leader would agree with that.
11:57
How do you get that end user feedback I think falls on a spectrum.
11:57
There's the extreme quantitative end of the spectrum and there's the extreme qualitative end of the spectrum.
12:04
And I like to believe that you need to be across that spectrum.
12:09
So on the extremely quantitative part, we all love data. We all look at charts.
12:15
And in those charts, every user is an MAU or a DAU.
12:23
Every user is a number and collectively as those numbers users are showing certain behaviors and we learn a lot from that and that's a very valuable exercise.
12:29
You move a little bit to the left of that spectrum and now you may go into things like surveys where you're actually sending out surveys to thousands of riders.
12:40
You're quantitatively understanding, do they like A, do they like B, do they like C which is also very, very valuable.
12:45
Then you get into a little bit to the center, maybe I have a room of 10 drivers and I'm asking them, it's a focus group, it's a discussion, and for an hour I learn a lot from them.
12:57
It becomes very interesting.
12:57
We do this anytime I go to a Brazil office, I go to our India office, I go to a UK office, at a minimum I do one or two roundtables with drivers.
13:08
I also do roundtables with riders.
13:08
It's just extremely valuable to hear from them directly.
13:14
You move one step further and now I may have a one-to-one conversation with one of them.
13:20
By the way, in the rides that I take every single day, it's such an amazing opportunity for me to do two things.
13:25
As a passenger, I'm talking to the driver for 30 minutes off my commute and I'm basically one-on-one interviewing them and because the driver app is placed in a certain way, I can actually see how they use the driver app.
13:35
So that actually drives me a lot of learning and every day there's an aha moment that, "Oh, this is what he just did."
13:47
A text message came in and then he left the driver app and then he went and addressed the text message and so on.
13:51
So there's a lot to [inaudible 00:13:54]. All of that is great.
13:56
However, until I get behind the wheel or I'm a passenger or I'm using Uber Eats, what I miss is the visceral reaction that you get when something happens. So now, think about it.
14:02
I sit in my office, my team and I, we design these amazing features.
14:09
They look so good on my MacBook or on a Zoom screen, and then you get in the car and you have a phone which is sitting three feet away from you.
14:23
You're driving at 45 miles per hour. The word just changes.
14:23
This thing that was looking so great in an office setting now maybe makes no sense or there's a lot of stuff that you have to change. So that's what you feel.
14:35
But then as you are trying to make money doing that, every single little inefficiency, every single flaw is costing me my earnings.
14:42
And while I don't do it for money, so I don't think I can get to the extreme, extreme end of how drivers feel, I think I can get close.
14:56
And the only way for me to feel that emotion is by doing it myself.
15:02
And I do think that that emotion is very important.
15:02
I'm a big believer in then bringing that emotion to my work, whether it's an emotion of joy, I have many, many moments of joy and I bring that back and I celebrate that and that's very motivating for my team and I.
15:16
But there are many emotions of outrage.
15:16
And I want to bring that to work as well and make sure that we can use it as a motivator to do better.
15:22
So those are my reasons to be dogfooding.
15:28
What's an example of something that's super outraged you that you experienced as a driver especially and/or something that maybe is the most surprising that you learned as a driver?
15:40
Let me start with the surprising and then I think we'll get to the outrage part of it and there may be things that you may not expect.
15:45
The surprising thing as I mentioned before was how the app just feels different when it's actually in a moving car at 40 to 60 to 70 miles per hour.
15:51
That was a surprise for me.
15:58
The positive surprise is there's a dopamine hit you get when you get an offer.
15:58
I get an offer, it's 20 bucks.
16:05
I'm picking someone up from point A, I'm dropping them off at point B.
16:11
There's a bit of an excitement angle and I don't want to overplay it.
16:11
Earners, drivers work really, really hard to do their job, but I think that part of our product, there's a bit of an excitement to it.
16:23
That was a surprise to me.
16:23
I will go into a few individual features, but even at a more meta level, the thing that surprised me the most is we obsess about our app and our app is the center of our universe, but when a passenger gets in the car, the center of the driver's universe is that interaction with the passenger.
16:41
That human touch between the driver and the passenger is something that we don't talk about in the office enough.
16:52
And if you talk to the driver and when I'm a driver, if the passenger is friendly, it lifts my mood.
16:56
If the passenger is grumpy, then I'm unsure what to do.
16:56
If the passenger is rude, which has also happened, that can really be demotivating.
17:02
I want to share one anecdote with you.
17:02
This was way back almost eight years ago when I started driving.
17:09
This was probably my fifth or sixth trip, so I was very, very early in my journey of dogfooding.
17:15
So it was early morning, I think it was a Saturday morning maybe about 7:00 a. m.
17:19
I was picking someone up, it was a lady, I stopped my car on the curbside.
17:25
She got in, she was going to the airport, she was going to San Jose airport and she had a suitcase, but she left on the curb and she's inside the car and I'm like, "What's going on?"
17:37
And it suddenly dawned on me that she's expecting me to get out of my car, grab her suitcase, open the trunk, and put the suitcase in the car, but she didn't say anything.
17:47
And then I suddenly realized that there is an expectation of me as the driver doing that part of the job as well.
17:54
And since I was very early in my journey, I had not internalized that and the kind of emotions that ran through me at that time just explained to me, "Hey, what does a driver actually go through?"
18:05
That was a somber ride for me.
18:05
That was a somber day for me.
18:10
And since that day, I have not forgotten what it feels like to be a driver and I carry that emotion with me and that has built a level of empathy in my mind for drivers.
18:15
And I want to make sure that my team understands that's their life and everything we build for them has to have that level of empathy.
18:27
And I would say that was a big surprise for me.
18:27
It's not just about the app.
18:32
Yes, there are many features in the app that we can talk about which are also surprising, but what does a driver feel day in and day out as they interact with other human beings was a big one. I love that.
18:43
I want to talk about how you actually disseminate this culture within the product team, but out of curiosity, I always wonder what pisses off a driver as a rider?
18:48
For example, do drivers like short rides or do they like long rides? Does it not matter?
18:58
I think the short or long rides can be very personal.
18:58
There are some drivers and the great thing about our driver base is there's a lot of heterogeneity in terms of their preferences.
19:03
Some of them do like shot trips because it keeps them closer to their home, they can make a decision to go back whenever they want.
19:14
Sometimes we have incentives which if they do 20 trips, then they get something more. They're able to do that.
19:20
And some of them just like a long trip because they don't want to do multiple pickups and drop-offs.
19:25
So I think there's a lot of heterogeneity.
19:29
If you want a tip for what not to do to annoy a driver, I'll give you a couple of tips.
19:29
It's totally okay for you to be on the phone. We have to do that.
19:35
Just ask them for permission to say, "Are you okay if I'm on the phone?"
19:41
They will always say yes, but it will just annoy them a lot less.
19:47
So that's number one thing that I've learned.
19:47
The number two, on your way out do not slam the door. It is their car.
19:52
It is their final moment of interaction with you.
19:52
Close the door and do it softly.
20:01
That will ensure that you actually get good ratings from the driver. These are great tips.
20:06
I know you can check your actual rating now on the Uber app as a passenger and so it's not 5.
20:09
0 for most people including myself, so this is good to know what might've caused that. Oh, man. Okay.
20:15
One other just that I want to follow here is you spending so much time on the experience implies that that is a huge lever for growth and retention and long-term success.
20:28
A lot of people in growth and product don't see it that way.
20:35
They focus on what's a metric we can drive, what's a conversion opportunity here?
20:39
They're not just like, how do we just make this experience better?
20:39
Can you just talk about that insight of just why you believe just making that experience better so important?
20:50
I think it's not an or, it's an and.
20:50
I do think that obsessing over your metrics, especially when it's funnel metrics and understanding all the points of friction, numbers do tell the story.
21:01
So I think you do need to have the experience points whether they are observed anecdotally or they come through some other means as well as understanding the points of friction because of drop off in your conversion funnel, they're very important.
21:13
But you can't just have that. Because if...
21:18
I mean, you're talking about end users, if it becomes just a mathematical exercise, I don't think that is enough, because as I mentioned, what they're actually feeling in the moment is something that numbers will never show.
21:29
And numbers will not bring the level of intensity that emotions will bring.
21:35
Now the question is, do you have a culture that allows coexistence of both?
21:41
Does it allow a coexistence of you bringing emotion as well as the numbers?
21:47
And that's what I am trying to create every day.
21:47
Our leadership team is trying to create every day, the product team, the engineering team is trying to create every day.
21:58
I think we have done a good job of having a coexistence of those two in the company. Okay.
22:04
And then, so how do you operationalize this and make this a part of the culture at the company?
22:09
Clearly, part of it is modeling it and just doing a lot of it yourself.
22:09
What else do you do to help product people build this empathy and dogfood as much as you do or just some extent? I mean I do it. Tara does it.
22:20
So there is definitely just doing by example or teaching by example.
22:27
We also organize around it, because I think creating a little bit of an organization around it just helps.
22:33
We are a large company now and we try to assist with this by centrally organizing.
22:41
This means let's make sure that everyone understands what it means to be an Uber driver, how to sign up for an Uber driver.
22:47
Some people may be on visas and so on.
22:47
So there is a regulatory part of that.
22:53
So we make sure that everyone understands how to sign up.
22:53
Then we also organize, every year we organize, every quarter we organize one week of driving and delivery.
23:04
And we have a little bit of a competition.
23:04
So we have hundreds of employees who go out and drive and deliver and there's a competition of who can do the most number of trips and who can bring in the most amount of feedback in terms of things to improve and we have prices and so on.
23:15
And that's just taken off over the last, I would say especially the last maybe three years post-COVID because during COVID I think it was a little bit harder for us to gather everyone together.
23:31
And then the last thing I'll say is, what do you do with all of that feedback?
23:31
Because that's when the "boring" part starts.
23:37
Now you're gathering all these documents, you're taking all those issues, you're putting them in your Jira tracking or whatever tracker you may be using and you're fixing them.
23:46
So what we have done is you have set a six monthly OKR.
23:46
So we have set an OKR that we are going to fix 300 of these issues.
23:52
Every team actually, the driver team will fix 300 of these issues.
23:57
The rider team will fix 300 of these issues, and same for the Uber Eats team.
24:02
Now, I don't really care whether it's 300 or 200 or 500.
24:08
The number actually doesn't matter, but the fact that there is a number is what matters, and we have been hitting those targets every single half over the last two to three years.
24:17
So these are some of the things that we do to operationalize it and make sure that people understand that this behavior is not just expected but it's also rewarded.
24:26
I imagine what's challenging about this is there's also work to do that will clearly drive growth, will clearly cut down support tickets, will help with infrastructure costs.
24:31
So these OKRs, there's "Just fix these things that we know suck.
24:39
We don't know what good they'll do, but we know they're not good."
24:42
How do you balance that in resource and prioritization versus all these other things on the roadmap that you know are going to drive a metric?
24:51
We have a pretty good balance set of OKRs.
24:51
So let's say one of our teams has six OKRs just illustratively over the next six months.
24:59
One of them would be ...
24:59
We call them fix-its, so we sign up for 300 fix-its for the next six months.
25:04
In addition to that, we do sign up for a growth goal.
25:10
We sign up for a retention goal.
25:10
We sign up for a cost savings goal.
25:10
And each of them have projects behind them and the executional rigor is a pretty big part of Uber.
25:17
So for all of those projects, we actually have, and I get engaged in that as well, we have pretty serious executional rigor.
25:31
Now, OKRs are not meant to be hit hundred percent.
25:31
I think we celebrate if we hit about 75 to 80% because you do want to set stretch targets, and then every half we do planning.
25:38
And we will allocate resources between all of those OKRs.
25:44
So fix-its, for example, don't get all the resources, but they also don't get zero resources, which may have been the case if we had not set a culture which celebrated this and allocated resources towards this and rewarded this. Awesome. Okay.
26:02
Clearly, an important ingredient here is the CO and C-suite on down values of this as something we will invest in and we'll trade off all focus on growth for this because we believe this is going to help us grow and win long term. Absolutely, absolutely.
26:22
Some people hearing this just as a final question along this area are going to feel like, "Okay, Uber so easy to dogfood.
26:27
Of course, I order Uber Eats every day. I'll go and ride. So good. This is what a dream.
26:32
I'm dogfooding like crazy."
26:32
But a lot of people work at Oracle and Snowflake and I don't know, Fivetran, all these companies where it's much less, non-consumer product essentially where it's more difficult to dogfood.
26:46
Just any advice for folks that want to create a culture of dogfooding and up their dogfood game at a product that isn't consumering.
26:57
So first of all, we thought the same when we thought about the driver app as well because it was a little bit distant for people.
27:02
Now, I would say while it is more difficult than just ordering food and taking a ride, it is not as difficult as perhaps some of the other products that you mentioned.
27:13
But I do think that you just stretch that even further.
27:13
We do have a product we call Uber for Business where it's a business admin using it.
27:18
We do have our merchant product.
27:25
So Uber Eats is actually three sides of the marketplace.
27:25
The eater, the consumer, the courier who's bringing your food and there's a restaurant.
27:31
Now, how do I dogfood the restaurant product, for example.
27:36
So these questions are very, very valid.
27:36
So my guidance would be, and what I try to do is try to get as close to that customer as possible.
27:42
So in the case of a merchant, we go spend time at restaurants.
27:50
I'll give you a latest one.
27:50
So we have a cafe here at Uber on campus. We also have a barista.
27:59
So we enable the barista to be an Uber Eats restaurant, and they have an Uber Eats tablet and I can actually order from that barista.
28:07
So now you're dogfooding being an eater, you're dogfooding being a courier and you can dogfood being a merchant.
28:13
So some of my product team, they actually will go to the barista, they'll stand behind the counter and they'll experience what it feels like for these orders to come in.
28:24
We have very close relationships with our big customers such as McDonald's or Domino's and all of them.
28:29
We also have a lot of small-medium business restaurants and we have relationships with them which allows us to go and spend days with them so that we can observe what life is like behind the counter for a restaurant worker, for a kitchen worker, how does our product show up, and we learn a lot of stuff, Lenny.
28:49
And I would ask everyone who's listening, irrespective of the product that they're working on, go to the nth degree of how much you can stretch.
29:01
And if it's really an IT admin at your customer who's using your product, I would do two things.
29:05
A, go to that IT admin and sit next to them and observe them using the product.
29:12
B, set up what we call test accounts and try to pretend to be an IT admin and set up that account.
29:19
I went through our merchant setup flow, which I can just do by setting up a test account and I learned so many things of all the points of friction that we are putting in that flow.
29:30
So I think there is still a lot that you can do even if you do not have a consumer product. Awesome.
29:35
So the advice there is just find a way, get as close as you can.
29:35
You may not be able to ride an Uber, drive an Uber car, get an Uber ride, but there's some way you can get really close.
29:46
You may not be exactly do the thing. Awesome. Okay.
29:46
I'm going to go in a different direction and there's another theme that came up again and again when I was asking people about you and what you're amazing at.
29:55
And a motto came up again and again it was ship, ship, ship. Talk about that.
30:01
Why is this motto, this way of working so important to you?
30:07
Why is it something you try to encourage everyone to always think about?
30:11
I think it's actually a natural extension of what we just talked about.
30:11
If we are going to go dogfood and experience this pain and you're going to document it, what's next? You have to ship this.
30:22
I'm a big believer in you don't ship documents, you don't ship brainstorming meetings, you don't even ship designs in a Figma.
30:29
What you ship is code in your product.
30:29
That is the only thing that actually ultimately has impact on the end user.
30:35
So as I mentioned before, I develop a sense of impatience.
30:41
It's a positive impatience if you will, that, "Okay, I just observed a problem for our riders or for our drivers.
30:47
They have to get a solution to that as soon as possible."
30:52
And for that, you have to ship, ship, ship.
30:57
Now, you could interpret ship, ship, ship to be just mindlessly shipping.
30:57
No, this is not about mindlessly shipping, but once you know that you're working on something that is important, that is solving a problem, it could be a big feature or a small feature, then you have to cut down the cycle time.
31:12
My biggest enemy is the cycle time of we know it is a good thing all the way to our users seeing it, you have to minimize that time.
31:18
So that's why ship, ship, ship has become such a motto for me and we try to do a lot of things internally operationally for us to enable that.
31:30
And I'll say we still are not there.
31:30
In my opinion, I think we can still be faster, we can still cut a lot of that cycle time, but we are on a constant quest to get as good as we can.
31:42
What does this look like in practice other than just reminding people, "Ship, ship, ship. Ship, ship, ship."
31:46
Just repeating this mantra.
31:46
How do you help people ship faster, to your point, not just work all night and all day, staying long hours, just what are some ways to actually operationalize this?
31:59
I think it stems from the cutting down cycle time, which is the time it takes to make decisions.
31:59
I realized that a lot of the time is not...
32:06
I mean, all the time that you spend documenting the requirements or designing the product or actually writing code is time really, really well spent.
32:18
All the time that is spent in between those activities to align, to discuss a decision, to not make up your mind, then have a meeting the next week, "Oh, let's regroup two weeks from now."
32:30
That happens in our companies a lot.
32:36
The question is, how do you cut that down?
32:36
And for that what we do is we have product reviews and as much as possible we try to come out with decisions in those product reviews.
32:43
And you have to operate in a one-way door versus two-way door concept.
32:50
If something is a two-way door, you have to be able to make the call quickly and you want to make a good call, but if you get it wrong, it's okay, we can turn it back.
33:01
If it is a one-way door, if it's a massive launch, sure, it's okay for it to take some deliberation.
33:06
Maybe it even goes through the C-suite and we have a discussion, but my guidance to my team is cut down the decision-making that has to happen or to cut down the time required for the decision-making between those various steps.
33:22
A few things I heard that you do that you didn't mention that I think helped with this, one is that there was a period where you ran daily stand-ups with cross-functional teams during COVID just to keep things moving.
33:31
There's another story I heard where you wrote a 15-page PRD over the weekend to unblock a stall team.
33:36
It feels like these are ...
33:36
look for opportunities to unblock essentially.
33:42
Well, I want that to be an exception rather than the rule, but yes, I want to be able to step in or I want all our leaders to be able to step in when needed.
33:48
I think you have to be hands-on.
33:52
You have to be hands-on when needed.
33:52
Those two examples, we were coming out of COVID, we had a massive shortage of drivers on our platform and we were talking about growth earlier and we had clear goals on how we grow the number of drivers on our platform all the way from getting new drivers in and improving the funnel as well as regaining existing drivers.
34:13
And Lenny, there was no shortage of the number of ideas, but again, there was a problem with cycle time.
34:19
So I said, "Great, our cycle time is going to be 24 hours."
34:24
And so I think there was a period of about six months when we did do daily stand-ups and they were extremely helpful to unblock the teams on making certain decisions and we shipped a lot.
34:36
And I think it took us about nine to 12 months, but we were able to come out the other side having a great number of drivers.
34:46
And the great thing is that what we did in our driver and courier funnel at that time is still paying off because the number of drivers and couriers continues to grow and now we are worldwide at about an eight million number.
34:58
And we were probably close to half of that when we were just out of COVID.
35:04
So that's an example of where daily standards help, I try to minimize that.
35:09
The PRD exam, yes, I did write a PRD. It was a 15-page PRD.
35:09
We were working on a pretty big concept and it was a risky concept.
35:15
And it was driver facing and that's always high stakes.
35:23
And by this time, we have had about 20 conversations and brainstorm meetings and documents that had been written with the best intentions, but thing that was not becoming clear is, hey, what should we actually ship?
35:36
Because there were a lot of good ideas and not so good ideas, but no one was actually putting it all together to say, "This is what we're going to ship."
35:47
So it was one Sunday night and I literally, it took me two hours and I just wrote down, I'm like, "Guys, this is my stake in the ground.
35:53
We can completely change it, but this is the product we will ship. Agree or disagree."
35:57
Now, the product that we eventually shipped was at least 50% different than what I had written down, but it just catalyzed us getting to that endpoint.
36:03
Now, if I were to do that every single day or every single week, then I would say there's a problem.
36:14
Yeah, you wouldn't have time to deliver food and give rides. That is exactly right.
36:18
Or I would not have time to do hiring or I would not have time to do performance reviews and so on.
36:22
But once in a while, I think it's totally fine for product leaders to step in, especially if you're trying to remove ambiguity and you're trying to just speed decisions up.
36:32
It's okay to step in and be hands-on.
36:37
Another tactic that I heard that I want to ask you about, someone anonymously said that you ask for live demos.
36:43
And here's the question, he always insist on doing live demos that make everyone nervous when they present. The question is, why?
36:53
So I think they may be talking about when we do product announcements.
36:53
So when we do product announcements, externally, on stage, et cetera, I do insist on actually live demoing the product.
37:09
Part of this comes from, I grew up as a product manager back in the day in the 2000s at Palm.
37:16
Palm where the Palm Pilot.
37:16
Palm was cut from the Apple clock.
37:16
There were a lot of ex-Apple people there and we were in the consumer electronics business.
37:22
And every single year the way you would announce your product is you would get up on stage and you show the product working.
37:28
And leading up to that moment was a lot of massive amount of work to make sure that it actually works on stage.
37:40
Now, I will say there's an internal reason and an external reason to do this.
37:44
The external reason to do this is you have to tell a story.
37:44
One of my other beliefs is, which I may not have shared before, is no one actually really cares about the feature that you want to talk about.
37:59
People are way too busy.
37:59
You are obsessed about your products, you, your products are your life, but for your end users, your products are maybe 10 minutes of their life once a week.
38:11
So their life does not center around your products. They don't care.
38:18
How do you make them care?
38:18
You have to tell a story and that story comes from actually showing the product in action rather than a lot of slideware and pictures.
38:24
So that is my belief that I don't want to have hubris about my own product.
38:29
I need to be able to tell a story about how an end user will actually get value out of that.
38:34
That's my external reason for why I insist on live demos.
38:40
My internal reasons for why I insist on a live demo is while it is a lot of work, I think it creates a certain level of rigor ahead of time before the product is ready to launch.
38:55
Because if it is going up on stage, it has to work.
38:55
And two, you may ask the same person who asked you this.
39:02
It creates such a sense of pride amongst the product team and the engineering team to actually see their product live on stage and now it's on video and it'll stay forever, to see the product that they put their blood and tears into actually working.
39:19
I think that's a huge moment of pride and I take it as personal responsibility to make sure that I can help bring that moment.
39:23
But yes, leading up to that is a lot of work. A lot of stress.
39:28
Okay, that was a great answer.
39:28
Today's episode is brought to you by Coda.
39:28
I personally use Coda every single day to manage my podcast and also to manage my community.
39:35
It's where I put the questions that I plan to ask every guest that's coming on the podcast.
39:41
It's where I put my community resources, it's how I manage my workflows.
39:45
Here's how Coda can help you.
39:50
Imagine starting a project at work and your vision is clear, you know exactly who's doing what and where to find the data that you need to do your part.
39:55
In fact, you don't have to waste time searching for anything because everything your team needs from project trackers and OKRs to documents and spreadsheets lives in one tab all in Coda.
40:03
With Coda's collaborative all-in-one workspace, you get the flexibility of docs, the structure of spreadsheets, the power of applications, and the intelligence of AI all in one easy to organize tab.
40:20
Like I mentioned earlier, I use Coda every single day and more than 50,000 teams trust Coda to keep them more aligned and focused.
40:26
If you're a startup team looking to increase alignment and agility, Coda can help you move from planning to execution in record time.
40:30
To try it for yourself, go to coda.
40:35
io/lennytoday and get six months free of the team plan for startups. That's coda.
40:41
io/lenny to get started for free and get six months of the team plan. Coda. io/lenny.
40:49
I'm going to shift our conversation a bit to general career advice.
40:49
I've got a bunch of questions.
40:54
I actually asked the community, my newsletter, Slack community, what to ask you, and they had a few really good questions that I want to run by you.
40:59
So one is just, what are some of the most impactful and recurring pieces of advice that you share with early career product people that you find most helps them get unstuck and most helps them move to where they want to move in their career?
41:19
One is go to a job where you can ship multiple products as fast as possible.
41:19
Comes back to cycle time where you can ...
41:31
Let's say over a period of two to three years, you can go through three to five product cycles as fast as possible because you want to put that under your belt.
41:37
Don't worry too much about how you show up in a room.
41:44
Don't try to spend too much time on being the smartest guy in the room.
41:49
Just define the product, work with your engineering, with your design team, with your data team and ship that product.
41:54
Rev that product, iterate on that product, ship the next product.
41:59
What you will find as you do that over those two to three years, three years later, you'll suddenly realize that, "Wow, I've actually shipped a lot."
42:10
And the other thing that I strongly believe is what makes a great product manager is not the five amazing strategic ideas you had over those three years.
42:16
It's the thousand micro decisions that you made, not the five macro decisions that you made, but the thousand micro decisions that you made.
42:29
Where should I put the button?
42:29
Should I put the screen there or not?
42:29
What should the copy say?
42:34
Do I launch this product in Atlanta or in New Jersey?
42:34
You go through these and what you want at the end of that is you'll start developing an innate sense of judgment.
42:41
So you want to increase your cycle time to be able to get to judgment as fast as possible, because when you make a similar decision the 10th time, you don't even have to think about it. It just happens.
42:58
And we constantly talk about experimentation versus gut or data versus gut.
42:58
As you progress in your career, you want to develop that gut as fast as possible, because in my opinion, that's what differentiates a good product manager from a great product manager, that they have developed some extra level of gut and they're able to make the correct decisions because of the experience that they have had over that time.
43:23
Another way to label what you're describing as product sense, which is something a lot of people are always told they should be building and are very unsure how to go about building it.
43:32
And what I love about your advice here is this, is you find a place where you could just ship over and over and over and that keep that cycle time down.
43:36
And there's almost like implicit advice here is if you're working at a company that takes a long time to ship, this is probably not a good place to be starting your career. That's right. That's right.
43:51
Another question from the community, this is from James Conway. He asked that.
43:51
Okay, so you've been in product for about 25 years, something like that at this point, right? That's right. 24 years. 24 years. 24, okay. Almost 25, almost there.
44:04
Oh man, that's a long time.
44:04
Okay, so the question is, what's remained constant throughout your career in product and do you think that constant will hold even as AI starts to change the way that we build product and build teams?
44:21
I think there are many things AI is changing and AI will change even if ...
44:21
What did I work on 24 years ago or how did I do product management 24 years ago is different than product management today.
44:35
The way we write PRDs, the way you do designs, the way you make decisions, the fact that you have access to data today didn't really exist 24 years ago.
44:40
The only data that you had was surveys and those surveys took six months to run and for data to come back.
44:50
And now we have instrumentation and telemetry and everything that we ship and we can analyze data.
44:55
So I would say things have definitely evolved.
44:55
With AI, things are just going to change even more rapidly.
45:01
I'm already seeing on a weekly basis we are introducing a new way of doing things and if you look at a typical product cycle all the way from doing market research and user research to writing the first draft of a PRD to doing the first draft of Marks, doing user research and even storytelling that we talked about, how do I actually tell the story about my product?
45:19
AI is making all of that much, much easier.
45:24
And I think that transformation is going to happen very fast.
45:30
What has not changed in those 24 years is you having to understand what your end users want.
45:36
And I know that's kind of the cliché of product management.
45:36
"Yes, of course I have to understand what my end user wants."
45:40
I think it is harder than we think.
45:40
I think it takes more than we typically do to actually really get your finger on the pulse of what your end user feels.
45:52
And I think that that has stayed constant.
45:52
I think that is the holy grail, not just to understand one archetype of an end user.
45:58
At Uber here, not only are we operational in 75 different countries, even in a single country like the US, we have different kinds of end users, whether it's demographics, whether it is socioeconomic status, and for us to internalize what do end users want, AI will definitely speed it up, but for a product manager, it is going to stay constant.
46:24
The one thing I do think product managers need to be aware of is the importance of knowledge is going to stay, it's going to continue, but the importance of judgment and as you mentioned product sense and the gut feel that is going to become even more important in the world of AI.
46:44
As a result of that, we need to increase that or reduce that cycle time even more so we can develop that product sense and judgment.
46:55
Sort of along those lines, there's a question I got from, I don't know if he's the CTO of Uber, CTO of part of Uber, Praveen who I believe you work with at Uber. Yeah. Okay.
47:05
So the question here was, ask him what he hates from engineers.
47:08
Does that ring any sort of bells?
47:12
I think Lenny, that's a trick question.
47:12
He wants to get me into trouble. I'm sure.
47:17
Let's get you into trouble. Let's do it.
47:22
I'll tell you, I actually know the answer to this question.
47:22
What I hate about engineers is when you on a project and they say, "Hey, I can't do anything because I do not have a PRD from a product manager."
47:35
Now, while I consider that to be important for my job security and a PM's job security, for a PRD to be required, I also think that the best engineers are the ones who actually don't need to start with a PRD.
47:49
They can have a whiteboarding session with a PM, with a designer, they should have just an equal seat at the table in terms of what we are going to build and then they co-create.
48:01
They should co-create and they should write a document together and they should just build.
48:05
One of the things that some people like about me and some people probably hate about me is I'm very boundary-less from an art perspective in how I operate.
48:16
So if someone has a product opinion, it doesn't matter whether you're in the product org or the ops team or customer support or engineering or design.
48:20
I think everyone's opinion matters and is valid and I welcome that.
48:27
I welcome their participation.
48:27
I don't care about org boundaries.
48:32
Now, at the same, time, I will have an opinion about how we message our products.
48:38
I will have an opinion about maybe some engineering architecture to the extent that I understand that or I will have an opinion about an ad campaign or how we operationalize our product.
48:46
So in that spirit, what I hate is if we are blocked because a certain document was not written to the nth degree of specificity.
48:58
I love that you answered the question. That's great.
48:58
And I think what's also interesting along those lines is it's a lot easier now to write a good version of a PRD/one pager with AI tooling where you just like, "Here's what I want to build, help me craft it into something very specific that we can use as a product requirements document."
49:17
And you have so many tools now.
49:17
We do a lot of brainstorming, we do whiteboarding.
49:21
You could actually convert summaries of those conversations into the first draft of your PRD.
49:25
So there are many tools available now.
49:25
So I don't think PMs or engineers have any excuse to be blocked because of lack of PRDs. All right, amazing.
49:35
I want to ask a few Uber questions.
49:35
I think there's a few that a lot of people think about when they think of Uber today.
49:39
The first is, it feels like there's just this big elephant in the room of what does Uber become as self-driving emerges, we see all these self-driving cars around Uber's historically been drivers, people driving cars.
49:50
What is the vision of where Uber goes in a world where self-driving becomes more and more of a thing?
50:00
Yeah, it's definitely a big area of conversation internally, externally.
50:04
So absolutely agree with you on the importance of it.
50:04
If you rewind back the clock, we actually had our own self-driving unit, and during COVID, we decided to divest off it. I remember.
50:16
And we decided that we're going to take a partner-oriented approach and that is the approach that we have taken since then.
50:21
Over the last year, we have signed about 15 partners.
50:21
In fact, we just announced another partner earlier this afternoon called May Mobility.
50:28
So we have been announcing many, many partners when it comes to AVs.
50:34
Now, that's our strategy.
50:34
Now, what does that actually mean?
50:40
What do these partners do?
50:40
What we are building is we are building a hybrid network.
50:45
The hybrid network has a hybrid of human drivers and AVs on the same network.
50:53
And we work with partners.
50:53
We have a set of APIs that we have built.
50:53
It's a very rich set of APIs and we have partners who basically integrate with those APIs and get access to this hybrid network, this hybrid marketplace.
51:05
And let me explain why that is beneficial.
51:10
These cars are expensive.
51:10
These are assets that need to be utilized.
51:10
And if let's say a city has 500 Avs. When it is 10:30 a. m.
51:16
on a Tuesday mid-morning, we know that demand is going to be really, really low and you may have a lot of these vehicles actually sit underutilized.
51:24
Whereas when it is Friday evening at 6:00 p. m.
51:31
, you don't have enough cars to fulfill the demand that you will have at that point.
51:36
So if any of these partners try to operate that service themselves, it can definitely work.
51:41
But you're going to leave either customers dissatisfied or you're going to have underutilization.
51:46
So our hybrid network really allows them to equalize that because when you don't have enough cars, we are able to send drivers to fulfill that demand.
51:57
And then given that demand density that Uber brings, we are also able to utilize all the vehicles even if it is 10:30 a. m. on a Tuesday.
52:02
So that is the value proposition of our network.
52:09
We are big believers in autonomy.
52:09
We think autonomy is going to lead to more safety.
52:09
It is going to take some time.
52:14
So as we project out, we don't know if it is going to be five years or 10 years until it sees a certain kind of scale across the world, across many, many cities.
52:26
We are in about 75 countries today, roughly about 15,000 cities.
52:26
So there's a huge geographical landscape for us to capture as well. But that's our strategy.
52:34
And so far, we recently launched in Austin with Waymo and that's been going really well.
52:44
And we of course have Phoenix with Waymo.
52:44
We are going to be launching Atlanta soon.
52:44
We've also launched internationally in Abu Dhabi with our partner WeRide.
52:51
And then there are many more launches that are going to come down the road. So very excited.
52:57
And by the way, I talked about mobility.
53:01
Delivery has something similar.
53:01
So we have sidewalk robots with many partners who in a downtown area for a short distance delivery are actually doing thousands of deliveries today. Oh wow. I've seen those robots. They're very cute.
53:15
I always wonder how they get to where they're going, but clearly it's working.
53:20
The Waymo thing, I've totally noticed that Waymo basically in some form has decided to move away from their own app in many cities and Uber is how you get a Waymo, which I don't think a lot of people understand.
53:30
So what I'm hearing is essentially the strategy is just be the super app for getting around and getting food and Uber drivers are a part of that and part of that is autonomous vehicles.
53:41
And is the idea over time it will stay that way?
53:46
There will always a need for humans to drive cars for a long time?
53:51
I will answer that in two ways actually.
53:51
I think for a long, long time there will be a need for humans to drive cars.
53:56
There are pickup and drop-off spots that AVs can't go to today.
54:01
There are going to be regulatory hurdles that we have to get across before AVs are everywhere and so on and so forth.
54:10
Second, we have a mission to make sure that we can provide a great earnings opportunity on the Uber platform.
54:18
Today, that is fulfilled by driving people and delivering food or delivering groceries.
54:25
But that list is going to continue to expand, and we want to make sure that we offer more and more earnings opportunities on our platform.
54:35
Even me going back to dogfooding about, I think about a year ago I also started picking up groceries because now we have a pretty large grocery business.
54:41
And I have gone and I've carried crates of water up a flight of stairs and delivered that to the doorstep.
54:47
And there are more things that we will do as well.
54:53
We have a partnership with Walmart where we may send a courier who may pick up 15 different bags and deliver it to 15 different homes.
55:04
So the need for human drivers on our platform or human earners I would say is just going to continue to increase because we're going to have multiple earning opportunities for them over time.
55:15
I love that visual of you carrying up big jugs of water to some random person's house.
55:19
I think they tipped me well.
55:19
I think that was a good tip.
55:22
By the way, what do you do with the money you earn from this?
55:24
Does you just keep it or it's like it goes somewhere?
55:28
So we have a policy that the money actually is not kept by us.
55:32
We refund it back to the riders or to the eaters. Yeah.
55:42
What do they see when they get this refund? They're just like enjoy.
55:45
Actually that's a good question.
55:45
I don't think that it says that this is because it was delivered by an Uber employee. It's a good question.
55:54
I should go back and see what it says.
55:54
But we do refund it back to them. Yes.
55:57
We're dogfooding right here. Yeah. Okay.
56:00
Another question that I want to touch on is Uber, it feels like Uber was one of the first companies to take profitability really seriously.
56:05
I remember Dara meant tweeting or sharing this insight during COVID I think of just investors and public markets are now valuing cash flow profitability.
56:17
We need to change the way we operate.
56:17
You guys have clearly done that.
56:23
It must've been a pretty hard thing to do for a company that was used to just, famously all these companies raised a bunch of money, scaled like crazy, blitz... What is it?
56:31
Blitz scaled as Reid Hoffman called it and then making a big change like that.
56:37
I imagine it was very hard.
56:37
What was that experience like?
56:37
Just switching cultures from just raise money, go crazy to profitable cash flow business? I think it was awesome.
56:47
I think it was a great, great period, because if you are taking inefficiency out of your system and you're taking the dollars you saved back into the business or back into the end user's pockets or the driver's pockets, that's just a great feeling.
57:04
And as you said, doing that is not easy.
57:04
It's actually very, very hard.
57:09
But that's where real innovation actually kicks in.
57:09
Let me give you an example.
57:14
One of the ways that you would take cost out of the system in the delivery world is to batch trips together.
57:20
So when a courier is delivering from point A to point B, maybe they can deliver two packages.
57:27
They pick both of them up at point A and they deliver them to point B and to point C and point B and point C happened to be close together and there we just delivered a lot of efficiency with the courier's time.
57:37
And now you multiply that by a thousand different ideas that we may have both in the mobility business as well as the delivery business and it's a grind, Lenny.
57:47
There was a grind in terms of, how do we take inefficiency out of something like that, which is our core marketplace logic.
57:53
How do we take inefficiency out of how we do promotions?
57:53
How do we take inefficiencies out of support costs, payment costs, background check costs?
57:59
There are many, many pockets and some of that required building new paradigms and building new software.
58:11
And when you actually save that money and you're able to put that back into the business and become a profitable company so that you don't go under and your stock price goes up, that's actually feels very, very rewarding.
58:20
I think that Dara has done an excellent job driving that.
58:20
I think the company has done an excellent job responding to that.
58:28
I feel that that still goes on today.
58:28
Yes, we became profitable about two years ago and now we are driving great margin quarter over quarter, but the spirit of making sure that our core is solid in terms of efficiency, it continues to feed the system and continues to allow us to then make more bets that may be outside of the core.
58:55
I imagine a lot of companies are going through this same experience now, trying to cut costs, try to be efficient, the government, the US government's going through this, it's very difficult to do well without cutting things that are really dear to people, especially at a product company, big bets that are going to maybe pay off in the future.
59:08
Big innovations, things people are really excited about.
59:14
For people trying to do this at their company, do you have any tips?
59:19
I don't know, one or two tips to do this well to keep morale up, to not lose sight of big opportunities in the future? Anything there?
59:27
Yeah, I think we divested out of two big businesses, which were our self-driving unit as well as what we call Uber Elevate and we divested out of both of them during COVID.
59:32
COVID was a very special time for us, if I could use that term.
59:39
Our mobility business dropped by 80% overnight and I think that Dara and the leadership team actually showed tremendous leadership during that time.
59:51
And one of those was to make those calls to divest out of those businesses.
59:57
Now, I think an important part to remember is you could be on the quest to just drive profitability and you may lose your sight on new growth bets as we call them here and how do you do that.
1:00:05
So I have my framework on this, which is I think of our portfolio as concentric circles.
1:00:13
The center of that concentric, the core is our core product.
1:00:20
We do 33 million trips every single day.
1:00:20
And if you think of a typical Uber trip or an Uber delivery, there is a multi-point failure possibility all the way from you opening the app to a driver, accepting or you doing the pickup, doing the drop-off and you getting a receipt.
1:00:40
And then similarly, there are a number of steps on the Eat side.
1:00:46
And we obsess over making sure that that core is as flawless as possible.
1:00:46
And also to our point about profitability is as efficient as possible as well.
1:00:54
So imagine taking even 1 cent of efficiency out of 33 million trips every day, the numbers just really start to add up.
1:01:00
So you have to make sure that you're paying a lot of attention to that.
1:01:06
As I go out and dogfood, I'm dogfooding the core mostly and making sure that it is flawless in my own little way.
1:01:18
And if you're spending enough attention on that, then you get the license to expand out of that core and start to think about other growth bets.
1:01:23
By the way, this is where Uber Eats came from back in the day.
1:01:28
This is where Uber Groceries come from, which is now a very large business for us.
1:01:34
And on the right side, Uber Reserve, Uber for Teens, which is something I launched a couple of years ago, our focus on taxis and there are many other growth bets that are now in the billions of dollars in total.
1:01:46
That's where it came from because we are making sure that the core is being paid attention to and it's giving us the license to go out off that in that concentric circle.
1:01:58
There's another area of Uber that it feels like is shifting a bit.
1:01:58
So historically, I know a lot of people that have worked at Uber over the years, Uber has this reputation of being very data-driven, very ruthless with prioritization.
1:02:06
A lot of what you've been talking about feels like the opposite of just anecdotes and experiences, empathy.
1:02:13
So let me ask you this question.
1:02:13
What's something that you as a company, as a leadership team chose to do that was not what the data was telling you to do?
1:02:25
Could be a feature, it could be a whole big bet that ended up being successful.
1:02:32
I have a few and each of those could have gotten me fired, Lenny.
1:02:32
It was that sort of gut-based.
1:02:39
I'll give you a few of those examples.
1:02:39
When I joined I was working on safety, and safety is a big issue on the Uber platform.
1:02:44
Was a very, very big issue back in 2017.
1:02:44
We had not built enough software to be able to manage safety in the physical world.
1:02:56
So while safety in terms of incident reduction, look at the number of incidents you're having and then just like a growth hacking project, you just drive that number down needing to do whatever you can.
1:03:07
I said in addition to that, we also need to incorporate a lot of functionality and features and UI in the product, both on the rider side and the driver's side, which is delivering a sentiment of safety and making sure people feel safe. 99.
1:03:25
9% of our trips are totally fine from a safety perspective, but that doesn't mean that people are not concerned.
1:03:31
So in addition to driving down the incidents, we also have to develop a bunch of features so that they feel safe.
1:03:36
And that actually helped a lot because we started hearing from our users that, oh, we measure something called safety sentiment and that safety sentiment has been going up a lot.
1:03:47
So I would say that was back in the day that I drove.
1:03:54
Another crazy one was I pushed the company to work with taxis.
1:03:54
If you look at my resume, I worked at a company called Flywheel, which is a taxi app company.
1:04:02
And so I spent three years there and that's what got me to transportation or being interested in transportation and eventually ended up at Uber.
1:04:11
And Uber and taxis have always had an issue with each other. We all know the history.
1:04:18
And I was saying, "No, we should actually launch with taxis."
1:04:18
Now, if you look at the data, the data will tell you that the number of taxis is going down, that the taxis are not very reliable, the taxis have antiquated software in them, the taxi drivers may not want to provide the same level of service, so on and so forth.
1:04:35
And there's a little bit of a gut feel and I said, "No, look, there are drivers who want to make money who do not always have enough demand. We have a lot of demand.
1:04:47
If we put the right incentives in place, we will be able to create a structure where we can work together productively."
1:04:53
And today in the US I think we operate taxis in about 10 to 15 different cities.
1:04:58
In New York City, by the way, every single yellow cab is hailable through the Uber app and that is generating a pretty large amount of revenue for us and we have taxis all over the world as well.
1:05:10
So that was probably another one where the data just didn't support my gut decision.
1:05:15
And the last one I would say is Uber for Teenagers.
1:05:20
There are so many reasons to not do that, by the way.
1:05:20
You go talk to parents, they're like, "I am never putting my teen in an Uber."
1:05:25
And if you talk to the risk people in the company, they're like, "We are not going to put teenagers in our cars and there's a lot of liability."
1:05:35
I happen to be the father of a 16-year-old and a 12-year-old.
1:05:35
And I can tell you that one of the biggest problems in our household is kids' transportation.
1:05:40
And I'm like, "How can that not be a problem?
1:05:48
And if anyone is going to solve a transportation problem, shouldn't that be Uber?"
1:05:54
And so how can we build a product?
1:05:54
So I started with let's just build a product that is going to solve every single one of these problems that parents have in terms of safety.
1:06:04
So we launched that two years ago.
1:06:04
It's been growing really, really well.
1:06:04
And in fact in California for several reasons, we had to turn the product off and I get emails every single day from parents in California about why did we have to turn that product off.
1:06:15
So I would say a lot of the growth bets that we actually make may not have the best data supporting them, but as long as we understand intrinsically what the end user wants and the problems that we need to fix, I think you have to go with your gut.
1:06:33
And I would say when it comes to transportation, we still have not solved a lot of problems that people have and so that's all headroom for us. Very cool.
1:06:44
On the taxi's point, I've always thought taxis are great in so many ways.
1:06:44
They just know exactly how to get places.
1:06:50
They're very knowledgeable and they just get you done.
1:06:50
It's not the most pleasant experience, but they're very good at what they do.
1:06:57
And so I think it makes so much sense to integrate.
1:07:01
And it reminds me also, my wife always jokes.
1:07:01
There's always startups that come in, disrupt the whole industry, like Airbnb is another example and then they end up just providing the same service. Oh, okay, cool. Now we got taxis again. Here we go. Exactly. They're very efficient.
1:07:14
They know the city like the back of their hand. Yeah.
1:07:17
Especially in the UK, their whole knowledge thing. Okay.
1:07:17
I'm going to take us to a couple of recurring segments on this podcast. The first is AI Corner.
1:07:28
Is there some way you've figured out how to use AI in your work, some AI tool to help you work better, to help you work faster, to get better work done?
1:07:36
I would say the team is using a lot of tools now.
1:07:36
The team is using tools to build initial prototypes, initial mocks I should say.
1:07:42
The team's using that for user research to be able to process information as well.
1:07:52
What are some of the tools, if you can name them?
1:07:55
We are using both ChatGPT as well as Gemini quite heavily in the company.
1:07:55
We have close relationships with both the companies.
1:08:01
Specifically I end up using ChatGPT a lot for summarization of documents. That's one thing.
1:08:07
All of us have too much to read and we don't have enough time.
1:08:12
I've not used NotebookLM yet.
1:08:12
I know a lot of people who have started using it and that is the next thing that I'm going to use just to be able to build an audio podcast based on a bunch of information that you can consume. I think that's awesome.
1:08:28
What I've been doing is I have been using ChatGPT and Gemini to summarize long user research documents.
1:08:34
Some of these reports, they're 50 to 100 pages long.
1:08:34
I will never have the time to read them.
1:08:39
That's really making me much, much better acquainted with, this is what writers in South Africa feel and this is what's happening in Brazil, this is what's happening in Korea.
1:08:49
So that has been a huge enabler.
1:08:49
The second thing that I've really, really started liking is the deep research feature.
1:08:55
The deep research feature in, again, both Gemini as well as ChatGPT, where you feed them a prompt, you actually present a pretty hard ...
1:09:02
The other day we were thinking of a driver feature.
1:09:08
It's a driver feature that has been discussed a lot in the past. We haven't done it.
1:09:13
I just asked ChatGPT deep research, here's the feature, this is how it would work.
1:09:20
Can you let me know if drivers will like it or not like it?
1:09:20
Is there anything we can change about that feature for the drivers to like it?
1:09:25
And it presented me an answer, which I would say it's an amazing research assistant and it's absolutely a starting point for a brainstorm with my team with some really, really good ideas.
1:09:38
So I think that that is making us not just more productive.
1:09:42
I think it's actually going to make us better at our jobs.
1:09:48
That is an incredible use case.
1:09:48
Essentially it's a thought partner on strategy and helping you find gaps in your thinking.
1:09:53
And I think for Uber, it's easier because these crawlers can learn a lot about what Uber is.
1:10:00
For some products, they won't know as much, but clearly, there's something you could do there, whatever your product is.
1:10:05
And by the way, it's cool that you guys give access to ChatGPT.
1:10:08
A lot of companies don't.
1:10:08
Look, a lot of companies only have Gemini or something like that.
1:10:12
Okay, I'm going to take us to another corner on the podcast. I call Fail Corner.
1:10:18
And what Fail Corner is about is people come on this podcast, they share all these wins, cool taxis worked and teens worked and safety worked.
1:10:23
And look at this, we got profitable, everything's just going great all the time.
1:10:28
But in reality, that's not how it goes.
1:10:32
Is there a story you could share where something went wrong, something didn't work out the way you expected in your career and what you learned from that?
1:10:40
I still somehow, when I think about it, I don't end up calling it a fail, but absolutely you could call it a fail because the product didn't work or the business didn't work and so on.
1:10:51
I was at Palm a long time ago.
1:10:51
Palm developed a Palm Pilot and then it went into the phone business.
1:10:59
Some of your listeners may have used them, some of your listeners may be too young to remember.
1:11:03
And then there was a period of time when it was really Blackberry and Palm or the two smartphones on the market, especially here in the US.
1:11:10
And then Apple announced the iPhone and then soon thereafter, Android was announced by Google.
1:11:16
And it took us a long time to respond to that.
1:11:23
It took us maybe three, four years to respond to that.
1:11:23
And in the end, HP bought Palm.
1:11:32
We did not have the level of scale or the level of hardware quality that we needed to have to be able to effectively compete with the likes of Apple.
1:11:38
When you are going through the motions, when you are actually in the moment, you don't always realize what's going to happen.
1:11:44
You also think that, "Oh, we will be able to recover and we will be able to compete."
1:11:49
But I think there are many, many things that I learned from that and I always keep that very close to my heart.
1:11:56
In consumer markets, if you are player number three or player number four, it's a little hard, because the two top players are going to compete with each other so ferociously especially if they're scale players like Apple and Google, it's difficult for number three and number four to be able to keep up with that unless you have a very, very unique angle.
1:12:20
And the way I use that now is I am, Lenny, always paranoid about the number three or the number four player or the number two player who may have an innovative idea, who may try a different strategy, who may try a different go-to-market.
1:12:38
I just never take the status quo for granted.
1:12:38
So that's sort of one learning.
1:12:38
And I think the second learning is speed and hustle and resilience.
1:12:45
After Palm, I spent time at a couple of startups.
1:12:52
I mentioned Flywheel, which was a taxi company, which is also not able to scale.
1:12:52
But what I learned through that period is that hustle and sense of urgency is extremely important.
1:13:05
So when I think of my portfolio of skill sets right now, really, that's portfolio that I have developed from a lot of failures over my career.
1:13:12
And those are, I would say sense of urgency because you need to move fast because status quo can never be taken for granted.
1:13:19
And second is resilience because times will be hard and things will be difficult and you just have to power through them and you have to keep building for your end users through that and the times will change.
1:13:36
So I would say those are the things that I've learned and those are my failures.
1:13:41
I can see where this motto of ship, ship ship came from. It at all connects.
1:13:45
There's absolutely a causation there.
1:13:48
Sachin, is there anything else that you want to share or leave listeners with before we get to our very exciting lightning round?
1:13:55
I can talk a little bit about how I think about, I think I've talked a lot about end users, but I cannot talk about end users enough.
1:14:00
So in terms of offering my unique take, which may be completely flawed, but that is my unique take, I'll offer this. Intriguing.
1:14:11
Which is I don't think our relationship with our end users is reciprocal, which means when I wake up in the morning, all I'm thinking about is, "How can we make our products better for our end users?"
1:14:22
When end users wake up in the morning, all they're thinking about is, "I have to do this at my job.
1:14:27
I have many complexities, many problems in my life."
1:14:34
And that's what occupies their mind all the time.
1:14:34
They're not thinking about you, they're not thinking about your product.
1:14:39
As much as that may be a hit to your ego.
1:14:45
However, for those, and I'm talking about let's say Uber Rides, for those 15 minutes in the day when they do think about your product, that is your opportunity to absolutely dazzle them, give them an experience that they will not forget or it'll be so seamless for them that they don't have to think about.
1:15:00
So next day when they're again going to work, they just open the Uber app just as muscle memory, but do not expect their life to be centered around your product. Their life is complex.
1:15:12
All of our lives are complex.
1:15:12
You are just trying to be a part of their lives.
1:15:17
You're not trying to take over their lives. You never will. I love this advice.
1:15:22
There's a book that I loved that is for writing and the title is Nobody Wants to Read Your Shit. Yes.
1:15:33
And that applies exactly to products.
1:15:33
Marc Andreessen has this quote where I think it's, "Everyone's time is already allocated.
1:15:39
They don't have time for your product, they have stuff to do in their day."
1:15:42
And to hit that bar is very high.
1:15:42
And so I think that's a really important reminder.
1:15:46
Thank you for sharing that.
1:15:46
With that, we've reached our very exciting lightning round.
1:15:49
I've got five questions for you. Are you ready? I am.
1:15:56
Number one, what are two or three books that you have recommended most to other people?
1:16:02
Maybe a boring answer, but these are the books I believe in.
1:16:02
I recommended the Steve Jobs book by Walter Isaacson.
1:16:08
I have recommended Elon Musk book.
1:16:08
I have read the Ashlee Vance one fully.
1:16:16
I'm in the middle of the Walter Isaacson one and I have recommended Hard Thing About Hard Things by Ben Horowitz.
1:16:23
These are the three books.
1:16:23
I think that there was a period of time in Silicon Valley maybe 10 years ago, and maybe this is still true, but everyone wanted to be Steve Jobs and they thought that they can just recreate being Steve Jobs.
1:16:36
There was this fever in the valley, and the reason why you want to listen or read that book is you need to understand the context of what makes up the entire person.
1:16:48
Yes, the person part is important, but how they apply their own context to the products that they ship and they change the world is important because not that I or anyone else can recreate what they did or what they're doing, but take the learnings from them and see how you can apply the same principles to your context.
1:17:10
You're not going to behave the same way that they did.
1:17:10
You're not going to build products the same way that they did, but you're going to give it your absolute freaking best because that's what your end users deserve and you're not going to compromise on that.
1:17:19
That's sort of what my takeaway is, and I think that takeaway is extremely strong in the books that I talk about.
1:17:28
And the Ben Horowitz book, it goes back to my concept of resilience.
1:17:28
Some people have this really optimistic view of how life is in the corporate world or in the product world or in the CEO office. It's not so.
1:17:40
These are lonely jobs, these are hard things, and how do you persist through that?
1:17:46
I felt that that book was amazing at teaching me that. Amazing. Okay, next question.
1:17:52
Do you have a favorite recent movie or TV show?
1:17:57
My favorite recent TV show is Peaky Blinders.
1:17:57
Peaky Blinders is on Netflix.
1:18:02
It's based in the 1920s and now they're in the 1930s. I'm on season six.
1:18:02
It's an Irish family that's in a town called Birmingham and they go through all these years and they do all sorts of bad things.
1:18:15
But again, it's an amazing story, so I highly recommend watching it.
1:18:19
I started watching it, but I didn't stay with it, so maybe I got to revisit it.
1:18:19
Do you have a favorite product that you have recently discovered that you really like?
1:18:30
It is the product that makes me or evokes that emotion in me right now is deep research from ChatGPT, because I've never had a research assistant and I keep going back to that research assistant and I think it's an excellent product.
1:18:42
I'm sure it will improve further, but that would be my favorite product recently.
1:18:51
Do you have a favorite life motto that you often come back to in work or in life?
1:18:57
Yes, and my kids are sick and tired of hearing this from me now, which is just focus on the inputs.
1:19:00
You have input metrics in life and in work, and you have output metrics.
1:19:07
You can't really control the output metrics.
1:19:07
All you can control is inputs, whether you're trying to lose weight, whether you are trying to get good grades or become very good at high school sports, or you're trying to ship a great product or you're trying to improve retention or churn or whatever, you can only control the inputs or focus on the inputs.
1:19:30
I love that your kids are sick of that.
1:19:30
I wonder what context you use with your kids.
1:19:30
But anyway, final question.
1:19:35
You shared a really good tip for how to not annoy an Uber driver.
1:19:35
What about as a food delivery person, what can a customer of Eats do to avoid pissing off a delivery courier?
1:19:50
If they're delivering food to your home at night, turn the porch light on because they've just driven in the dark.
1:19:58
They don't know your house.
1:19:58
They're trying really, really hard to find your house, and after that, they turn the flashlight on, on their phone and they're trying to find the door where they can place the food. Just turn the light on.
1:20:14
I'm super guilty of forgetting that sometimes, and they also don't want to be shot by someone thinking they're sneaking around.
1:20:16
Sachin, this was incredible. Two final questions.
1:20:21
Where can folks find you online if they want to reach out, follow up on anything, and how can listeners be useful to you?
1:20:28
Send me a message on LinkedIn.
1:20:28
I will say that I've been guilty of not establishing an online presence and writing.
1:20:34
Life and work have just taken over.
1:20:40
I need to do a better job.
1:20:40
Maybe your blog will or your podcast will really help you with that. Here we go.
1:20:46
But yeah, hit me up on LinkedIn.
1:20:46
That's probably the best place.
1:20:50
And how can listeners be useful to you?
1:20:52
Well, back to my dogfooding theme. Send me feedback. Send me feedback.
1:20:52
Hopefully everyone is an Uber user, whether it's Rides or whether it's Eats.
1:20:58
Send me feedback of how we can get better as a product, and the more detailed your feedback, the better it is.
1:21:04
It's very easy to say this doesn't work or that doesn't work, but if you actually describe and say, "Hey, this is what happened.
1:21:15
I was doing this and then this happened. This didn't make sense.
1:21:19
This is how you can change it."
1:21:19
I think that'll be extremely valuable to me.
1:21:22
Only 40-page reports accepted. Just kidding.
1:21:22
Sachin, thank you so much for being here. Thank you, Lenny. Thank you for having me. Bye everyone.
1:21:30
Thank you so much for listening.
1:21:30
If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app.
1:21:36
Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast.
1:21:42
You can find all past episodes or learn more about the show at lennyspodcast. com.
1:21:47
See you in the next episode.