The Fast Lane to Learning in Ops | Kristin Kaiser Mulligan, Heard

Ken Babcock (00:01.142)

Hey everyone, welcome to the Change Enablers podcast. I am so pumped today to have Kristin Kaiser Mulligan, my friend, my former Uber colleague. And since Uber, her career has just gone on like a crazy trajectory, so we're gonna talk about that too. But Kristin, I'm stoked you're here.

Ken Babcock (00:25.41)

Can you not hear anything?

Kristin Kaiser Mulligan (00:26.821)

Sorry, hold on a minute. It says uploading 99%. I lost you. I froze. Sorry, you have to do it again. I can hear you now. Sorry, it was like... Now I can see that it's working now. I didn't see anything. It went pixelated. You go.

Ken Babcock (00:32.47)

Oh, bummer. That's all right, you can hear me now though, right? All right. You just saw facial expressions?

Ken Babcock (00:45.678)

All right, all right. Welcome everyone to the Change Enablers podcast. I am so stoked to have Kristin Kaiser Mulligan here today. My friend, my former Uber colleague, since we both left Uber, Kristin's career has just gone on like a crazy trajectory, so we'll talk about that. But this is gonna be a fun one. This is like catching up over coffee or beer. So we're gonna talk about a ton of stuff, the crux of which is...

when to centralize, when to decentralize, when that works, when that doesn't work. So Kristen, I am pumped to welcome you. So welcome.

Kristin Kaiser Mulligan (01:22.873)

Thank you, Jen. I'm so excited to be here. Thanks for thinking of me.

Ken Babcock (01:26.83)

Of course. We always start with some rapid fire questions. And so I'm just gonna jump in. What are three software products you can't live without right now?

Kristin Kaiser Mulligan (01:39.449)

Okay, this is going to sound extremely lame, but my three, where I am now in this very remote world with my company being fully remote too, G Suite, Slack, Notion, that's where we live and breathe. And so I don't think I could live without those.

Ken Babcock (01:55.918)

Yeah, I mean, it's lame for a reason, right? It's like a no brainer. I mean, that's where Tango lives and breathes too. All right, what's the craziest ops, tasks, projects, playbook, I mean, you name it, that you've had to tackle in your career?

Kristin Kaiser Mulligan (02:17.409)

I love this question because I feel like especially like both of us being at Uber, we've just seen and been through a lot. After I left Uber, I went to Lime from 2018 to 2021. So we went through a ton of different things there. One of the most memorable projects that I worked on at Lime was our launch into Germany in 2019. It was a very huge project. The country, it was the first country to come out with.

nationwide scooter micromobility regulation. And there was just a race to who could get on the ground first. And so this was a collaboration across every single team at the company. And it required us to retrofit 10,000 scooters from a warehouse in Rotterdam, ship them to 20 cities in Germany, have warehouses and teams on the ground there to receive those shipments, have the app ready when it comes to, you know,

There's parking, there's geo-fencing, and we only had one month to do it, and it was just like an incredible experience. I have never spent so much time in my life in a warehouse or in Rotterdam, but I can say I know how to fix a throttle on a scooter pretty well and pretty fast, pretty efficient at it, as well as countless amounts of stickers put on 10,000 scooters, but it was incredible experience from like a collaboration and just get things done standpoint.

Ken Babcock (03:44.942)

And did like, did the Germans embrace scooting?

Kristin Kaiser Mulligan (03:50.093)

They embraced it so much. Germany that summer and for years after was one of Lime's largest markets, largest countries. They just have incredible infrastructure. Really all of Europe has incredible infrastructure for micro mobility. They have awesome city centers. They have a lot of bike lanes. And so Germany embraced us in a massive way, which was awesome. And we had incredible local teams on the ground too to help echo our message.

Ken Babcock (04:18.446)

Do you remember the Lagos, Nigeria story from Uber?

Kristin Kaiser Mulligan (04:22.549)

No, I don't.

Ken Babcock (04:23.634)

Oh my goodness. So the reason I asked whether like this rush to Germany worked was, um, I feel like really early on, uh, we were getting all this signal that people were opening the Uber app and Legos because you know, what turned out, uh, when we didn't get the best GPS ping, it would just default to zero zero. And so zero zero, the closest city to zero zero, like zero latitude, zero longitude was Legos.

Kristin Kaiser Mulligan (04:44.853)

Yes. OK.

Ken Babcock (04:53.554)

And it's a huge city in Nigeria, huge, but it was like millions and millions of eyeballs. So we thought, oh my God, there's all this pent up demand. And so we like raced.

Kristin Kaiser Mulligan (04:55.001)

Mm-hmm.

Kristin Kaiser Mulligan (05:02.441)

I do remember that actually.

Ken Babcock (05:05.162)

We like raced to go there and everyone was like, yeah, no, we don't take taxis here. So anyway, I'm glad to hear.

Kristin Kaiser Mulligan (05:13.22)

Like a classic well to do engineer that just said, oh, anytime that we get this, just set it to zero.

Ken Babcock (05:22.078)

Yeah, exactly exactly. It was probably good in the moment Speaking of geographies our last rapid-fire question Where would you go in a zombie apocalypse?

Kristin Kaiser Mulligan (05:34.593)

I've actually thought about this more than most people. I went through a very deep zombie fiction phase around the time that I was actually interviewing at Uber. Could be. I think that if you can survive the first wave of the apocalypse, which is when everybody is becoming a zombie, if you can get out of a major city, and I'm in San Francisco, so I don't have great odds, but if you can get out, I think the very best place to go in the zombie apocalypse is Iowa.

Ken Babcock (05:45.671)

Could be nonfiction soon.

Ken Babcock (06:03.838)

Wow. Just, just a hub of corn. You could eat a lot of corn.

Kristin Kaiser Mulligan (06:08.061)

I just think, yeah, exactly, it's probably got a good food to people ratio right now. Not a high population, so there's a lot of food there. So if you could get out and get to Iowa, because I think after being attacked by zombies, obviously, the primary concern, the next one would probably be starvation. So you got to go somewhere where there's food. I thought this through.

Ken Babcock (06:15.05)

Right now, yeah. Yeah.

Ken Babcock (06:30.098)

Sure. Okay. I mean, I was kind of flat though. I don't know. I'll think on it.

Kristin Kaiser Mulligan (06:35.266)

Can I ask you, you need like almost some terrain to protect you?

Ken Babcock (06:38.654)

Yeah, yeah, I'm more of a terrain, terrain zombie avoider, but

Kristin Kaiser Mulligan (06:44.129)

Where would you go? Are you going in the mountains?

Ken Babcock (06:46.862)

I think I'd want to be on skis.

Kristin Kaiser Mulligan (06:49.043)

Okay.

Ken Babcock (06:50.443)

Like, I mean, I'd have to ski for the rest of my life, but that doesn't sound so bad.

Kristin Kaiser Mulligan (06:54.573)

Yeah, but you get down the mountain you have to get out of the skis pretty fast.

Ken Babcock (06:59.242)

Or just get on the chairlift and hope that they don't get on the chairs. So.

Kristin Kaiser Mulligan (07:03.135)

And your zombie apocalypse, the chairlifts are still running? Yeah.

Ken Babcock (07:06.814)

Oh yeah, someone's... the zombies are running, for sure.

You just, you just, you just picked out the fatal flaw in my idea, but, um, yeah, ideally they're running. So I want to jump into the topic because, because it's, it's meaty, you know, this decentralization versus centralization sort of, I wouldn't even call it a dilemma. Cause I do think that there's, and we'll unpack this, but there's a lot of people that probably think centralization standardization is critical, uh, you know, and I'm sure they're probably surprised to hear like, Oh, the tango CEO is saying

Don't standardize, don't centralize. So I want to dig into that a little bit, but can you start with kind of telling us your framework for when you think about centralizing versus decentralizing and then we'll jump into some examples.

Kristin Kaiser Mulligan (07:57.113)

I think about centralization and decentralization as you centralize what you know, you decentralize what you want to learn. I think there's pros and cons to each and when you think about organizational structure as well.

When I think about centralization, the things that come to mind to me from a pro standpoint are consistency and control, right? The cons of a centralized organization are closer to, you know, slow decision making, you're lacking innovation, there may be some bureaucracy because centralized organizations tend to be a little hierarchical. But again, you're centralizing what you know, so that might be okay for you in that time and place. And then when I think about decentralizing what you wanna learn, you're almost optimizing for the opposite.

The pros of that type of organization are incredible flexibility, responsiveness, people can work at speed. You see a lot of innovation in decentralized organizations done really well. And you see a lot of autonomy, which is really, really empowering if you're hiring really smart super motivated people. You put them in a decentralized organization, they're really going to flourish. The cons of course to decentralized organizations are the opposite of the pros to the centralized. So you know, it's a lack of control.

lack of consistency, and if not done well, a lack of coordination.

Ken Babcock (09:13.826)

Yeah. And I think for a lot of the folks that are listening, there's, there's probably some, I'll call it a virtue around centralizing where decentralizing might be the opposite of what they're trying to do and achieve. You know, they probably want some of that control. They probably want to say, Hey, you know what? We need to follow like this set of processes across the board so that we're consistent, so that we're efficient. What would you say to those folks to help them maybe get over the hump a little bit on, on the.

decentralization piece.

Kristin Kaiser Mulligan (09:46.537)

Yeah, I would say, you know, I've had the fortune in my career to be at Uber at Series D, I joined Lime at Series C, I joined Wheel at Series B, and I joined the company I'm at right now at Series A. And what I've noticed is, regardless of the stage, my experience has been that the company that learns the fastest wins. This is a mantra that I learned

Kristin Kaiser Mulligan (10:16.711)

across all stages. And if you're really optimizing for learning, I think, and speed, I think that decentralization is a virtue and could be an end all. It doesn't mean that everything should be decentralized. There are certain things like compliance and safety and planning that really benefit from centralization. So what I would say to those people is you probably are looking for more of a hybrid approach. A central operations team, something that can

mentioned compliance, those scalable features like customer support where you really want that customer experience or that internal operational experience to be the same. You can still have that team while also decentralizing those things that you still want to learn. So a decentralized org could be a highly execution org. It can be an org that does the things that don't scale. I think people are really afraid to do things that don't scale, but some of the best products

from those learnings. And if you're testing, experimenting, if you're still trying to find your product market fit, if you're still trying to find that right ICP, that right customer, I would say that you want a small decentralized arm that's really fostering that environment of speed and innovation.

Ken Babcock (11:35.414)

Yeah. And there's some, there's some decision in there around like, what do we just want to execute on consistently versus like, what do we want to innovate on? And I, and I think there is almost like some introspection that needs to be done of like, okay, if we're going to win, what do we want to be really good at? What do we want to learn about versus like, I think you gave the example to me when we were prepping like, Hey, the way we should repair a scooter should be the same across the board, right? We don't need to innovate on that.

So I think that was, that was a big takeaway for me too. Can you maybe talk us through an example of decentralization and when that worked really, really well and ultimately like benefited the company?

Kristin Kaiser Mulligan (12:18.233)

Sure.

Yeah, just staying with the scooter example, and you're absolutely right, we had at Lime a central repair team, and that team worked incredibly closely with our hardware team because we're not trying to innovate on how to fix a scooter when it comes to safety and standards. And so that's something that we would really, inside of our warehouses, want to centralize and to standardize. But at the same time, we hired a lot of very smart people and put them in these warehouses.

were often like a different shape, they were in a different location just based on availability, based on price. Some of these warehouses at Lyme had different models of scooters, some had scooters and bikes, some had just bikes, some had mopeds as well, and the flow of those warehouses especially in the early days are something that we were purposefully not prescriptive on. We had a lot to learn about what is the right flow of a warehouse, how should we make sure that

as possible and so that was something where we decentralized that early and we learned so much I'd love to share with you more on that and but beyond that how we shared that

We shared that innovation at our all hands. We had a weekly all hands, I'm sure you have one too. Those as companies get big can become like very like reporting out on metrics and roadmaps. And what we really optimized for at Lime was innovation. We were constantly asking people in those warehouses to share what was working, share what wasn't working. We use data to, you know, guide us there. So if you saw something, and we did this at Uber too, you probably remember.

Kristin Kaiser Mulligan (14:01.181)

you would see a city just light up on the leaderboard and every single person in local ops was reaching out to the ops manager that they knew in that city being like what the heck happened there how did you do it let me take that playbook and make it a little bit better so i can collaborate and compete with you and we had that we had that energy at lime as well and we were able to learn so much and so fast by just fostering that

Ken Babcock (14:29.654)

Yeah, I think the, uh, it's cool to hear the example from Lyme and it's cool to think back to like how we were doing this at Uber because it was this combination of like, yeah, let's, let's try to learn once we figure out the best way to do something, then let's scale that. Um, but it also was really competitive and you were, you started on the Austin city team at Uber, correct?

Kristin Kaiser Mulligan (14:54.146)

Yeah, that's right.

Ken Babcock (14:56.106)

And so like, talk to me a little bit about some of those dynamics where it was like, yeah, we were decentralized. You're tasked with your own market, but you know, were things competitive? Like, did you guys compete with Houston and Dallas and, you know, everyone else in your region?

Kristin Kaiser Mulligan (15:11.75)

Oh yeah.

Absolutely, that was one of the best parts of Uber. I think in those early days, like 2014, 2015, when we were launching Uber, we were in Austin, we were launching a lot of Central Texas. At the time, the team in the region that was innovating the most was Chicago. They had a lot of riders, they were doing a lot to onboard drivers. And so you wanted to talk to that team, you wanted to know what they were doing. And we would watch their metrics pretty closely,

even though we were way behind in terms of trips and drivers, just from an age standpoint. But yeah, my colleagues and I in Austin, we wanted to be the innovative team in the central region. We wanted people to reach out to us and ask us for our playbook for how we handled ACL, for how we handled South by Southwest. And that was really the beginnings of this hybrid world that we moved into at Uber, which was centralization and decentralization. So.

It happened so naturally, which was so cool. You were just passing playbooks on, and over time, it was like, okay, let's have a central team that's blessing these playbooks, that's doing these things that do scale, so that we can keep letting the local team innovate. So in Austin, Texas, maybe, you know, we're not best suited to run driver payments every single Monday.

somebody in Chicago or somebody in SF could run driver payments and now on Monday the person in Austin that was running payments can do something else and that was the beauty of Uber and how that evolution happened and then too we got to take a lot of those learnings and Expand our own careers as operators So everything that we learned in Austin about managing a huge citywide event like South by Southwest I got to take that learning and go to Rio

Kristin Kaiser Mulligan (17:02.331)

in 2016 and I spent two months in Brazil preparing for the Rio Olympics. And that was an incredible experience in using the learnings from South By because we just didn't really need to have that team learn it all over again after we'd done it for two years.

Ken Babcock (17:16.258)

Yeah. And that was what was so cool. Right. It was like the cities with the most airport traffic became like the airport's product team eventually. And, um, that was that flow where we were decentralized. We learned a lot and then we're like, okay, we don't even, we don't even necessarily learn anymore. Let's just go like productize it, which I think is really cool. But, you know, one thing I did want to ask about is, you know, as you're, as you're making that playbook or as you're saying, Hey,

Kristin Kaiser Mulligan (17:28.099)

Yeah.

Ken Babcock (17:45.734)

Austin, we figured this out. Um, what was like, what was the challenge associated with saying, okay, we, we've got to win. Let's go to go and apply it. Like what was challenging about that?

Kristin Kaiser Mulligan (18:00.409)

I think the challenges there were the documentation around it, like how do you get that word out to make sure that you were making the full company better? And it was really about how do I scale this beyond just what I was able to do in Austin? And I think

unique to Uber, everybody in local operations at that time was thinking that way. And so one of the challenges was also, as you remember, Team Dot.

Okay, now you have a lot of people who are trying a lot of different things, and they're throwing it all into this centralized location. None of it's blessed, all of it's kind of messy, and you could, in 2018, find a playbook from 2014 and execute on it if you wanted to.

Ken Babcock (18:36.14)

Yeah.

Ken Babcock (18:53.982)

Yeah. Team.for those who are listening was, was a confluence, you know, Confluence page, many Confluence pages. And, and, and people were constantly stumbling into, you know, if I searched driver payments or stunt promotions, or, uh, the one that I'm the one that I'm like getting PTSD from is like setting up dispatch radiuses, like it was just like.

Kristin Kaiser Mulligan (19:03.062)

Yeah.

Kristin Kaiser Mulligan (19:20.857)

Bye.

Ken Babcock (19:23.282)

You don't, you know, you, you'd pull stuff because you're like, okay, this looks good enough. Like, let me go do it. My head is, you know, spinning constantly. I'm not going to like dig into validate or like look at if that person's even still at the company, I'm just going to do what come. Yeah. This has a lot of views. I'll just go do it. So we would, we would often, we would often stumble on that. Um, so that, you know, the documentation piece is so critical and

Kristin Kaiser Mulligan (19:37.129)

Yeah, you're like, this has a lot of views. I'm gonna use it.

Kristin Kaiser Mulligan (19:44.098)

Yeah.

Ken Babcock (19:52.086)

Making sure you just have like the right platform and stage. I think the difference that I'm hearing between our experience at Uber and what you guys did at Lyme was that there was a very intentional platform for you to share the innovation, which I mean, Uber it was happening, but it was just like, everything was happening, you know, we didn't have necessarily a stage to do it. And Travis spent most of the all hands answering questions for 45 minutes. So.

Kristin Kaiser Mulligan (20:13.878)

That was important too.

Kristin Kaiser Mulligan (20:18.389)

Right, exactly. No, that was very intentional by our leadership team at Lime. Because when you have this mantra, which is the team that learns the fastest wins, you have to set up your organization to actually be in a place to do that. And

People say you learn from failure, but we were learning through success at Lyme. And that's what we were really trying to optimize for too. Let's share the things that work. Let's make sure everybody has access to them. And let's replicate those things. But at the same time, let's also share the things we tried that didn't work so that other teams don't try that. And that was a really important piece too. It was like, hey, we tried this. We thought it was a good idea. These were the results. And we're kind of gonna put it

this group of let's not try it because that is the huge big one of the biggest risks of decentralization is you have a bunch of teams that are all failing the same way and you really don't want to see that that's not going to help you move the company faster.

Ken Babcock (21:21.662)

Yeah, trying to recreate the wheel. So I want to go back to the lime warehouse example. And this for everyone that's listening, we're sort of, we're transitioning into the centralization side of the house. But with that example, you know, what were the things that ultimately needed to be standardized and. What were the big learnings that you took away and said, okay, stop trying on this. Every warehouse kind of needs to look this way now. Um,

if you can walk us through that, because I think it's fascinating.

Kristin Kaiser Mulligan (21:55.317)

One of the things that we measured pretty religiously at Lime was what we called scooter decay. The concept that a vehicle could enter the warehouse and not ever come out.

which is easy when you're working at a warehouse that might have 10,000 scooters in it, right? So think about how to manage a warehouse at scale. And like I said earlier, in the beginning, we were purposefully not prescriptive about what that warehouse flow looked like because we didn't know and we wanted to learn. One of the things that we did is we went around to a bunch of different cities of all different sizes and we took note of the flow.

and say, walk me through the flow of this warehouse. What we started to see was we were going to a lot of warehouses where, again, this tribal knowledge passed on from warehouse to warehouse, no centralization. A lot of warehouses were starting to organize their scooters by easy, medium, hard.

And like that makes sense on the surface, right? Like a scooter comes in, oh, super hard repair. Let's put it in the bucket with the hard scooters. Easy repair goes here, because we can get those easy repairs in and out really quickly. But I told you in the beginning is one of our biggest metrics was scooter decay, time that a scooter, an individual scooter stayed in the warehouse.

So if you're organizing your warehouse by easy, medium, hard, but we're measuring the proficiency of your warehouse by time, there was a huge disconnect between what was actually going on the ground and how we were reporting out on the metric. And so we had an idea to test changing the warehouse layout to actually visually represent time.

Kristin Kaiser Mulligan (23:45.681)

So that, and we called it FIFO, right? The first one that comes in, the first one that goes out, a scooter that comes in Monday morning, we want that one to come out of the warehouse before the scooter that comes in Monday afternoon. And now extrapolate that by, again, thousands of scooters moving through this warehouse, let's organize the warehouse by time. So we said that to the teams, no more easy, medium, hard. That was like definitive.

And we said we want you to do a FIFO and here's like some ideas of how you could do it but your warehouses are all different and so go forth and see how you do. The metric immediately improved.

It was just incredible. I always say the best way to move a metric is to measure it. But first people have to understand how it's measured and you have to talk about it all the time and you have to measure it on a frequency that makes a lot of sense for the rhythm of the business, which for us was weekly. So we were looking at this metric every single Monday. How did we do the week before? And then we went back out to all of the markets and it was so awesome to see like every different shape and warehouse, how they were doing it. Some warehouses.

literally had like corrals Sunday through Saturday. And so the Sunday corral had to be emptied before the Monday corral, before the Tuesday and so forth. Some of the warehouses bought like Livestrong bracelets, different colors, like Monday scooters were red, Saturday were blue. When we also saw like actual flows of warehouses totally change where you might bring a scooter into the back of a warehouse, it goes all the way up to the front. The mechanic stations are closest to those scooters that are closest to them.

So they're pulling from the front of the queue and now also they're not cherry picking the easiest scooters to repair. So now those hard scooter repairs are happening in time and during the day and not piling up and becoming this like physical representation of like the hardest part of your job. I think people do that too. It's not just about scooters and warehouses. Like you probably prioritize your day that way sometimes as well, which is like, I'm going to put off the hard thing.

Ken Babcock (25:35.031)

Yeah.

Ken Babcock (25:41.39)

Oh my gosh.

Kristin Kaiser Mulligan (25:44.961)

But like it was physical for us where it was like, oh my gosh, there are 100 hard things versus if you dealt with them every day, maybe there's just two or three. And it changed our entire business model, really it did. And it also changed how the people on the ground were thinking about the business, which was really meaningful.

Ken Babcock (25:45.068)

Yeah.

Ken Babcock (26:03.734)

Yeah, that's super cool. And a PSA for the audience, next time you abuse a Lime scooter, think about what Kristen had to go through and probably, yeah, and think about the person that is now at Lime that has Kristen's role and has to think about all your willy nilly actions with these scooters. Be nice to the scooters, yeah. That's the big takeaway.

Kristin Kaiser Mulligan (26:12.813)

Think about how hard it is.

Kristin Kaiser Mulligan (26:24.781)

Yeah, be nice to the scooters.

Ken Babcock (26:33.194)

No, it's really cool. And, and like, it's, it's just a part of like a consumer product that you don't think about, um, which I think is fascinating, but you know, when it came to, you know, this, this metric you measured, you know, you had, you had the specific metric, you, you landed on the right frequency, you educated the teams to know how they could move it and what mattered, um, is there anything else like

you know, maybe what were you measuring before and how did you sort of create that behavioral change?

Kristin Kaiser Mulligan (27:12.31)

Sorry. The behavioral change that made the biggest difference for us, I think this was across all, this has been across all of the companies that I have been at, but you really celebrate that centralization as a win. When you think about that...

you know, all of the warehouses before that had the easy, medium, hard, and that first warehouse that maybe thought of that idea and everybody followed it. And you're kind of, you know, at the same time that you're celebrating this new idea, you're kind of saying that one wasn't a good one. And that can be very hard in how you take those learnings and those lessons and apply them and celebrate them. And so like the easy, medium, hard, you could maybe take your Monday and divide it that way and then...

Ken Babcock (27:49.303)

Mm-hmm.

Kristin Kaiser Mulligan (28:03.049)

mornings you're doing your hard scooters and Monday afternoons you're doing your easy ones and so we still try to embrace those learnings while celebrating and really amplifying the one that we wanted standardized and centralized so that was a really big thing and the other thing too is we talked about how that standardization would scale not just the company but also the people in the company people think of especially in operations I think you get very defensive of like this is the thing that I own

and I do this very well, and so I'm gonna hold on to it forever and never let it go. But what you actually get when you centralize something is time to do something else, which is really cool. You and I have talked about payments at Uber a bunch of different times. At one point, there was a person in every single city pressing a button every single Tuesday to pay drivers.

Ken Babcock (28:55.018)

Yeah, it was, it was, and it was maybe your whole day leading up to that press of the button.

Kristin Kaiser Mulligan (28:59.193)

It was your whole day. It was your whole Monday. You could not get that wrong. It was just everything. And then they centralized that, and every single person in the city that was pressing that button on Tuesday now had their whole entire Monday free. And we hired the type of people that were like, with that free time on Monday, I'm gonna go freakin' figure something else out.

And I think celebrating that too, celebrating like centralizing something means that now you have time to do the other things is something that you have to do very, very well to keep that learning and that like, you know, wheel going to be honest.

Ken Babcock (29:38.142)

Yeah. And so, you know, documentation is obviously a big part of this. We talked about playbooks, creating those resources, you know, I think celebrating the wins and letting people know like, Hey, this is the, this is the Chicago team of this, of this process, like follow their lead. That's really important, but you know, for that documentation to sort of. Work and persist and to live beyond that all hands meeting where you put someone on a stage, like what's required.

for it to work well.

Kristin Kaiser Mulligan (30:10.345)

Yeah, this is one of the hardest things. And I think for your listeners, if you're at a big company and you feel like you haven't figured out documentation, like you're in good company. This is a very, very hard thing. I saw, like I mentioned, I've gone from series D to A, and at every single stage, one of the hardest things I think in an organization is documentation. Where I've seen it work really well is,

Where it's standardized, it's written for the masses, which means you can open it and understand what's going on, even if it's a highly technical spec. Maybe there's a summary that helps you understand it as an operator. It's succinct. The thing that you're doing is well-documented, centralized, you know where to find it. That's a really important thing, like going to TeamDot and just typing into the search bar. That's one way,

more organized whether you're using Notion, whether you're using Google Drive, wherever it is that you're like documenting everything. Do you have a folder that has playbooks? Are you labeling the playbooks as blessed? That was a great thing at some point that happened at Uber where you remember blessed queries.

It was like a big deal for the local team because you used to have to write your own query and you never knew if you were like maybe going to a table that hadn't been updated in a while. And so the central teams and the data team started blessing queries. That was incredible. That gave you so much leverage, even if you wanted to like slightly adjust it, you knew you were looking in the right tables at the right places.

Ken Babcock (31:27.838)

Oh yeah. Yeah.

Kristin Kaiser Mulligan (31:47.041)

That concept of blessed queries, we expanded it at Lime and at Wheel to blessed playbooks. These are playbooks that like the centralized team looked at and blessed and this is like what we wanna do. It has a date on it so you know how old it is. It has an author so you know who to reach out to. If something was the opposite of blessed, yeah, totally, you wanted your name on those blessed playbooks.

Ken Babcock (32:04.59)

And that became competitive too. Yeah, like that became competitive too. Oh my gosh, like all I wanted in this life was to get my query blessed. It was a different time, it was a decade ago. Especially me, yeah.

Kristin Kaiser Mulligan (32:18.205)

Oh my gosh, I bet, especially you. And for me as an operator, I wanted one of my playbooks blessed. I wanted people to reach out to me and ask me about the Rio Olympics. Like I was super proud of the work that we did there. And I wanted my name on that playbook so that the next time the Olympics were coming around and I was at Uber, like they would reach out to me so that I could go to the Olympics again, too.

Ken Babcock (32:44.351)

Yes, of course. I mean, that's so cool that you got to do that. Meanwhile, I was toiling away in San Francisco. They sent me to Miami though, that was fun. We launched Miami, which was a wild, we tried to use all the playbooks, but it was like, oh, and it was also like, oh, this is a different animal, this place.

Kristin Kaiser Mulligan (32:54.989)

There you go.

awesome.

Kristin Kaiser Mulligan (33:02.669)

That was early.

Kristin Kaiser Mulligan (33:07.533)

Okay.

Ken Babcock (33:09.806)

And so that was valuable too, because it's like, we didn't really know what we were getting into there. So as we kind of close out the centralization versus decentralization discussion, if you were to kind of summarize for the audience, like what are the big takeaways? What are the big things that like, you can learn from either side of it? What would you say?

Kristin Kaiser Mulligan (33:36.929)

I would just go back to.

centralize what you know, decentralize what you want to learn. It works at every stage. I think it works in every industry. I've been in transportation, Lyme was a transportation, but a big hardware company, Wheel, a healthcare company, and now I'm at FinTech. The learnings are the same. And most likely, you need a hybrid approach, to be honest. So you think about, again, centralizing those things you want to learn, centralizing those things you want to know, decentralizing those things you want to learn.

Central teams can cover averages really, really well. And so if you can centralize a lot of the work that you're doing, a lot of the...

outputs that your company measures. What you can do is really focus your decentralized teams on those P90 experiences, those things that don't scale, but they make a really meaningful difference for your customer. I'll tell you a quick anecdote just about Wheel. So Wheel is a B2B to C healthcare, virtual care staffing company. And one of our biggest metrics that we looked at was our performance on

Sorry, let me start that again. I wanna give you a quick example from my time at Wheel. So we had a centralized marketplace team that was matching clinicians and patients. And that centralized operations team and the product team, their focus was getting that right 99% of the time. And we were very, very good at it. But when you looked at that 1% and you look at a 1% at any company, especially at scale, you might be looking at hundreds or thousands of defects, depending on your size.

Kristin Kaiser Mulligan (35:17.599)

And so we used that central team to get to the 99. And then we were able to use our customer operations team to focus on that 1%. At the time at Wheel, that might have meant texting or calling a clinician and just begging them to come online, telling them we have a visit that's coming up in 30 minutes. It'll breach without you. And that made magical experiences for the customers and for the companies that we were partnering with.

It doesn't scale, right? You can't do that forever, but it enabled our centralized team to really focus and deliver for the 99% and get that average up. And...

gives the product team time to catch up as well. And it gets you to that next stage because you're creating this magical experience. The customer doesn't know that they're the 1% for you. They're coming in and just hoping to get matched with a clinician and have a visit at that time. So we did the things that didn't scale in our decentralized teams and we centralized the things that did. And I think if you're at any, like I mentioned, any stage bringing in that...

hybrid approach is going to help you deliver better and faster for your customers.

Ken Babcock (36:27.982)

I'm glad you mentioned that because I think even the way we structured this conversation, it was, we made it seem a little binary. Let's talk about decentralizing. Let's talk about centralizing, but taking that hybrid approach and sort of blending the two and saying, Hey, this isn't, you know, it's not a one size fits all. In fact, we can actually, we can actually do both at once. I think that's a, that's a great takeaway for the audience. And with that.

Kristen, I'm so glad you joined us for this because it's just so fascinating to hear all this and follow your career. I've been following from afar, so I feel very lucky to have you on the podcast. But thanks again. And the audience is gonna love this one. This will be a good one.

Kristin Kaiser Mulligan (37:13.909)

you. Thanks Ken for having me.