Tango Blog
The Fast Lane To Learning: 7 Lessons from Uber and Lime

The Fast Lane To Learning: 7 Lessons from Uber and Lime

A Tango-branded illustration of a stopwatch

How often do you talk to someone who makes you want to order another cup of coffee, pull up your Notes app, and pretend you haven’t noticed the time?

We’d run late to just about any meeting for five more minutes with Kristin Kaiser Mulligan, Chief Operating Officer at Heard (previously Uber and Lime). With experience at Series A, B, C, *and* D startups, Kristin is a Swiss army knife of an operator. See also: a wealth of knowledge, a quote a minute, and an undeniable asset for any team in a race against the clock.

On the latest episode of Change Enablers, Kristin and Ken (previously Uber) go beyond their shared history of transforming how people get from Point A to Point B and dig into seven tactical tips to build a learning organization. 

And if you happen to have an electric scooter with a broken throttle, or need a playbook to orchestrate transportation for the next Olympics? 

Kristin can help with that, too.

1. The team that learns the fastest wins

Working at hypergrowth companies turned into household names has been the ride of a lifetime for Kristin. No matter where she’s worked, one thing has always held true: 

A quote from COO Kristin Kaiser Mulligan about the correlation between the speed of learning and the likelihood of sustained success.

If that mic drop of a soundbite wasn’t enough, check out the clip below for more about Kristin’s experiences at Uber, Lime, Wheel, and Heard firsthand—and to find out where this quote-turned-career-mantra came from:  

The last decade of Kristin’s career has shown her that it doesn’t matter whether a company has 10 people, 100 people, or 1,000+ people. Or if a brand is dreaming of its first birthday, celebrating unicorn status, or recognizing one of the many milestones in the middle. 

The most operationally excellent teams ask better questions, experiment more, fail faster, and adapt before everybody else.

2. Centralize what you know; decentralize what you don’t

Another one of Kristin’s takeaways from successful startups in the fast lane to learning? 

They approach learning differently. But not learning as in L&D, training, or people development. Learning as in the deeply human exercise of learning about, understanding, and experimenting with their business.

While most companies are grappling with whether to centralize or decentralize that kind of learning, winning teams are coming up with creative ways to benefit from the best of both worlds.

A Tango-branded image of a quote from Kristin Kaiser Mulligan, Chief Operations Officer at Heard.

The fundamental pros and cons of centralization and decentralization still apply, and we’ll touch on those in a minute. But what’s served Kristin most isn’t a dictionary definition or Venn diagram. 

Kristin wouldn’t be where she is today without:

  • Being willing to operate in the gray area (and take a hybrid approach to learning)
  • Surrounding herself with people who are [just as] curious, open, and forward-thinking
  • Getting really, really good at identifying what companies have a handle on already and what they haven’t fully understood yet 

Centralized learning: Upsides and downsides 

Before trying to design and implement a hybrid learning strategy, you need a good grasp of both models.

 At the end of the day, there are two big advantages and two big drawbacks to centralizing learning:

Reasons To Centralize Learning Reasons Not To Centralize Learning
Control Slow decision making (👋🏿, bureaucracy!)
Consistency (see also: standardization and scalability 🙏🏾) Little to no innovation

Maybe increasing safety or compliance is critical to your company’s success. Maybe standardizing your internal processes is going to have the biggest impact. Or maybe creating a uniformly delightful customer experience is more important to your organization right now than experimenting. 

In those cases, Kristin says centralizing learning makes a lot of sense.

Decentralized learning: Upsides and downsides  

Decentralized decision-making has its pluses and minuses too, of course.  

Reasons To Decentralize Learning Reasons Not To Decentralize Learning
Speed Lack of control
Learning Lack of consistency
Flexibility Lack of coordination
Continuous improvement/change

Maybe optimizing for learning and speed is going to take you where you want to go. 

In that case, Kristin argues decentralization is a virtue—and suggests that not striving to standardize everything is a sound business decision.

Why Kristin votes for a hybrid approach to learning 

In Kristin’s experience, the magic happens at both/and, not either/or. 

The most successful companies didn’t become “learning organizations” by centralizing or decentralizing all their learning.

They built winning teams by:

  • Juggling when to standardize (centralize) vs. when to learn (decentralize) 
  • Prioritizing control and consistency where and when it was logical
  • Not being afraid to do the things that didn’t scale, sometimes

Here’s Kristin on the benefits of having a small, decentralized arm focused on speed, learning, and innovation (especially for early-stage companies):

At Uber, leveraging a hybrid approach gave Kristin and her team:

  • The ability to centralize things like strategy and customer support
  • Freedom to decentralize execution, stuff that wouldn’t scale, and everything Uber was still figuring out as they expanded into new cities
  • Insulation from one of the biggest risks of decentralization 👇
“One of the biggest risks of decentralization is you have a bunch of teams that are all failing the same way. That’s top of the list of things you don't want to see, as an operator. That's not going to help you move the company faster."

-Kristin Kaiser Mulligan, Chief Operating Officer at Heard

3. Optimize for innovation

In Kristin’s opinion, the [massive] trade-off for centralizing learning is a lack of innovation.

At Lime, that was a non-starter. Innovation was a north star, company-wide focus. It didn’t just pop up occasionally in their All-Hands meetings—it was the purpose of their All-Hands meetings. 

“As startups grow, All-Hands meetings usually 1) decrease in frequency, and 2) morph into a forum for reporting on metrics and roadmaps.

But what we really optimized for at Lime was innovation. We used our All-Hands to create an intentional platform to share experiments and learnings from around the company."

-Kristin Kaiser Mulligan, Chief Operating Officer at Heard

Lime leadership asked themselves over and over again: 

  • What do we want to innovate on?
  • What DON’T we want to innovate on? 

Here’s Kristin’s response to that question: 

While most companies learn from failure, decentralizing (some) learning and optimizing for innovation set Lime up to:

  • Learn through success
  • Spend their energy wisely (i.e. not on exceptional scooter repair) 
  • Increase autonomy across the board
  • Attract top performers
"When decentralization is done well, you see a lot of innovation. And a lot of autonomy. Which is really, really empowering for super smart, motivated people. They're the ones who are going to flourish—and help attract more self-starters as you grow."

-Kristin Kaiser Mulligan, Chief Operating Officer at Heard

4. Be flexible about your approach and accountable to your outcomes 

Let’s talk a little more about Lime. 

If the brand rings a bell, you already know why Kristin knows how to repair an electric scooter. 

If you aren’t yet familiar with Lime, all you need to know is that their electric bikes and scooters have powered over 500 million rides in over 250 cities in five countries—and they need *a lot* of warehouses to keep their fleets charged and ready to ride. 

More than you ever wanted to know about scooter warehouse workflows

When Lime was getting off the ground, Kristin and her team gave each warehouse the autonomy to more or less run their own show (and report back). 

Decentralizing learning was the right call at first. No two warehouses looked the same back then, and their goal was to learn how to manage warehouses at scale. 

The problem (with decentralizing learning without guardrails) 

Over time, lots of warehouses started organizing their scooters by repair level difficulty, despite centrally measuring warehouse efficiency by how long a scooter was waiting for repair. 

When a broken scooter with a tricky repair came in, the team would set it off to the side. When a scooter with a quick fix came in, they’d wheel it over and knock that one out first. It was a reasonable way to manage the fleet, but it was negatively impacting global fleet depreciation metrics in a big way.

It exposed a fundamental break in the way Lime went about decentralizing learning. Without a clear understanding of the metric to drive the desired behavior from the mechanics in their warehouses (first in, first out—regardless of the maintenance effort required!), there was a serious lack of accountability. Sunday’s corral of scooters would literally never be emptied before Monday’s corral, and so on and so forth, because there wasn’t 1) a clearly communicated goal, or 2) a system to achieve it. 

The solution

To bridge the gap between what was happening on the ground and the business outcomes that would be most beneficial to the business, Lime introduced “scooter decay”—a metric you’ll probably never need to report on.

Scooter decay data measured how long damaged vehicles  would sit in a Lime warehouse, unable to be used. Once Lime had the decay metric in place and agreed “first in, first out” could apply to more than just how to stock a grocery store shelf, they asked a bunch of questions to instill behavioral change:

  • What if there were visual cues to indicate which scooters arrived on which day of the week?
  • What if the warehouse layout could be rearranged to visually represent time? 
  • What if mechanics’ stations were placed at the front of the queue, so they had to pull from the front of the line?
  • What if it were a lot harder to cherry pick the easiest scooters to repair, and a lot easier to keep 100 hard things from piling up? 

The impact

Those questions teed up a series of tests that ultimately:

  • Put an end to “easy,” “medium,” and “hard”
  • Ensured the first scooter to come in would always be the first one to go out
  • Improved Lime’s lifetime value and scooter decay metric on a global scale 
  • Not only changed their business model, but also transformed how people on the ground thought about their work

Kristin’s own learnings were just as monumental: 

  • If you’re going to decentralize, you have to help people understand the outcomes they’re accountable to
  • The best way to move a metric is to measure it

What’s the real takeaway to this story, though? Be nice to the scooters. 🫠

5. Double down on (better) documentation

There’s one prerequisite to building a successful learning organization that Kristin hasn’t seen anyone crack the code for yet. And that’s documentation.

Specifically: documentation that 1) doesn’t take forever to make, and 2) actually gets distributed/used.

What makes documentation so critical, especially for fast-moving teams? 

"Without good documentation, it’s nearly impossible to share learnings effectively. 

But when the important things are documented well, people can replicate processes that worked and avoid spinning their wheels on stuff that didn’t work."

-Kristin Kaiser Mulligan, Chief Operating Officer at Heard

What good looks like

In Kristin’s experience, good documentation comes down to a few best practices that 1) sound simple enough in theory, and 2) are still so rarely put into practice. 

"If you feel like you haven't figured out documentation, you're in good company."

-Kristin Kaiser Mulligan, Chief Operating Officer at Heard

But what *has* Kristin figured out, over the course of her career?

Properly documented processes and playbooks are:

  • Succinct. Another heavy ➕ for minimum viable context!
  • Visual. Bonus points if your screenshots are cropped, annotated, and focused on what matters most. 🏆 
  • Clear. The clearer you are, the fewer 🆘 calls you’ll have to answer.
  • Accessible. No one wants to look through multiple knowledge bases. Embed training in the tools where people work. 🙏🏻
  • Dated. Help people understand if they’re looking at something from two days, two weeks, or two years ago. 🙄
  • Archived. Instead of deleting, deprecate as part of your learning.
  • Trustworthy. Uber had a [centralized] system for “blessing” queries and playbooks that Kristin still thinks about to this day. 😍

Here’s a little bit more about that last one—and the beauty of sharing centralized playbooks and continuing to let local operators innovate:

"All I wanted in this life was to get one of my playbooks blessed. I'm joking...but also not. 

I wanted people to reach out to me and ask me about the Rio Olympics. 

I was super proud of the work that we did there. 

I wanted my name on that playbook so that the next time the Olympics were coming around and I was at Uber, they would reach out to me so that I could go to the Olympics again, too."

-Kristin Kaiser Mulligan, Chief Operating Officer at Heard

6. Don’t be afraid of a little healthy competition

As you may have gathered, Kristin has a competitive side. As did most of her co-workers at Lime and Uber (Ken included).

In her opinion, there’s nothing like a little friendly competition to kick learning into overdrive:

"Every so often at Lime, we'd see a city just light up on the leaderboard...and every single person in local operations would reach out to the ops manager they knew in that city and say, '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.'

That energy (and healthy competition) taught us so much, so fast."

-Kristin Kaiser Mulligan, Chief Operating Officer at Heard

Competing with other cities was also one of the best parts about being at Uber in 2014 and 2015: 

7. Always be building

To wrap up their conversation, Ken asked Kristin about the one piece of advice she’d give to operators today. Her response? 

A Tango-branded image of a quote from Kristin Kaiser Mulligan, Chief Operations Officer at Heard.

As an operator, it’s easy to get obsessed with the idea of being the “specialist.” You’re the only one who can run payments; you’re the only one who can update a complicated spreadsheet; you’re the only one who can onboard a customer. 

If that sounds familiar…“centralize what you know; decentralize what you don’t” might be as good of a mantra for you as it’s been for Kristin.

The bottom line is, hoarding knowledge—and shrinking your bandwidth for learning new things—is bad for your company and bad for you. In both cases, it stunts your growth. 

If you’re someone who excels at a process, go train someone else on it. And then go learn how to build another one. According to Kristin, this is how you:

  1. Prove you are an incredible asset to your company
  2. Stay engaged and keep growing in your career 
  3. Learn fast and win big
"If you have nine people pressing the same button every week and you can get that down to one, you’re left with eight people with the opportunity to do something new. 

That’s a huge win for the company and those people."

-Kristin Kaiser Mulligan, Chief Operating Officer at Heard

Want to hear from more operations leaders like Kristin? Subscribe to Change Enablers Podcast on Apple Podcasts, Spotify, and YouTube.

The bottom line


Keep in touch

We'll never show up
empty-handed (how rude!).

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
This is some text inside of a div block.