30-day Impressions; or Sonder vs Uber

Apparently, the honeymoon period at a new company lasts anywhere from six months to a year, depending on whom you ask. Well, I recently passed the 30-day mark at Sonder. Probably too early to give you the honeymoon-period-is-over inside scoop, but at least I’ve made it past setting up my email and vaguely know what I’m doing.

During this time, I’ve received an overwhelming influx of support and well wishes from friends. Funnily, almost every single person who’s reached out or come by the office to say Hi has asked the exact same question: How does Sonder compare to Uber?

So, here goes: my 30-day report of random things I’ve noticed at Sonder, with Uber as a backdrop for comparison.

Early Birds

One of the first things I learned was that my team likes to start their day very early. In fact, the whole company seems to have a culture of seizing the day. As a father of two who needs to balance work and family, this was a very pleasant surprise. And it makes a ton of sense given Sonder is in hospitality. I love this.

My day: I get into the office around 8:20 am, and some engineers are already there. I brew myself an Americano, settle into my desk, and get ready for the day. The team kicks off with a standup at 9:30 am (sub-teams have their standup at 9:15 am), then we break out to take on what we’ve committed to for the day, circling back to sync if needed, but otherwise operating autonomously and untethered.

The tight alignment we have schedule-wise goes a long way in allowing us to move quickly. This is in stark contrast to my prior team at Uber where engineers strolled in as late as 1 pm. By the time we have standup, half my day is already over, and it’s difficult to avoid last-minute disruptions at that point. We made it work, but it was far from ideal.

6 lbs

This is how much weight I’ve lost since I left Uber over a month ago.

They feed you pretty well at Uber; lunch buffet along with salad, sandwich, and soup bar. It’s nothing compared to its neighbors Square and Twitter, but it’s quite generous in the absolute sense. Outside of lunch, the multiple kitchens on each floor are stocked with an array of snacks; some healthy but plenty of not-so-healthy guilty pleasures at arm’s reach.

Trust me, I never complained about the food options at Uber. But like any human, I’m weak, especially when I’m stressed out. And oohhh boy was I stressed out at Uber. One could gauge how poorly my day was going by how many bags of Cheetos, cans of Dr. Pepper, and packs of It’s It ice cream I would go through.

In Sonder’s kitchen, you’ll find plenty of healthy stuff: gourmet coffee, tea, fruit, popcorn, and fizzy water. For lunch, we order from Forkable, which has a nice, varied and balanced selection. It’s most certainly not a buffet: after $15 we foot the bill ourselves. To me, this frugality is actually a benefit, because we all know what happens when we walk into a buffet.

If I’m stressed, I just step outside for fresh air and a stroll along lively Hayes Street. Not that I need to though since the office itself is amazingly calming. Which brings me to the next thing.

A Happy Office

Sonder’s office is simply beautiful. Bright, airy, almost serene. This is what an office should be like. This is where I feel I can be creative and collaborative.

Meanwhile, Uber’s offices can be best described as cavernous and chaotic, with the 5th floor of HQ (where I sat) affectionately referred to as The Death Star.

I’m really glad Sonder put so much thought into building a workplace that is warm, welcoming, and, for lack of a better word, happy. The office does make me happy, and all of my visitors have echoed similar sentiments.

Kudos to Ashley Redmond and her amazing Interior Design team — great job!

Standing Desks

At Uber, if you wanted a standing desk, you’d be given one of these really ugly and clunky Varidesk add-ons. When you adjust it up and down, you end up having random shit fall off or otherwise cause a ruckus. It essentially turned everyone’s desk into a mess.

I know this is small, but a lot of people at Uber liked standing, so these monstrosities were everywhere. And when your office is tight on space and desks are crowded together, it becomes a real mess. This was a big reason why I rarely stayed at my desk.

Sonder’s desks are all adjustable and do so without being a hazard. You have a complete, smooth surface no matter what height you want to work at. No fear of it toppling or making a mess. Now I actually enjoy being at my desk. Yay for good design.

Building Empathy

My first week at Sonder, we took a trip to the field to better understand our internal customers in Operations. I was excited to take a look behind the scenes to see how the sausage was made.

We started by shadowing Housekeeping in one our new developments. This gave us a chance to ask for in-the-moment feedback as they tried to use our app to report issues and assess the cleaning. Even though I’d never seen our Housekeeping app before, the scene was all too familiar as they expressed their frustrations and schooled us on how broken our tech was. It was great.

Next, we visited the SF city team’s warehouse where we got to see what goes into onboarding a unit (i.e. getting a unit ready for rental). It was really eye-opening to 1) see the dizzying array of moving parts people in Operations need to deal with and 2) the sheer hustle they possess to overcome any shortcomings in technology they faced. Their get-shit-done attitude was inspiring.

The week after, we flew to see the Chicago team for more shadowing and a deep dive on Housekeeping. These field experiences galvanized our team, and today two of my engineers spend one day each week working out of our SF warehouse. They work alongside Ops to make sure they have their ear to the ground and that we maintain a tight relationship with our customers. The best part is that they decided to do this on their own. Amazing.

At Uber, there was certainly a desire to build empathy between HQ and Operations, but I feel we could have prioritized it more highly. I didn’t visit our SF driver support center until probably four months in, and this was when it was located in the same building (it was later separated from HQ). My first time driving was about three years in. The experience left a strong impression on me, but it was long overdue.

To be fair, as a leader, it’s on me to reinforce the importance of building empathy for our customers. So this is a good lesson learned, and an opportunity to raise the bar at Sonder.

Foundational Investments

During our field trip to Chicago, I met one of our Regional General Managers, Craig Rashkow. Along with the local team, we took a deep dive on Housekeeping, focusing on our biggest problems/opportunities and the bets we’ll make towards them.

We started our brainstorming session by scribbling every Housekeeping-related problem we could think of on Post-It notes and slapped them up on the board. Then came the task of distilling and prioritizing them, and finally committing to the top problems. As expected, this was the most painful part of the brainstorm as we had to challenge ideas and priorities (as we should, and passionately so) in order to whittle down the list.

At a moment when the conversation had elevated a touch, Craig pulled us back and said (paraphrasing), “Remember: our core competency is in execution, and we need to make sure to invest in building a unified platform that ensures our long-term vision of operational efficiency.” While others may have seen this as a non-event, I was stunned. To hear a leader in Operations, or any non-technical leader for that matter, hold the line on making a technical investment towards a long-term vision was shocking. I couldn’t believe my ears. From that moment on, I had a newfound respect for Craig. The best part is, I continued to experience this perspective as I collaborated with others across Operations. (Shout out to Peter Bowen, GM of New Orleans, as well!)

At Uber, I’d been conditioned to expect short-term and reactive approaches to problems. In our quest to execute as quickly as possible, we starved long-term technical investments. As unpaid debt piled up, the on-call rotation got absolutely brutal. We did “move fast” but it was unsustainable. Most engineers burned out; those who survived were stuck maintaining brittle systems and performing tricky, years-long migrations. Eventually, engineering leadership wised up to the droves of complaints and ushered efforts to pay down this debt. At that point however, it was too little too late for many.

Up-Leveling

Uber is notorious for down-leveling. Couple that with a very competitive promotion process, and you get a whole lot of disgruntled employees. I’ve sat through more interview debriefs than I would have liked where I heard an interviewer say, “Well, I’m not sure, so let’s just down-level them.” For a period, it felt almost as though Uber took pride in down-leveling, or what some would euphemize as “raising the bar.” It was frustrating, to say the least.

When I joined Sonder, to my surprise, they decided to up-level me. That is, I’ve been given a big opportunity with high scope and remarkable autonomy. The areas I’m accountable for I can see one day being 30- to 50-person orgs, or possibly their own business units. Best part is that the teams are still very small and I get to build them out, which is what I love doing. For that reason, about half of my time is devoted to recruiting, where I coincidentally have an opportunity to get this leveling thing right.

Leveling and titles is a philosophical debate you can fill a novel with. Details aside, there are really just two things good leveling should achieve: 1) you feel you’re being treated fairly, 2) you feel challenged by high expectations. If either of these is not true, then you got something wrong. Fix the leveling, or accept that it won’t work and make the hard call to walk away from the candidate. And mistakes will happen, in which case you need to have the conviction to manage someone up or out.

When you place high expectations and big opportunities in front of people, amazing things can happen. People grow. They meet or exceed your expectations, and may end up surprising you. You just need to be open to the fact that sometimes up-leveling is the right move.

Press

Uber was in the press all the time, and in a very bad way. I could safely say this was true for the entire duration of my 4-year stint. As we all know, 2017 was a particularly brutal year. But even when Uber was at the height of its growth, the negative press barrage was non-stop. From the outside, it may not seem like a big deal — it’s just news, right? — but on the inside, as an employee, the stories amplified and the dissonance against reality was not only distracting but at times heartbreaking.

Sonder is nowhere in the news. I’m kinda hoping it stays this way for a while :-) Of course, it makes recruiting harder so it’s a double-edged sword. Gotta figure this one out.

Data

So, here’s where Uber actually does really, really well. I can wholeheartedly say that Uber is a data-driven company to the core. It is ingrained in the culture, from Engineering to Product, from Design to Operations, and more. I might go so far as to say this is a major part of why it’s so successful.

From the early days, Uber’s weekly company all-hands devoted a full half-hour to reviewing city performance metrics, and to this day any presentation at the all-hands would be chock-full of numbers. Walk through the office and you’ll see walls of monitors teeming with graphs.

On Growth, we published a report every single week on our progress against our goals. Above the fold was our trips and below were further cuts of data across markets and teams. Every chart had a progress line held up against a longer-term goal line; we were religious about it. In our weekly Growth meeting with Travis, we would supply him dense sheets of data and he would go through every single fucking metric and actually call us out when things were off. It was unreal.

What’s most impressive is that every single person in Operations knew SQL/Vertica/Hive and leveraged this skill to accomplish amazing feats. Without sharing any secrets, I can say that had it not been for this superpower, Uber honestly may not have survived the last few years. That Uber had gone to such great lengths early on to democratize data and ingrain it into the culture I believe was a gamechanger.

At Sonder, we are behind. Sad to say, but compared to Uber, I feel somewhat blind. True, it’s still early, and we’re extremely short on resources, but I wanna push. I am passionate about building a culture that lives and breathes data, so I need to get to work on this.

Marketplaces

My manager Vivek often uses the analogy, Airbnb is to Ebay as Sonder is to Amazon. The key difference is that Ebay started as a marketplace while Amazon was the seller itself.

Back in the day, Ebay was the darling of the Valley. Everybody was talking about them, and everybody wanted to work there. On the other hand, Amazon stock took a pounding as skeptics criticized their thin margins, capital-intensive business, and lack of profitability. Hard to imagine that now, right?

Fast forward a decade-and-a-half later to today: Amazon is a $761B business, and Ebay has been left in the dust.

So, why? What happened?

I don’t have the exact answer — otherwise I might be a very rich man — but I’ll assert it starts with having control over supply. In a marketplace, you could just be the facilitator of transactions between buyers (demand) and sellers (supply), levying a commission on these transactions. At first, you grow quickly because the marketplace is hot, but over time you find that you’re at the mercy of supply and demand. The only way to influence them is with money, whether through adjusting pricing (commission) or providing incentives.

Owning supply may be a slow and difficult climb towards critical mass, but the silver lining is that you have full control over it. As a startup, if you want to be huge, you’ll want to enter a high demand market. Naturally, such a market would be supply-constrained, thus control is paramount to efficiency. Amazon has been able to leverage this control, innovate the crap out of it (they have their own planes for Christ’s sake), and is now able to ship billions of products with “free” 2-day shipping. Of course, it wasn’t easy (it took 7 years to turn a profit), but I believe taking on those challenges early shaped Amazon into the company it is today. Once they got it right, it was easy to turn on their marketplace. Imagine Ebay (or Airbnb) attempting to pivot away from the marketplace model today; it would require a massive and fundamental change.

Is Sonder anywhere near Amazon? Nope. But we have a similar path ahead of us. We just need to stay on course and be willing to make huge bets that seem crazy now but could end up revolutionizing hospitality like Amazon did for retail. Haters gonna hate, just like they did Amazon.

Uber is a marketplace and has no direct control over its supply i.e. drivers are contractors, not employees. It’s no secret that both Uber and Lyft spend a ton of money on incentives in the hopes that drivers stay on their platform and drive at the right times and places. Imagine if they instead owned their supply: they could dispatch cars exactly where they want, when they want, and in the exact quantity they want. This could be done if they employed all their drivers, but of course, that ship has sailed. Driverless cars… now that would do the trick :-) Costs would drop dramatically (e.g. incentives, acquisition costs, insurance, even the 75% of the fare human drivers take home), so much so that Uber would be profitable overnight. You can see now why self-driving cars are such a huge deal to everyone and their mom who’s building one.

So the trillion-dollar question is Can Uber make the huge leap from being a marketplace to owning supply?

I sure hope so. Time will tell, and the world is waiting. If not Uber, then my money is definitely on Tesla. They just so happened to have started with owning supply and will have the luxury of “listing” on multiple marketplaces…

…

Well, that’s it for now folks. Hope this was helpful, or at least entertaining. I’ll have more to report at the 90-day mark!