The DoorDash Operating System, part 0.
The discomfort of building.
I have this tinfoil hat theory.
The biggest goat-rodeo companies out there get that way because they accept, rather than fight, the natural human desire for comfort. And in doing so, they ultimately seize into a state of inaction, decay, and dysfunction.
Think about every group project you’ve ever done. How many times have those projects resulted in you doing all the work?1 Or you deciding it’s easier to just agree with your teammates, and get a B?2 Or you all wasting so much time debating an approach, you hand in the assignment late?
Why?
A reversion to comfort.
In short, human beings naturally crave comfort.
We generally don’t like risk, we don’t like conflict, we don’t like failure, and many of us, given the option, simply don’t like working all that hard. Which makes sense, since all of these things, for a caveman, could end with a woolly mammoth yeeting you into a tar pit.
As a result, we naturally build our modern organizations to maximize our own comfort and, in turn, ensure we have to do very, very little. Specifically:
We avoid risk with bureaucracy: you’ll never have to risk something personally if you require total agreement to do anything
We avoid conflict with groupthink: you’ll never argue with a colleague if you simply agree 100% of the time
We protect our pride with statusism: you’ll never have to take direction from anyone with a lower title
We avoid failure with ‘shiny object syndrome’: you’ll never fail at your current thing if instead you just keep starting new things
And we avoid working hard with ‘empire building’: you’ll never have to put pen to paper yourself if you have an expensive army that’ll do it for you
Put it all together, and who wouldn’t want these things? Never risk anything, never advocate for anything, and never do anything, and keep getting paid and promoted?
The discomfort of building.
You can say a lot of things about the early years at DoorDash, but one you can’t is that it was comfortable. After all, what’s more uncomfortable than:
Knowing that you’re continually being given, as a team, completely ludicrous company goals?
Knowing that you alone are singularly responsible for delivering a critical input to those goals?
Knowing that you will be held accountable for your pacing to that input every single week, in front of all of your peers?
Knowing that you are expected to move at full-speed in pursuit of your goal, 365 days a year?
And knowing that if you ultimately fail, there are a half-dozen other loons chomping at the bit to take over where you’ve left off?
Hell, my hotel room my very first week at DoorDash was at The Murder Inn in the Tenderloin, in a room so small I could reach out and touch the not-even-remotely-lockable-enough doorknob from my bed. Not comfortable.
I paint a dismal caricature; early DoorDash was indeed many other adjectives as well, including energizing, exciting, and rewarding. But it was absolutely a grind, and that grind was driven by an operating system designed to fight complacency and self-satisfaction.
Recently I’ve found myself wondering, ‘why’? Why did Tony, Christopher and others design it like this? Why was this model good? Why not orient around “unity” or “collaboration” or “creativity”? Surely there are plenty of other successful companies with completely different organizational programming?
The answer, I think, is that you ultimately build your operating model to fit your business. And DoorDash’s business was and is, quite obviously, a terrible one. It requires the pinpoint pickup and delivery of a rapidly expiring, relatively inexpensive product, with effectively no lead time, for an easily-angered customer. And it requires doing all that for effectively no money. So you can imagine that scraping margin out of that business was challenging and required the lights out execution of everyone involved.
So given those cards, you start slowly understanding why you’d build an entire company’s battle rhythm around the unusual idea of ‘discomfort’.
Deploying discomfort.
So, you want to make your team less comfortable, where to begin? DoorDash focused on just a few areas:
Hammer a stake in the ground (quantitative, immovable goals): To have your team feel the discomfort of potentially missing a goal, you must first - shocker - have a goal. So, ffs, do that. (And importantly, those goals aren’t squishy in that they never, ever change mid-stream; if you’re gonna miss, you miss and try again).
Put the ‘individual’ back in DRI (clarity of ownership): Yes, obviously having one person own an important outcome drives urgency; nothing lights a fire under your ass faster than knowing that if your thing fails that it was your fault. But beyond that, it’s unbelievably helpful for a large, complex, cross-functional project to have one person gathering all necessary inputs, and ensuring that the collective group is delivering them, on-time.
An empowered leader among equals (single-threaded leadership): The flip side of the coin above is that each DRI is, within their sphere, empowered to make decisions, break ties, and override vetos.
Sunlight is the best disinfectant (Weekly Business Reviews): DoorDash’s WBR accomplished two important things. First, it was uncomfortable to report a miss, in red, week-in and week-out, in front of your most senior peers, so you really tried hard not to do that. Second, the WBR was anchored in quantitative goals and actuals, such that there were objective hits and misses, and executive soliloquies and grandstanding were kept to a minimum.
A team of title-free builders (no/ few titles): Uncomfortable companies fight statusism and entitlement by flattening or, in some cases, eliminating titles. Square famously championed nearly everyone being a ‘lead’ of something or other, and DoorDash sought to collapse titles similarly (no doubt encouraged by DoorDash3 legend and chief title antagonist Gokul Rajaram).
Discomfort’s brand problem.
Once upon a time I worked as a consultant for a large, national, casual dining restaurant. As part of this project I was responsible for leading a group to reinvent the restaurant’s bar area. One of our main suggestions was to expand their beer taps from a neolithic 6 to a more modern 12, such that they could add some millennial-appropriate IPAs and stouts to their Bud Light, Bud Lighter, and Mexican Bud rotation.
This project lasted 12 grueling months. We spent eons scheduling meetings, gaining alignment, losing-then-regaining alignment, and resourcing an effective plan. And then, at the 1-yard line, the facilities team determined that adding another half-dozen taps was literally impossible. It could not be done. That was that, the project died.
The result? The team was dejected, with their hard work effectively down the drain. The restaurant’s customers continued to defect to new, innovative upstarts. And, not for nothing, the company’s stock price continued its slide, eventually wiping out huge amounts of employee wealth.
That is the cost of comfort. No one led. No one made a decision. No one disagreed or fought for what they believed in. And ultimately, the company crashed into the ground, needing a full rebuild.
So next time you’re sitting in a meeting wondering “wtf are we even talking about here?” ask yourself whether the problem might be that everyone involved is a little too comfortable.
Not my energy.
There we are.
And Square, and Meta, and Google…