The DoorDash Operating System, part 2.
Finance sets the number.
It was the summer of 2018 and I was sitting listening to DoorDash’s weekly business review from our sometimes-air conditioned Austin HQ / rat Four Seasons.
“Steve, I think that was for you.” The voice was unmistakably [and alarmingly] my boss, Keith.
I sat up in my chair. I had not been paying anywhere close to what you might call a world-class level of attention.
“Could you repeat the question?”
“Yup, what’s customer service’s average handle time?” This was a new voice, emanating from a postage stamp-sized video feed of DoorDash’s main conference room in San Francisco. I couldn’t figure out who’d be asking such an inside baseball CX question.
“It’s about 14 minutes,” I answered back, with a level of confidence that could only be described as ‘not very confident.’
“Okay, that’s too high. What’s our cost per agent hour?”
Realizing this would be a more lengthy shelling I dug deep and activated my maximum serious person business voice1.
“Sorry, can I ask who’s speaking?” There was long, concerning silence, and then Keith again.
“That’s Prabir, our new CFO”.
Oh.
What to do with your CFO.
As a startup scales, its founder(s) will invariably add a Head of Finance. Sometimes this is a brilliant, thoughtful decision. Sometimes it’s “I dunno, we make money, a finance person does money, let’s hire one so we don’t have to do all that garbage money stuff anymore.”
In the latter case, a founder will typically end up hiring a ‘Head of Bank Accounts’2, someone who simply makes sure that money comes in, money goes out, and all the associated paperwork gets done.
The problem with this ‘founder directed’ approach (left diagram below) is it misses the point of having a CFO. You’re just adding another somewhat fungible executive, and giving them a bunch of tasks3. The result is that you, as founder, are still playing the critical role of setting the right objectives for each of your directs and then enforcing that each of them hits. Meaning that if, for instance, your company plan is off-track because your CMO is missing on new customer acquisition then it’s still on you to spend a week running-down that problem. And at a certain point, as CEO, that’s not really your job anymore.
The DoorDash model was different, in that its CFO played the role of effectively running the company’s P&L (right diagram below). He, broadly speaking, took Tony’s direction on where the business needed to go in a given year, and then translated that into how the P&L needed to evolve and improve over that time period.
So how did this work in practice? In a prior post I wrote about how DoorDash built a very effective operating system for getting a large, complex organization to ‘hit the number’; a way of translating a clear, ambitious annual goal down into quarterly plans, weekly sprints, and daily iteration. The following explains how DoorDash’s CFO and finance organization played the critical role upstream, of figuring out how to set that number in the first place.
Finance sets the big number.
Founders don’t become founders - especially successful ones - by being all that firmly moored to reality. No, most choose to do what they do because they have a dream for how they can help build a slightly better world, no matter the daunting odds or obstacles that stand in their way.
In a similar fashion, you can imagine that a founder’s impulse, when it comes to goal setting, is to set one in the stratosphere and let that inspire the team to think differently, push themselves further, and ultimately achieve a new level of performance. The result of this approach, unfortunately, is that at some point, the team will start missing plan, lose faith that their goal is achievable, and develop a Catch-22-level of existential despair.
Enter the CFO. They set the number.
Let’s imagine that you as a founder have decided that next year your biggest strategic priority is that you want to conquer your local market, while holding firm at a given cash burn multiple. Rather than pulling an all-nighter to crank out a jank-dizzle4 Google Sheets model for how you might possibly achieve that, you instead get your CFO to do it. It’s their job to figure out what output metrics need to move, how their corresponding inputs need to move, how much needs to be spent on growth, how many new people need to be hired, and how much that will all cost, to arrive at the optimal solution5. And then in doing so, figure out where each of your teams can be pushed to achieve an ambitious yet credible top line objective.
Or, if helpful, let’s explain in terms of a Formula 1 racing team:
Your business is the car
Your CFO is the chief engineer who knows everything about the car, including exactly how fast it should be capable of getting around the track
Your COO is the driver who’s responsible for driving the car it at its maximum speed to win the race
Your CEO is the team owner/ principal who wants to win the championship, but delegates winning each individual race to his/her CFO and COO so he/she can focus on making sure there are enough sponsors to keep the team competitive (or at least solvent)
Your CFO doesn’t just take the reservation…
If you’ve seen even a few episodes of Seinfeld you’ll know that it’s not enough to just take the reservation, you must hold the reservation.
In the same way, finance is not just responsible for setting the number, it’s also responsible for holding the business accountable for ultimately hitting the number.
In a sense, the vibe is this: the company’s P&L is the CFO’s baby, and you better not mess with their baby (by either not generating enough revenue or, worse, spending a more than expected amount of money). Should you find yourself guilty of either you can expect an exciting conversation with your CFO - like the one I reported earlier - at your company’s weekly business review.
Enter the finance goon squad.
The next logical question is ‘ok, well I get the idea here, but how does the CFO do all this themselves?’
The answer is, ‘they don’t; they have a team of ‘finance business partners’.
Every important team at DoorDash - and in turn every important line of the P&L - was run by a cross-functional pod that brought together ops, product, engineering, and analytics. So, for example, ‘US Marketplace revenue’ would be nominally owned by an operator/ GM, and then supported by directors of product, engineering, and analytics, all working as a single cohesive unit. They would then be assigned a finance business partner, whose job it was to 1. develop their own point of view on the US Marketplace business such that they could set credible annual goals, and then 2. work hand-in-hand with the US Marketplace team to ensure they were tracking to that goal on a monthly, quarterly, and annual basis.
The sum total here was that DoorDash’s CFO had someone from his team expert in and keeping tabs on every single line of the company’s financial statements.
The art and science of sandbagging.
As the owner of one of DoorDash’s largest cost centers, my lived experience on this topic was 20 straight quarters of receiving relentlessly challenging goals for improving our efficiency. Four times a year my finance business partner would rock up to my desk, and say something like “you won’t like this, but we need CS costs to come down by 20% this quarter, and here’s the math for why we think that’s possible.”
My first reaction was typically to explode with the furious intensity of a supernova and then black out. After coming to, I’d circle up with my team to figure out what would need to be true, and what resourcing would be required, to achieve this insane target. And thus would begin a back and forth negotiation with finance to add more resources here, to reduce the goal there6, until a mutually acceptable solution was achieved. At that point the goal was locked, and the team was sent off to execute for the next 12 weeks.
Why was this relationship magic? First, my finance business partner understood the mechanics of my business so deeply that I couldn’t just brush him off as having no clue what he was talking about.
Second, the ambitiousness of our goals forced us to consider every option available. Were there sacred cows we should finally address? Were there processes that instead of optimizing we should simply eliminate? Could we push ourselves to go faster, and achieve more than we’d thought possible?
If you’re a founder and you find yourself asking “what does my CFO do all day”, there’s a decent chance that this blog is for you. Genuine CFOs don’t just count the money, they help set your annual objectives, they manage board expectations, and they ultimately hold the business accountable for delivering results. A good litmus test for a founder? Ask yourself: if I skipped every weekly business review this quarter, would I still be confident in my executive team hitting plan? If the answer is no, you’ve probably got to start rethinking your CFO and the role they’re playing.
Next time on Missing & Incorrect: Pods.
Somewhere between Tim Calhoun and a very courageous cartoon mouse.
Or sometimes it’s just a ‘Head of Fundraising’, which sort of feels like hiring a random from Craigslist to go to all your kids’ parent-teacher conferences.
For example, how many zillion startups have given their new CFO the first task of renegotiating their Stripe contract?
Modern slang: ‘largely wrong’.
Very importantly, this is a bottoms-up view of how inputs need to move - informed by a deep understanding of the underlying business - to achieve your top-line goal. It is not a top-down ‘what you have to believe’ model. If your CFO comes back to you with “well, the total market is $15B and if we take another 2% of that, that’s $300M in new revenue”, stuff them in an oil drum and launch them into the sun.
Rarely.