The CFO role has evolved faster than the tools built to support it.

Most finance teams are still running infrastructure designed for a job that no longer exists. The reporting. The reconciling. The close that bleeds into the next month. That's not finance - that's overhead with a title.

Agentic Finance shouldn't multiply your output - it should eliminate the work that was never worth doing in the first place.

That's why I run Mostly Media on Brex - an intelligent finance platform with AI-powered agents that do exactly that. Expenses handled automatically, policy enforced before the spend happens, books closed in minutes. So I can spend my time on the work that actually moves the business.

Gee wiz. I run a recruiting company now. Life comes at ya fast.

We do finance and accounting recruiting. And we place at the manager, director, and VP levels.

👋 Hi, it's CJ Gustafson and welcome to Mostly Metrics - my newsletter for CFOs about the metrics and financial strategies behind the world's best businesses.

I've been in finance my whole career and I have never worked with a product you can touch.

(I did lay blue stones for multiple summers growing up. And while there was no math involved, I became an expert in lifting heavy things up and putting them down)

I’ve helped budget for software, transactions, and data. The closest I've come to participating in a physical supply chain was when I worked at a marketplace that helped local auto mechanics source tires and aftermarket car parts.

So when I sat down with Claire McDonough, the CFO of Rivian, on the Run the Numbers podcast, my brain basically short-circuited.

Claire helped take Rivian public in one of the largest U.S. IPOs ever (~$12 billion raised). She came up through J.P. Morgan investment banking, where she worked on the Flipkart acquisition for Walmart, and also helped take Peloton public. And now she's running finance for a company that vertically integrates everything… from designing their own silicon chips, to building drive units, to manufacturing vehicles in a 4 million square foot plant in Normal, Illinois.

I want to bring you three things from our conversation that I think apply far beyond the auto industry:

Let's get into it.

Rivian made a decision a couple of years ago that surprised people: instead of building the R2 (their new, more affordable ~$45K SUV) at a brand new greenfield plant in Georgia, they decided to build it first in their existing Illinois facility.

Claire told me:

By squeezing R2 production into the Normal plant, Rivian saved roughly $2.25 billion in near-term capital expenditures. They'll still build out Georgia at a later date (with 400,000 units of capacity), but only once they've maxed out what they’ve got.

This is the manufacturing version of something I preach constantly in the software world: don't hire for what you might need. Staff for what you can actually utilize today.

Rivian's Illinois plant was built to handle more. So before breaking ground on a second facility, they expanded the existing footprint by 1.1 million square feet and are targeting 215,000 total units of annual capacity across R1, R2, and their commercial vans.

There’s a parallel for all CFOs: before you greenlight that new product line, new geography, or new go-to-market motion… are you sure you've wrung everything out of the current setup?

I asked Claire what's scarier: over-building or under-building capacity. Her answer was nuanced, but leaned one way:

While she’s not saying "always under-build,” she’s saying your starting position should be finding creative ways to extract more from what exists (like adding automation, optimizing line rates, and increasing robotics) before committing massive capital to net-new capacity.

But there's a wrinkle. In automotive, there are efficiency cliffs. Claire told me the sweet spot for a mass-market vehicle line is typically 200,000 to 300,000 units. If you go below 150,000, you're losing a lot of the value on the investment. And if you go significantly above 300,000, you hit diminishing returns.

So Rivian has honed in on a planning range. It’s directionally helpful, but also not a pinpoint estimate, which means you have to underwrite your capacity bets with conviction about your market opportunity… while knowing that if you're wrong in either direction, it really hurts.

For those of us in the software world, the analog is server infrastructure, headcount planning, and sales capacity. Under-invest and you choke off growth. Over-invest and you're burning cash that takes years to earn back (or might force you into raising additional funding off your back foot).

Claire also shared something about headcount planning that made me see things differently. Rivian doesn't plan headcount around a single "atomic unit" the way a SaaS company might plan around an account executive. They plan around the product roadmap, and work backwards to staff accordingly.

And they layer in a (big) constraint:

Then they use that affordability constraint to make trade-offs on timing, sequencing, and where to augment with contract labor to smooth out spikes.

It's all multidimensional in a way that makes annual headcount planning at a SaaS startup look like a tik-tac-toe.

Here's the nugget of wisdom that really put my brain in a pretzel.

You'd think adding a brand new product (like the R2) would push profitability further out. It calls for a new supply chain, new production ramp, and brings new inefficiencies. Launching a new car line costs you a lot of money before you make money (or at least I thought).

Claire explained why the opposite is true:

By adding R2 into the same manufacturing facility, Rivian spreads its enormous fixed costs (plant, equipment, depreciation) across way more units. The R1 and commercial van lines get cheaper on a per-unit basis just because R2 is now sharing the overhead.

Claire calls this fixed-cost leverage.

If you've already built the infrastructure (the data platform, the security stack, the customer success team, the billing system) launching a new product on top of that shared base doesn't require you to rebuild all of it. The marginal cost of the new product is lower, and the blended cost of the entire portfolio goes down.

As a bonus, R2 is being dropped into a running, scaled manufacturing facility. That means it ramps faster than it would in a standalone greenfield site. The institutional muscle memory of how to build vehicles already exists, the workforce is trained, and the suppliers are established.

That's the equivalent of launching a new product line at a SaaS company where the sales team already knows the buyer and the CS team already has the relationships. You're not starting from zero.

And to prove the point, Rivian's CashOpEx in 2024 was the same as it was in 2022, but they 3x'd revenue over that period.

For any CFO thinking about capacity planning (whether that's servers, headcount, or a four-million-square-foot plant):

Claire was generous with her time and I walked away with a completely different appreciation for what it means to be a CFO of a company that builds physical things. In the software world, we can ship a beta and iterate. In Claire's world, you can't ship a car that's like 90% done.

As she put it: when you're a chef, you can swap an ingredient and still make a great meal. When you're building a car, you need every single part to be there.

I talk with Rivian CFO Claire McDonough about financing one of the most capital-intensive businesses in the world. We cover:

Hoping there are more EV tax credits to go around,

CJ