The SaaS mental model colonized finance. Despite being roughly four tenths of a percent of global output (40 bps for all you fancy financiers), it absorbs so much more of the conversation. It’s almost like it generates a gravitational pull.
Why? SaaS is so clean, predictable, and teachable that it gradually became the default frame for how we talk about building and running businesses (and I'm guilty as charged for contributing to this).
It’s also apparently dying, depending on who you talk to, but that’s not the point of this report.
The metrics that came out of it (ARR, NRR, CAC payback, Rule of 40) are useful. They're also built for a pretty specific type of company, one that sells software by the month, has roughly zero marginal cost, and doesn't need to know where its product is physically located at any given point in time.
I come from that world. I've spent most of my career thinking in those terms, writing about those metrics (shameless plug for mostlymetrics.com), and interviewing CFOs at the top of the SaaS game. And for a long time I didn't notice how much of the economy that excluded. It's easy to sequester yourself off into a specific echo chamber of X (I still call it Twitter).
Then I started talking to a different kind of CFO.
Scott Haralson used to be the CFO at Spirit Airlines (RIP) before recently moving to Hertz. He described his job to me like this:
"We have to sell every seat on every plane, over and over and over again."
A seat that flies empty is gone forever; it's a fleeting monetization moment. The economic logic of the business lives in the shape shifting gap between a fixed, massive cost structure and perishable inventory that reprices itself daily and expires at gate close.
Dustin Pederson, CFO of Locus Robotics, put an even finer point on the physical reality he lives: "Inventory is cash sitting on the shelf." Robots don't generate revenue when they're waiting in a warehouse for shipment or repair. They contribute to the bottom line when they're deployed in a customer's facility, picking items off shelves, and actively running. Every dollar tied up between manufacturing and deployment is a stranded dollar.
And when it comes to AI chips, the playbook is being written on the fly. Look no further than CoreWeave for inspiration. Nitin Agrawal, the company’s CFO, told me his accounting team is "forming precedents, not following them." They're writing financial structures for infrastructure contracts that have no template in any accounting standard anywhere (as much as I'd like to sell these templates to paid readers on my website).
SaaS was once all the rage. But what I've realized over the last year interviewing CFOs who operate “in the real economy” is we are entering the era of capital intensive companies. Chip makers, humanoid robot builders, and a new wave of defense companies are defining this moment in the macro. AI is making stuff you can touch cool again, which complicates how we think about the CFO's role as the chief resource allocator.
I was inspired to write this report at this very moment because something is happening. It doesn't have a name, but you can see it in the capital flows and the policy debates and the types of companies commanding the attention of the largest pools of money in the world. Data centers are going up across the American Southwest at a pace that hasn't been seen since the Eisenhower interstate highway system. Semiconductor fabs are being built on timelines measured in years and budgets measured in the tens of billions. Automotive companies are designing cars that can go 600 miles on a charge. The physical world, supercharged by AI, is having a moment.
And yet, the CFOs sitting inside these companies are doing a job that most finance content has never described.
So I got curious. I've sat down with ten CFOs from businesses that operate squarely in this world asked them what their finance function looks like and how they're solving their biggest problems. My hope is this serves as a guide for the next wave of capital intensive companies.
A note on what this isn't: this report isn't an argument that SaaS finance is wrong, or that ARR is a bad metric, or that anything you learned in the standard tech toolkit doesn't apply. It's that the standard toolkit was built for one type of business, and a huge and growing part of the economy runs on a completely different operating system. And now many of those business models are converging. This report is written to make you sharper on operating companies that touch the real world.
Hoping your supply chain is healthy AF.
CJ
