On the a16z piece on the rise of the engineering / builder CFO. There are some great ideas there. But some points are viewed through a stunningly narrow lens. I find that a little unambitious given the size of the opportunity in front of us. First, two things they definitely get right: 1) AI’s ability to make sense of messy data is a HUGE inflection moment for finance. Every bit of finance transformation advice always starts with “sort out your data”. Any idea how fucking hard that is? Teams have been grinding away on transformation and data hygiene projects for years. They often move slower than the businesses they’re supposed to support. AI unlocks the potential to destroy the biggest barrier to modernization of the CFO’s office. 2) Finance engineering is real, and should be thoughts of as a new core capability any ambitious finance team needs. The days of the systems accountants being the most tech-savvy people in the finance team, are thankfully ending. Real builders at the intersection of finance and engineering are worth their weight in gold and about to get more valuable. This is exciting, especially for smart young ambitious pros. Beyond that, much of the rest of the piece only makes sense through a software industry lens. It has the energy of tech people talking to other tech people about tech companies who sell to other tech companies. (Note - for the record, I’ve worked with some of the a16z portfolio companies mentioned in the piece, who definitely have the capability and appetite to serve outside tech.) For example, the piece talks about finance going from 5% of company headcount as standard, to 2%. In my last role, we had 1.5% of headcount in finance. We weren't especially efficient, it's just a function of the headcount dynamic for the industry I was in. So, unfortunately this falls into the trap that so much commentary on finance from the software industry does. It forgets that most companies are not tech companies. And therefore, most CFOs are not tech CFOs. Software finance is characterized by being scalable with a supersonically simple P&L structure (which goes to the finance capability you need.) I think it was Ben Horowitz himself who said that the biggest difference with hardware is "you need a much better CFO than in software." Yeah ... that’s because the economics of a hardware business looks more like conventional manufacturing than it does software. Now some of this has changed lately of course, as software consumes more AI, it takes on more of AI’s economics too; token costs, compute versus human decisions, usage-based pricing. More complicated margins and planning dynamics than before. But take a petrochemicals manufacturer, as example. They are dealing with feedstock prices, energy costs, production yields, plant utilization and selling prices moving together, by the day. And always has done. Now imagine doing that across a few hundred or a few thousand SKUs. Think about the dimensional complexity that brings. That makes the story much harder to find, to tell, and then ultimately to fund. Imagine what they could do with better tooling. So when the piece says (for example) finance can look forward to a "bigger mandate”, including pricing, capital allocation and people-versus-software decisions ... I’m a bit stunned. Any half-capable CFO I know would either be in the room when decisions of that financial magnitude were made. Or beating down the door to get in. And then there’s the framing of successive CFO eras: accounting, fundraising, FP&A / business planning, and now building. They acknowledge that the dimensions of a great CFO have endured. But the framing still jars. These aren’t things that most businesses have had the luxury of choosing between. I worked for CFOs twenty years ago who were incredible at controls, fundraising and business planning all at the same time. So, when I read things like that, I don’t see an evolution… it feels more like catching up?! Now … where plenty of finance teams outside tech ARE lagging is in the software available to them. Take that multisite petrochemicals business. Give it four different legacy ERP systems and years of workarounds holding everything together. The finance team is still delivering what the business needs despite poor, inflexible software. Impossible implementation conditions. So it's much slower and more manual than they need and would like. This is where AI’s ability to work with messy data gets really exciting. Helping bridge those systems and workarounds could finally let teams make progress on problems that have defeated successive generations of transformation programs. There is a world of businesses out there hostage to the legacy ERP industrial complex. I hope the investors and builders powering the next generation of CFO Tech have those businesses front of mind. Because the TAM is far bigger than the one on their doorstep.