It is wonderfully surreal to be living through the moment of the AI revolution. The productivity unlocks are so potentially massive in scale and impact and economic disruption, that we'd have to compare the world pre- and post-Industrial Revolution to understand just how massive this disruption could be.

And that means that yes, even the office of the CFO - historically notorious Luddites - are going to see some incredible changes. I got a bit excited thinking through this and thought I'd jot down some thoughts. In no particular order, here's a non-exhaustive list:

‍Streamlined Decisioning: CFOs will so often have a sense of the impact of a decision on the numbers, but will lean on their FP&A team to be thorough and run the appropriate analysis. To do this takes time, which varies depending on the availability and cleanliness of the necessary data and the necessary analyst. Productivity gains for a beleaguered analyst therefore allow for quicker and better informed decisions. At an end state, perhaps a CFO could even type in a prompt and have an analysis on the fly... one can dream...‍

Fundraising / M&A Due Diligence: Those lists of DD questions become more manageable when your FP&A team has a 'helpdesk.' FTE augmentation FTW here because AI doesn't get square-eyed after staring at fields of data for hours and make exhausted mistakes.

‍Accounting: FTE augmentation, better precision, quicker time to close and maybe even (dream) a *less painful annual audit*. If you take this through to its logical conclusion, think about how much smoother it would be to file for IPO if numbers are truly in order, and not burdened by 'accounting debt' and the inevitable cleanup.

‍Forecast Accuracy: When your limited FP&A resources are more productive, they can go from [re-re-]running the operating model to actually *looking at it* to ensure it makes sense. And speaking to colleagues across teams (RevOps, Legal, Product, Marketing, Sales, CX... truly anyone) to make sure there isn't some real-world obstacle - or tailwind - that doesn’t show up in the numbers.

‍Thoughtfulness at scale is the ultimate CFO pipe dream. *ALSO* I wonder if this helps companies with consumption-based pricing, and being able to hook into those usage data sets... what a morass of data...

‍RevOps: This dovetails with forecast accuracy, but it's a bit different. As a CFO, most of the metrics we look at describe present or past state, which are critical to understand and articulate, but the *meaty stuff* is where you're going on top line. *Accuracy in your funnel metrics and ability to anticipate your sales is so damn critical it feels silly to even write it.* If you're able to query against your CRM data (I know, I know, but we're dreaming) and augment your RevOps team's ability to give snapshots into what the F is happening in pipeline, you're a hero. By the way, same goes for customer health and renewals... again, we're dreaming.

‍*Problem*: The inevitable big old BUT. Data cleanliness. Hygiene. Training your LLM to understand the schema of your data (especially in your CRM). Having limited resources already, and then needing to pivot them to figuring out how to implement / train this LLM instead of doing the multitude of work that needs to be done (akin to the typical 'infrastructure' vs 'product roadmap' debate for eng teams). But I see this as a good problem to have, and while painful, an ultimately solvable problem.

For those building in the space, one other thought - it can be notoriously difficult to sell into CFOs. We're a prickly, skeptical bunch, but if you're solving a real problem in a thoughtful way (and don't pretend there won't be time needed to train and set up the LLM), then they're also quite a loyal customer.

What an awesome moment to be a CFO. Would love to hear how others are also thinking about this as well.

‍Michaela Lehr is a tech CFO with over a decade of experience working in startup finance teams. Most recently, she was CFO at ActionIQ, which was acquired in December 2024.