📬 Solid ground: Hiring a CFO to build the machine

Most finance teams are running AI on infrastructure built before the iPhone existed.

Campfire was built for the AI era from the ground up. The architecture, the data model, the agents; all of it designed with AI as the operating model, not retrofitted around it.

And that distinction isn't cosmetic. It shows up every month end, with a 5x faster close.

See what AI-native actually means 👉

✍️ If you’ve got an awkward CFO-level problem in your day job, and you want my help… send me your questions anonymously, and you might just see yourself in next week’s Mailbag. Submit using the button below:

Lots of fun stuff on tap today... here’s what’s up:

Now, let’s get into it.

Growth Entrepreneur from Tennessee, USA

Long time reader, first time caller.

I have built my career taking businesses from roughly $1m to $10m of EBIT. Most of my success has come from growth: Finding opportunities, winning customers, entering new markets, and building teams. In several cases, we grew quickly enough that growth covered up mistakes that would have become obvious in a slower-growing business. 

I appreciate operational and financial discipline and know how important it is, but I am realistic enough to know it is not where I create the most value. Your newsletter has convinced me that many of the problems growth CEOs create do not show up until years later.

I know how to hire sales leaders and operators. I have much less confidence in my ability to identify and hire a truly great CFO. If you were sitting in my seat, what would you look for? How can a CEO distinguish between someone who can close the books and someone who can genuinely help build a scalable business?

My fear is hiring someone who can report the numbers but cannot help build the machine.

Thanks. Great question. And it sounds like your diagnosis is spot on.

The short answer is: you need to find someone who has done it before. 

First, sector specialization matters. You need someone who understands your business model, because “the machine” depends on the business model. A great CFO in one business can be a very average CFO in another if the economic model is different. SaaS, services, manufacturing, distribution, marketplaces, agencies, all have different pressure points. Different margins. Different working capital behavior. Different capital allocation dynamics.

People can excel across more than one, and maybe even all of them, but why not pick a focused specialist if you can?

Second, you need someone with experience at your specific journey and scaling points. That means three things: experience working at your current scale, experience working at your target scale, and, most importantly, experience going from one to the other.

It is easy to assume that someone who has managed a $10m EBIT business in your sector can help you build one from $1m of EBIT. Maybe. But operating at that level and building to that level are two very different things.

This is one of the most common mistakes ambitious founders make when hiring CFOs. They hire someone from a bigger business and assume the capability will trickle down. Sometimes it does. Often it does not. Big-company finance can be excellent, but it can also create people who are very good at operating inside a machine someone else built.

You need a builder.

That means someone who can build the financial spine of the business while the business is moving. Planning rhythm, cash discipline, controls, reporting, capital allocation, KPI architecture, decision support, board communication, and the ability to tell you uncomfortable truths.

So how do you tell the difference between a book-closer and a business-builder?

Ask to see the scars.

Ask them about their previous experience building a scaling finance team, ask what broke. What they would do differently. A real builder will have a list as long as their arm.

Where should you look? Start with your sector network, then use a CFO-specialist recruiter who actually understands scale-up finance. Not a generic executive search firm that will spray impressive CVs at you and hope one sticks to the wall. Your requirement is specific, and specific is harder to find. Harder to find usually means a proper search.

I have regretted not paying recruitment fees many more times than I have regretted paying them. Good recruiters are worth the fee.

TLDR: Find a trusted recruiter to hire a CFO who has done it before. They have gone from A to B in your industry and have the scars to prove it.

Hi Secret CFO,

I work in the finance team of a mature B2B SaaS company, built through multiple integrated M&A acquisitions, with a strong, highly detail-oriented CEO. Their reviews are thorough and often insightful, regularly surfacing errors and valuable observations.

However, in practice this level of detail creates significant back-and-forth, making it difficult to establish a consistent operating rhythm.

Two examples:

In a previous founder-led business, finance played a stronger role in setting top-down expectations and framing targets, which created more clarity and pace in planning.

How would you expect a CFO/finance function to introduce a more structured operating rhythm—with clearly defined KPIs, consistent review cadences, and clear accountability—while constructively channeling a detail-oriented CEO away from ad hoc deep dives in a business of this complexity?

Thanks!

Thanks for the question.

I’m going to be slightly annoying here.

If your CEO is surfacing “errors and valuable observations” from ad hoc deep dives, then your problem is trust, not cadence.

That may be unfair. It may be irritating. It may make you want to gently bang your head against the desk. But if they keep finding real issues, they are not going to stop looking until looking stops being useful.

So start there: What errors are they finding? Where are they coming from? How do you stop them repeating?

A detail-oriented CEO needs to feel a warm fuzzy feeling from the numbers before they back off. Right now, it sounds like they do not. And from your description, they may be justified.

I would not start by trying to “channel” the CEO away from detail. I would start by making the detail boring and a waste of their time.

Track the recurring issues. Group them. Root-cause them. Fix the process. Then show the CEO what has changed. That earns the right to change the rhythm.

Then, once confidence is improving, separate three things that are probably getting mashed together today: 

And the more you can get these into separate forums, the better. But again, that will be earned, not negotiated. This might sound a little basic, but it sounds like that’s where you are.

On planning, I think you have diagnosed the issue well. You have bottom-up engagement, but no top-down anchor. So VPs submit what they can live with. Then the horse-trading begins.

It is kind of like budget karaoke.

Finance needs to create tension between two things: what performance the business needs, and what performance the business is prepared to commit to.

As a start point, think of it like a very simple two-person long-range planning process. Sit down with the CEO before the planning process starts and agree on a simple top-down frame. Revenue. Gross margin. OpEx. EBITDA. Cash by business unit. Put it on a single tab in a spreadsheet.

Sometimes, you just have to start with something very crude and build from there. It is the conversations that make a planning process, not a great model, or even a sophisticated project plan.

So, I would play this as “one for them, one for you.” For them: fix the recurring detail issues until the deep dives become less useful. For you: introduce a simple top-down planning anchor and a cleaner review cadence.

The obstacle is the way on this one, my friend. Good luck.

TLDR: You cannot cadence your way out of low trust. Fix the details, then earn the right to lead planning.

Norwegian Moose from Europe

Hi Secret CFO, 

Some time ago you proposed dividing companies into three groups based on their complexity from a finance/accounting/FP&A perspective. 

I’m curious to know where on this scale you would place AI providers like Anthropic or OpenAI, taking into account both their current business model and their future trajectory. How complex is financial management in companies like these?

Hey Norwegian Moose, thanks for the question.

I think you are referring to this post. That was where I set out my view that finance complexity is not really driven by scale. It is driven moreby what the business sells, how it sells it, and how much operational complexity sits behind the revenue line.

I stand by that. I won’t repeat the full argument here.

Crudely, I split CFO roles into three levels based on their business model:

Then you multiply that by product complexity. How many SKUs, revenue streams, business models, geographies, pricing structures, customer types, and fulfillment models you are dealing with.

The core point is this: the more inputs and outputs you have, the more exposed you are to accounting complexity, mix effects, margin opacity, and operational noise.

That makes it harder to understand the true story of the business… which makes it harder to tell the story… which makes it harder to fund the story. And that whole complexity snowball eventually rolls downhill and lands on the desk of the finance team.

So where do Anthropic and OpenAI sit?

Short answer: Level four.

Sarah Friar and Krishna Rao have extremely difficult CFO jobs (how well they are doing those jobs is a different question).

The scale, security, governance, and public interest alone would be enough. Add in unit economics that are still forming, a business model that investors are trying to underwrite in real time at Supernova scale, and a market where the cost curve is still moving around like a drunk octopus, and you already have a nasty finance job.

But then look at the operation.

These businesses are not just selling software in the clean old SaaS sense. They are consuming and contracting for enormous compute capacity. They are tied into physical infrastructure. They are dealing with capacity planning, power constraints, model training costs, inference costs, customer demand loads, and uncertain asset lives. 

Oh… and they are committing more capital to physical hardware (data centers) than any conventional hardware business.

That is a very different animal from “we sell seats at 85% gross margin and occasionally argue about whether CSM sits in COGS.”

In some ways, an AI model provider is part software business, part infrastructure business, part trading business, and part industrial science project. Compute is not inventory in the traditional sense, but capacity has real economic perishability. If you buy too little, you choke demand or degrade service. If you buy too much, you carry expensive idle capacity. And if demand spikes in the wrong place at the wrong time, you are into short-term decisions that look a lot more like capacity trading than clean SaaS planning.

There is real COGS. There is real gross margin pressure. There is real capital intensity. There is real unit economic uncertainty.

So yes, Sarah Friar, Krishna Rao, et al. have incredibly difficult jobs (especially if you add in that they’re both working for bosses with a propensity for saying dumb shit out loud).

Quick side note as you are a self titled ‘Norwegian Moose’. It has been a joy watching the Norwegian fans at the World Cup, and the Viking Row. I’m old enough to remember the last Norwegian World Cup team in 1998, when Ronny Johnsen locked up Ronaldo (yes, 9) to beat that great Brazil team. I’ll be tuning in to watch the match vs Côte d’Ivoire later today.

TLDR: Frontier AI CFO roles are about as hard as the job gets in 2026.

A few of the biggest stories that every CFO is paying close attention to. This is the section you might not want to see your name in.

Nike raids Pfizer for its CFO (with a $7.3m ‘new hire’ signing cheque…)

Going from selling drugs to selling sneakers…

A CFO just out-earned half the CEOs in America

Tim McHugh of Welltower out-earned Tesla's CFO, a hell of a lot of CEOs, and his own prior year ($7.2m) by a mile. It's a one-time 10-year grant though, so we’ll have to put a 9 year deduction on his Adjusted LTM Proforma Runrate Comp bridge…

CFOs think the world's getting riskier, so naturally they're betting bigger

Well… aren’t we a conservative bunch.

ICYMI, here are some of my favorite finance/business social media posts from this week.

People joke about this kind of thing, but honestly… this kind of grassroots research is how you uncover the insights other people don’t! All the best operators I’ve worked with always did this kind of mystery shopping:

The hot debate…

Let me know what you thought of today’s Mailbag. Just hit reply… I read every message.

Last weekend’s Playbook showed how to build a growth-friendly finance team.

Disclaimer: I am not your accountant, tax advisor, lawyer, CFO, director, or friend. Well, maybe I’m your friend, but I am not any of those other things. Everything I publish represents my opinions only, not advice. Running the finances for a company is serious business, and you should take the proper advice you need