The Story Behind the Story

How Propeller built embedded finance teams years before the market called them “fractional”, and the PropellerOS platform behind them

By Chris Fenster, Founder & Executive Chairman

Propeller was seven years old before I ever heard the word “fractional.” We started in 2008, doing something that didn't have a name: embedding full finance teams inside companies that couldn't yet justify building their own. Around 2015, the market finally coined a term for firms that looked like ours, and it described almost nothing about what we actually did.

Let me say something nice about part-time help first, because the idea deserves it. Renting expertise you can't afford to employ is one of the oldest arrangements in business, and it works. A founder who needs ten hours a week of controller time should buy ten hours of controller time. I have no issue with the arrangement, and plenty of good people make their living this way.

The issue is what the last decade did to it. “Fractional” took a sensible arrangement and turned it into a category, and then the category did what categories do: it optimized for what's easy to sell instead of what's necessary to win. Marketplaces of interchangeable operators. Rate sheets. A race to the bottom dressed up as a movement. 

About a year ago the ads started finding me: not for fractional services, but for services sold to fractional practitioners: lead generation, coaching, practice-in-a-box. There's precedent for that in B2B, and some of it is legitimate. But when the surest money in a category is made selling to the sellers, something has tipped. The customer at the end of the chain has stopped being the point. And underneath the branding, the same structural ceiling that part-time help has always had: a person splitting time across five companies is still just a person. The hours don't compound. The knowledge walks out the door when they do. There's no network underneath them, no accumulated data, no pattern recognition that gets sharper with every company served. And nobody stays long enough to be accountable for whether the advice was any good.

That last line is nearly word-for-word from the founder note I signed a week ago on our new website. If you've visited propellerindustries.com this month, you've seen that we launched a new site, a new brand identity, and for the first time, put a public name on the platform we run the firm on: PropellerOS. The site is the story we're finally telling; this is the story behind that story.

A website is a strange thing to get emotional about, and I didn't expect to. But this one is the visible edge of a decision that has been building for a long time. 

In 2008, the founding choice was to be embedded rather than episodic, and the unglamorous part turned out to be the whole ballgame: we kept the accounting, but not for the reasons you probably think. The daily transaction work is what let us stay inside a business between the big CFO moments, so we never had to leave and never had to re-learn a company when the stakes went up. Other advice-giving businesses arrived for the sexy fundraise, the board cycle, or the diligence sprint, and then left. We stayed. Staying turned out to be the strategy.

I mean strategy in the deliberate sense, so let me be careful here. Founders retrofit foresight onto their own histories all the time, and I'd rather show receipts than ask you to trust my memory. I had seen the embedded model work before Propeller existed, at the Tides Center in San Francisco, where I ran finance from 2000 to 2002 for a little subsidiary they formed to help political non-profits. I started Propeller convinced that the digitization of finance was about to make that model scalable in a way it had never been. 

While writing this piece, I dug out the first client deck I ever made, dated March 2009, a year after I started the firm. Slide six asks a prospective client, “How will your finance operation convert data into actionable intelligence?” That question has been the whole game ever since, and it's sitting on a slide older than the iPad. By 2011 we had sharpened it into an explicit thesis: data enablement was the highest-leverage move available to a firm like ours, because the durable asset wasn't hours billed; it was the accumulated record of how hundreds of companies actually operate, standardized well enough to learn from.

The record shows us spending real money on that thesis long before it was fashionable. In 2019 we acquired Bison Analytics to give the data thesis actual infrastructure instead of a shelf of spreadsheets; this was the earliest direct ancestor of what we now call PropellerOS. In 2020, we partnered with Newlight Partners for a growth equity investment, made specifically so we could push further into tech-enabled finance and keep building the data platform a firm like ours should have. 

In 2024, I recruited Raymond Nieuwenhuizen, who had spent his career leading transformation work inside enterprises, to take over as CEO and rebuild the firm while it was still working. That’s the opposite of the usual sequence, where companies transform only when a crisis forces them to, and I won't pretend the decision was painless (though that’s a story for a different day).

In late 2025, the current Propeller Data Lake and the Propeller Intelligence Engine went live, laying the foundation to turn eighteen years of client experience into infrastructure. In 2026, the front door finally caught up to the house.

So: were we prescient or lucky? Both. You can be prescient and lucky at the same time. We didn’t foresee AI, not in the form it eventually took; nobody in our world did. What we foresaw was that finance would come to run on connected, trustworthy data, and that whoever held the deepest and cleanest record would hold the leverage. The deck we raised our 2020 growth round on says so in as many words: trust earns access to client data, which we warehouse and benchmark into analytics and portfolio intelligence. 

The same deck contains something else, on a slide describing where the platform was headed: the name PropellerOS, six years before we said it in public. AI arrived as an accelerant poured on a thesis that was already a decade old.

The founder story on our new site says we'd been preparing for AI without knowing it, and that's true. The fuller truth is that we knew exactly what we were preparing for; we just didn't know it would eventually have a name. I don't bring this up for credit: I bring it up because it bears directly on the decision founders are making right now. In the agentic era, agents are only as good as the data underneath them. The firms worth betting on are the ones whose data discipline and financial intelligence predate the hype, and a record like ours can’t be assembled retroactively.

That's what PropellerOS is. Not software we sell, but the system we run on. Every engagement operates on it; it comes standard because the engagement doesn't work without it. PropellerOS is how eighteen years of financial intelligence across 1,500+ companies shows up in every engagement we deliver. Your Propeller team doesn't just bring their own judgment to the table. They bring the pattern recognition of everyone who came before you.

Which brings me to the part of this story I care most about, because a launch is only interesting for what it lets you promise next. Here’s the deal we're now built to offer, and the standard I want you to hold us to.

Companies that invest in a partnership with us should get more output for less money than they could achieve assembling it themselves. They should get flexibility that no hiring plan can match: a team that changes shape as the company does, scaling up for the raise, reshaping after it, and supporting the leaders you eventually hire instead of competing with them. And they should get the best of both models that founders have always had to choose between: the continuity, context, and ownership of a great in-house team; with the leverage, resilience, and cross-company intelligence of an embedded one. The choice between “my people” and “outside help” was always a false one. It just took eighteen years of infrastructure to make it false in practice.

And here’s one more promise underneath the others: when a number leaves our hands for your board, your lender, or your buyer, a named human stands behind it. We call it the last mile of accountability, and it's the part of finance that no amount of automation moves off a person.

Part-time help was always a good idea, but it was never a way to build. Building takes staying, and staying is the one thing we've been doing longer than anyone. We took apart what we'd spent eighteen years assembling because we decided winning mattered more than being comfortable. Now we do what we've always done: stay long enough to find out whether we got it right.