New

Peaky raised €1.1M, led by JK Invest

→

All articles

Company

/

Written by Ser Cappelle (CEO/CFO & Co-Founder at Peaky)

Why we founded Peaky.

“For fifteen years I was the last to know. And the first to explain.”
Ser Cappelle, CEO / CFO and co-founder
“For fifteen years I was the last to know. And the first to explain.”
Ser Cappelle, CEO / CFO and co-founder

I spent most of my career as a fractional CFO.

Different companies, different industries, different stages. I worked with founders trying to figure out their first thirty hires, CEOs trying to get to Series A, leadership teams trying to scale past Series B. I sat in operating reviews, board meetings, fundraising rooms. I closed books, built forecasts, wrote investor updates. I learned more about how finance actually works at scaling companies in those years than I had in any course or textbook.

What I learned, more than anything else, was that the finance function at most scaling companies is broken in the same specific way, and almost nobody is fixing it.

This post is about what I saw, what I came to understand about why it kept happening, and how that observation eventually became Peaky.

The pattern

I would arrive at a new company. Different industry. Different team. Different stack. I would expect to find a different version of the finance challenge.

I almost always found the same version.

The plan was a quarterly artifact. The forecast got refreshed at the end of each cycle. The actuals came in after the close and produced a variance that the team would try to explain. The CEO and the CFO would look at slightly different numbers. The CRO would have his own version of the truth. Each operating review would start with a few minutes of reconciliation before anyone could talk about strategy.

In between the cycles, the business was changing constantly. Decisions were being made. Hires were being pulled forward. Campaigns were being launched. Pricing was shifting. None of it was being reflected in the plan until the next refresh.

By the time the variance showed up in the close, the events that caused it had happened weeks earlier. The conversation about what to do about it was already too late to be a decision and only just early enough to be a reaction.

This was not one company. This was every company.

Why it kept happening

For a while, I assumed it was a tools problem. Maybe these companies just had not bought the right FP&A platform. Maybe a better tool would close the gap.

Some of my clients did buy better tools. Pigment. Anaplan. Mosaic. Various others. The forecasting got cleaner. The collaboration got better. The board packs looked more polished.

The fundamental gap did not close.

What I realised eventually was that the tools were optimising the wrong thing. They were making the batch cadence faster. Faster close. Faster forecast refresh. Faster board pack. The cadence itself was the problem, not the speed at which each step happened.

The business was changing daily. The plan was refreshing monthly. Faster monthly refresh did not change the fact that the plan was always weeks behind the business.

This was not a tools problem. It was an architecture problem. The traditional FP&A approach was built for a slower world, and at scale-up speed, it had stopped fitting.

What was missing

What was missing, in every company I worked with, was a layer that connected the financial plan to the operational systems where reality changed first.

Not a dashboard. Dashboards reported on data. They did not change the cadence of planning.

Not a BI tool. BI tools sliced and visualised, but they did not maintain a forecast.

Not a faster spreadsheet. The spreadsheet was the problem, not the solution.

What was needed was a layer that read the operational data continuously, mapped it to the drivers behind the financial plan, watched for drift, and surfaced what was changing the moment it changed. A live system that kept the plan honest.

That layer did not exist. I knew, because I went looking for it every time a client asked me to fix the finance function. The closest things I found were either too rigid, too slow, or too expensive to deploy.

This was, eventually, the gap that became Peaky.

The moment it became a company

There was a specific moment when the idea of building it crystallised.

I was working with a Series B company. We had just finished an end-of-month close. The variance was meaningful. A campaign that had triggered higher CAC. A hire that had been pulled forward. A pricing tier that had shipped without finance being looped in. Each of these had been visible in some operational system for at least three weeks. None of them had been in the financial plan.

We rebuilt the model. The runway changed by five months.

The CEO asked the question that every CEO asks in that conversation. How did we not see this sooner.

I went home that evening and could not stop thinking about it. The answer was: the information was always there. It was sitting in Stripe, the HR system, the ad platform, the CRM. It had been visible for weeks. It just had not surfaced in finance yet.

The structural failure was not human. It was that no one had built the layer that would have made this visible in real time.

I started sketching the architecture of that layer the same week. A few weeks later, I introduced the idea to Willem, John, and Guus. They had complementary skills and the same conviction that the gap was real and worth closing. Within a few months, the four of us were building.

What we built

Peaky is the layer that was missing.

The financial plan connects directly to the operational systems where reality changes. The CRM where pipeline shifts. The accounting platform where vendors get paid. The HR system where headcount grows. The billing platform where churn shows up.

A set of agents watches the drivers behind the plan continuously. When an assumption stops describing reality, they surface a signal. When the variance between forecast and actuals exceeds tolerance, they surface a signal. When the books grow new structure the model does not understand, they surface a signal. The plan stays current. The team sees what is changing the moment it changes.

This is the version of finance I wished I had access to when I was a fractional CFO. The version where the answer to “how did we not see this sooner” stops needing to be asked.

Why the four of us

Each of us brought something the others did not.

Willem had operated at scaling tech companies and understood the buyer’s perspective and the GTM motion. He knew what it would take to actually reach finance leaders and earn their trust.

John had design experience and a deep aesthetic instinct. He understood that a finance product that looked like every other finance product would never feel like the new category we were trying to claim.

Guus had the technical depth to build the agent architecture that the product required. He saw immediately how to design a system that would work continuously without breaking.

I had the customer empathy from years of being inside finance functions, and the conviction that the problem was structural rather than coincidental.

The four of us together had enough to start. We have spent the last year building, learning from our first customers, and shipping what we believed.

What is different now

The reason this product can exist now and could not have existed five years ago is that three things have converged.

The operational systems have become live. APIs exist. Data is available continuously rather than in batch exports.

AI agents have matured to the point where they can do the mechanical work of mapping, modelling, and monitoring without requiring constant human supervision.

The category of continuous financial planning has become coherent enough to position around. Customers can recognise what they are buying.

Five years ago, none of these were in place. Today, all three are. The opportunity to build the layer I was looking for as a fractional CFO is now available to anyone who is willing to do it.

We took the opportunity.

What comes next

The product we have today is the foundation. The agent set will keep growing. The integration depth will keep expanding. The audience we can serve will keep broadening.

The longer-term vision is that the financial plan stops being a deliverable that finance produces and starts being infrastructure that the whole company runs on. The CFO becomes the keeper of the layer, not the producer of the artifact. The role of finance becomes strategic in a way the category has been promising for decades without quite delivering.

We are not pretending we have built that future yet. We have built the foundation. The rest comes one quarter at a time, with the customers who are willing to build it with us.

If you have spent any of your career being the person who had to answer “how did we not see this sooner”, we built Peaky for you. You can start a trial at peaky.ai.

The financial infrastructure for scaling without surprises.

Previous

How to build a driver-based forecast in under a day.

Next

Continuous driver-based forecasting, built and maintained by an agent.

Keep exploring

Getting started

From your Excel plan to a live plan in a day.

Integrations

The systems Peaky connects to, and what each one brings in.

The Forecast Audit

Nine questions to find out whether your forecast is a report or a tool.

Peaky app icon

Sync or sink.

Ask AI for a summary of this website