Written by: Peaky
Date: June 15th, 2026
Every CFO has had the same Tuesday morning. You open the forecast. You compare it against the actuals. And you realise that something has been quietly off-plan for weeks, possibly months. The number nobody flagged. The assumption nobody updated. The line item that crept past tolerance without anyone noticing.
We call this drift.
Drift is a plain English word. You drift off course. You drift apart. You drift away. In financial planning, drift describes the same thing applied to a company's numbers: the slow, silent movement of reality away from the plan that was supposed to describe it.
What makes drift dangerous is not that it happens. It is that it is invisible until it lands in the close.
Financial plans were built for a slower world. You set the plan once a year. You refresh the forecast monthly or quarterly. You compare actuals to plan during the close. The cadence assumes that the business does not change much between refresh cycles.
The business does change. Constantly. A senior hire gets brought forward. A campaign triples spend in a week. A pricing change ships without finance being looped in. A new revenue stream shows up in Stripe before anyone has updated the model. Each individual change is small. Together, over a quarter, they are the difference between a plan that is roughly right and a plan that has quietly become fiction.
Drift is not a bug in how you do finance. It is the default state of any finance function that refreshes slower than its business changes. Which is to say, every finance function.
Most teams treat drift as one undifferentiated problem. It is actually three distinct problems with three distinct shapes.
ASSUMPTION DRIFT
Happens when the inputs your plan was built on stop matching reality. You assumed 15% growth, the actuals are showing 8%. You assumed a churn rate of 1.5%, it is now 2.7%. You assumed a sales hire would close eight deals in their first quarter, they closed three. The plan is still doing math. The math is just being done on numbers that no longer describe the business.
Assumption drift is the most common kind. It is also the most expensive, because it compounds quietly. A growth assumption that drifts a single percentage point in January is barely visible. The same drift, compounded across twelve months, can shift cash runway by a full quarter.
PERFORMANCE DRIFT
Happens when the actual numbers come in differently from what the forecast predicted for the closed period. The forecast said 380K in net new MRR. The closed period actually delivered 290K. The gap is real, the cause needs to be diagnosed, and the impact will compound into the next period if nothing changes.
Performance drift is the kind every CFO sees. It is the variance column in the monthly board pack. The problem is rarely that it gets missed. The problem is that by the time the books close and the variance becomes visible, the cause is already weeks old.
STRUCTURE DRIFT
Happens when the business grows shape that the model does not yet understand. A new entity. A new product line with its own ledger accounts. A new revenue category that does not map cleanly to the existing chart of accounts. The model is correct about everything it knows. It just does not know about the new thing yet.
Structure drift is the least talked-about and the most invisible. By the time you notice it, your model has been wrong about the size of the business for as long as the new thing has existed.
The cost of drift is not the drift itself. The cost is the delay between drift starting and finance noticing.
At a Series A company, one missed signal typically costs between 50,000 and 100,000 euros. Not because the underlying change was catastrophic. Because finance found out about it weeks late, and by then the window to act had closed.
A campaign that triples CAC needs a budget conversation in week one. By the end of the month, the spend is already gone. A churn rate that creeps up half a point a month needs a CS intervention in month one. By month four, the cumulative impact has shifted runway by an entire quarter. A board meeting that gets the wrong number on the slide because pricing changed two weeks ago needs a different kind of conversation than the one that gets the right number.
The expensive part of drift is not the change. It is the lag.
The reason drift survives until the close is structural. Finance refreshes its view of reality in batches. Operations changes its underlying reality continuously. The gap between batches is the gap where drift hides.
For most companies the cadence looks like this. Operations changes daily. Accounting closes monthly. Forecast refreshes quarterly. Plan resets annually. Each layer is slower than the one below it. Each layer is reading reality through the lens of the layer below, with a lag built in.
This is not anyone's fault. It is how finance has always been done, and it worked when businesses changed at the same pace as the plan. At Series A and beyond, businesses no longer do.
You can keep doing finance the old way. Most companies do. They will keep finding out about drift the same way they always have: in the close, weeks after the fact, when the only conversation left to have is the apology.
Or you can move finance into the same pace as the rest of the business. Not as a one-time fix. As a permanent change in cadence.
That is what continuous financial planning means. And it is what we built Peaky to make possible.
If you want to see what catching drift early actually looks like, you can start a Peaky trial at peaky.ai. Most teams have their first drift signal surfaced within a day of connecting their first integration.
The plan will keep up with the business. Or the business will keep surprising you. Those are the two choices.
