Reforecasting: when to update your numbers, and how

A budget is a plan made before the year happens. A forecast is your best current guess of how the year will actually finish, updated as real numbers come in. The two are not rivals. The budget is the line you committed to. The forecast is where you now think you will land. Reforecasting is the act of updating that guess when the world has clearly changed.

Founders tend to make one of two mistakes here. The first is never updating: clinging to a January budget through October even though a big customer churned in March, so every report compares today against a plan everyone quietly knows is dead. The second is updating constantly: rebuilding the whole model every time a number wiggles, until nobody remembers what was originally committed to or whether things are better or worse than expected.

The discipline that avoids both is simple. Reforecast on a regular cadence, plus whenever something material happens, change only what actually changed, and always keep the original budget intact so you can see the drift. Here is how that works in practice.

When to reforecast

There are two triggers, and you want both. The first is a regular cadence, quarterly for most small companies. Every three months you take a fresh look at the rest of the year given what you now know. This keeps the forecast from going stale and turns reforecasting into a calm routine rather than a fire drill.

The second trigger is a material change, something big enough that your plan is now plainly wrong. Use a threshold so this is not a judgment call every week.

• A large customer is won or lost, say more than 10 percent of revenue
• A funding round closes, or clearly will not on the timeline you assumed
• A new cost line appears, like a senior hire or an office, that was not planned
• Pricing changes, yours or a key supplier's, by a meaningful amount
• Actual results miss the plan by more than 15 to 20 percent for two months running

How to update without redoing everything

The mistake that makes reforecasting feel like a chore is rebuilding the model from scratch each time. You do not need to. A good forecast is built so that you can change the few assumptions that actually moved and let the rest flow through. Start from your current forecast, not a blank page, and ask one question: what do I now know that I did not before?

Say you lost a customer worth $8,000 a month in March. You do not rethink your rent, your payroll structure, or your software costs, none of which changed. You lower the revenue line by $8,000 a month going forward, let that ripple into your cash and runway, and stop. The whole update might take an hour. Reforecasting is surgical: touch the assumptions that changed, leave the ones that did not, and trust the model to do the arithmetic.

Keep the budget so you can see the drift

The most important rule is also the most often broken: never overwrite the original budget. The moment you save the new forecast on top of the old plan, you lose the ability to answer the question that matters most, which is how far have we drifted from what we committed to, and why. Keep the January budget frozen as its own column, untouched all year.

With the budget preserved, every report can show three things side by side: what you planned, what you now forecast, and what actually happened. That comparison is where the learning lives. If you budgeted $200,000 in Q2 revenue, now forecast $170,000, and actually hit $165,000, you can see both that you set the plan too high and that your reforecast was close. Over a year, that feedback makes you a noticeably better planner, which is the quiet long-term payoff of doing this well.

A simple rhythm

Put it together and the rhythm is light. Set the budget once, at the start of the year, and freeze it. Reforecast every quarter as a matter of routine, and reforecast off-cycle whenever something material crosses your threshold. Each time, start from the last forecast, change only what genuinely changed, and keep the original budget visible alongside.

Done this way, reforecasting stops being a dreaded rebuild and becomes a short, honest check-in with reality. Your numbers stay believable, your team stops arguing against a plan everyone knows is outdated, and you carry the clear-eyed view, plan versus forecast versus actual, that lets you decide what to do next while there is still time to do it.