Budget vs Actuals: Variance Analysis for Startups

Budget vs actuals is the practice of comparing what a company planned to earn and spend against what actually happened, line by line. The gap between the two numbers is called a variance, and the process of explaining those gaps is variance analysis. It is the fastest way to find out whether a financial plan is working.

Most founders build a budget once, usually before a fundraise, and then never hold the business accountable to it. The plan sits in a spreadsheet while spending drifts. A monthly budget vs actual report closes that loop. It tells you where reality diverged from the plan, by how much, and whether the divergence is a problem or a signal to update the plan itself.

This guide covers how to build a budget vs actual report, how to calculate variances, what variance thresholds actually deserve attention at each startup stage, and how to run the analysis without losing a week of your month to spreadsheet assembly.

What a budget vs actual report shows

A budget vs actual report places three or four columns side by side for every line on your income statement: the budgeted amount, the actual amount, the dollar variance, and the percentage variance. Most companies run it monthly for the month just ended and for the year to date.

The report only means something if the actuals are trustworthy. That is why budget variance analysis belongs inside a disciplined month-end close process rather than floating alongside it. If revenue is not fully recognized or a large invoice has not been recorded, the variance columns describe your bookkeeping lag, not your business. Close the books first, then compare.

Variances come in two flavors. A favorable variance means the result was better than planned: higher revenue, lower cost. An unfavorable variance means the opposite. The labels describe direction, not importance. A favorable variance can still be a warning, such as underspending on sales hiring because recruiting has stalled.

The discipline matters beyond internal decision making. Investors ask for budget vs actual comparisons in diligence, and a founder who can explain every material variance in plain language reads very differently from one who discovers the gaps live on the call. The habit feeds directly into board reporting that builds trust, because a board that sees plan-versus-actual every quarter learns that your numbers mean what they say.

How to calculate budget variance

The math is deliberately simple. For expense lines:

Dollar variance = Actual minus Budget. Percentage variance = (Actual minus Budget) divided by Budget.

An expense line that budgeted $10,000 and spent $12,000 has a $2,000 unfavorable variance, or 20 percent over plan. For revenue lines the interpretation flips: actuals above budget are favorable.

Two conventions keep reports readable. First, pick a sign convention and hold it everywhere, since flipping signs between revenue and expense sections is the most common source of confusion in founder-built spreadsheets. Second, show both the dollar and the percentage. A 40 percent variance on a $500 software line is trivia. A 4 percent variance on a $400,000 payroll line is real money.

The percentages only tell you what happened. The analysis is the sentence you write next to each material variance explaining why it happened and what you will do about it. "AWS ran 22 percent over budget because the new data pipeline shipped three weeks early; updating the infrastructure budget for Q3" is variance analysis. The number alone is not.

A worked example: seed-stage SaaS, one month

The illustration below shows a simplified monthly budget vs actual report for a hypothetical seed-stage SaaS startup with about $60,000 in monthly recurring revenue. The numbers are constructed for teaching, not drawn from benchmark data.

Read as a whole, the report tells a story no single line reveals. Revenue missed by 6 percent while marketing underspent by 30 percent, which suggests the pipeline was starved, not that the team saved money. Infrastructure ran 22 percent hot, which needs a one-line explanation before it compounds. And burn rate came in about $3,850 above plan, which quietly shortens runway if it repeats for two more months.

That last point is the reason this report exists. Small monthly variances compound into runway math. A startup that runs 6 percent over its burn budget every month for a year does not have a 6 percent problem, it has lost roughly three weeks of runway per quarter without ever making a single decision to spend more.

If you have not built the budget side of this comparison yet, start there. A budget is just a revenue forecast paired with planned spending by category, and the guide to building a financial model before you have revenue walks through the structure. A tool like our pro forma income statement generator can produce the baseline in a few minutes.

Variance thresholds by stage

Corporate FP&A teams investigate every variance above a fixed materiality threshold. Startups need a different posture, because early-stage budgets are guesses by design and the cost of investigating everything exceeds the value. A practical rule set, based on how we see companies actually operate at each stage:

Treat these as starting points to adjust, not audited benchmarks. The pattern behind them is what matters: as the company matures, the threshold for "worth explaining" drops, and the consequences of a miss shift from internal course correction to external credibility.

The dual threshold, a percentage and a dollar floor, keeps the exercise honest at every stage. Percentage alone drags you into investigating $80 variances on small lines. Dollars alone lets a 50 percent overrun on a growing category hide until it is large.

One more stage-specific habit separates clean operators from chaotic ones: when a variance reflects a permanent change rather than a one-time event, update the forecast. Budget vs actuals tells you the plan diverged from reality. A reforecast decides which one moves.

Why the report is usually late, and why that kills it

The standard failure mode is not bad analysis. It is timing. The books close on day ten, someone exports the P&L, pastes it into a spreadsheet next to the budget, fixes the rows that no longer line up because the chart of accounts changed, and circulates the report around day fifteen. By then the month being analyzed is half over again, and the findings are history rather than decisions.

The spreadsheet assembly step is also where quiet errors creep in. Manual paste operations against messy books produce variance reports that themselves need variance analysis, and the downstream cost of that compounds the same way bad bookkeeping compounds everywhere else. A budget vs actual report built on unreconciled actuals is worse than no report, because it manufactures false confidence.

The fix is structural, not motivational. Variance reporting has to be a byproduct of the close, not a project that starts after it. When the actuals are continuously reconciled and the budget lives next to the general ledger instead of in a separate file, the report is ready the moment the books are, and the meeting about it can happen in the first week of the month instead of the third.

This is how Futureproof handles it. Vic, our bookkeeping agent, keeps the books reconciled continuously instead of in a month-end sprint. Margo, our FP&A agent, tracks budget vs actuals against those live books all month, so the variance report is ready at close with material gaps already flagged, rather than assembled in a spreadsheet days later. The full team of six AI finance agents, covering bookkeeping, AR, AP, FP&A, RevOps, and investor reporting, is $1,000 per month flat. You can start with a free trial and see your first variance report against real numbers.

Frequently asked questions

How often should you run a budget vs actual report?

Monthly is the standard for startups, timed to the close of the books, with a year-to-date view alongside the single month. As of 2026, companies with continuously reconciled books increasingly monitor material variances weekly and reserve the monthly report for formal review. Quarterly-only reporting is too slow for a venture-stage company, because a bad quarter is three months of unexamined drift.

What percentage variance is significant?

There is no universal number, which is why fixed answers like "10 percent" mislead. Significance depends on stage, line size, and repetition. A useful startup default is to investigate expense lines that miss by more than 10 to 15 percent and more than a few thousand dollars, tightening toward 5 to 10 percent by Series A. Any variance that recurs for three consecutive months is significant regardless of size, because it means the budget is wrong or the spending is unmanaged.

What is the difference between budget vs actuals and a forecast?

The budget is the fixed plan approved at the start of the period, and actuals are the recorded results. Budget vs actuals compares those two. A forecast is the living estimate of where the period will land, updated as new information arrives. Mature teams track all three: actuals against budget for accountability, and forecast against budget to see problems before they become actuals.

What causes budget variances?

Four causes cover most cases: timing, where an expense or payment landed in a different month than planned; price, where the amount per unit changed, such as a vendor increase; volume, where activity levels changed, such as usage-based infrastructure costs rising with customers; and plan error, where the budget assumption was simply wrong. Labeling each material variance with one of these causes turns the report from a scorecard into a diagnostic.

Do pre-revenue startups need budget vs actual reports?

Yes, and it is arguably easier before revenue because the report is expense-only. A pre-revenue startup lives entirely on its spending plan, so a monthly comparison of planned versus actual burn is the single most direct check on runway assumptions. It also builds the reporting habit before investors start expecting it.