Flux analysis is the practice of comparing each account on the financial statements to the same account in a prior period, then explaining every meaningful change. It runs at month-end close, after the books are locked. The comparison is actual against actual across time, and the output is a written reason for every material move.
The word flux is short for fluctuation. A flux review walks down the income statement and the balance sheet, line by line, and asks a single question of each row that moved: why. Revenue rose eight percent. Cost of goods sold jumped a quarter. Insurance doubled. Each answer is either a real business event or a mistake in the books, and the value of the exercise is that it forces the distinction before anyone else asks.
For a seed-stage company, flux is the seventh step of a disciplined month-end close, and it is the one that separates a close that produces numbers from a close that produces trustworthy numbers. The founder or first finance hire runs it as a sanity check. It catches the miscategorized transaction, the entry posted to the wrong month, the annual bill booked in full instead of spread across the year. Those errors are cheap to fix in the week you find them and expensive to explain in a board meeting.
Flux analysis versus budget vs actuals
These two reviews get confused because both produce a column of differences with dollar and percentage figures. They answer different questions.
Flux analysis compares one actual period to another actual period. This month against last month, this quarter against the prior quarter, this year against the year before. It measures how the business changed over time, and it does not care what anyone planned.
Budget vs actuals compares actual results to the plan. It measures whether the business is tracking the forecast it committed to. A line can be perfectly flat month over month, showing no flux at all, and still sit far off budget because the plan assumed growth that did not arrive.
Both belong in the close, and both are forms of variance analysis, which is the parent discipline of explaining any gap between two numbers. Flux owns the period-over-period view. Budget vs actuals owns the plan-versus-actual view. Run flux first, because it catches bookkeeping errors that would otherwise poison the budget comparison downstream.
A worked month-over-month flux example
The clearest way to see flux at work is on a real profit and loss statement. The table below shows a seed-stage SaaS company comparing March to February. Every figure is illustrative, chosen to show the kinds of moves a flux review surfaces.
Illustrative month-over-month flux, seed-stage SaaS P&L (figures for illustration only):
| Line | February | March | $ Flux | % Flux | Explanation |
|---|---|---|---|---|---|
| Subscription revenue | $182,000 | $198,000 | +$16,000 | +8.8% | Nine new logos, net of one churn; expected |
| Professional services | $24,000 | $12,000 | -$12,000 | -50.0% | One implementation slipped to April; timing |
| Cost of goods sold | $46,000 | $58,000 | +$12,000 | +26.1% | Annual infrastructure commit booked in full; should accrue across the year |
| Gross profit | $160,000 | $152,000 | -$8,000 | -5.0% | Driven by the COGS posting above, not a margin problem |
| Sales and marketing | $88,000 | $121,000 | +$33,000 | +37.5% | Two account executives started; one conference |
| Research and development | $140,000 | $142,000 | +$2,000 | +1.4% | Below threshold; no explanation required |
| General and administrative | $31,000 | $47,000 | +$16,000 | +51.6% | Full-year D&O insurance expensed at once; should amortize |
Two of these rows are not business events. They are misbookings. The cost of goods sold jump comes from an annual hosting commitment posted as a single expense rather than accrued across the twelve months it covers. The general and administrative spike is a full year of directors and officers insurance booked in one month instead of spread as a prepaid expense.
A founder who runs this flux catches both before the numbers reach an investor. The fix is a pair of correcting entries in the general ledger that move the annual costs into prepaid accounts and amortize them monthly. Without the review, March looks like a company whose gross margin is eroding and whose overhead is out of control, when neither is true.
Setting materiality thresholds for a startup
Flux only works if it ignores noise. A threshold defines what counts as material, so the review focuses on the handful of lines that moved for a reason and skips the rows that drifted by a few hundred dollars. The standard approach is a dual threshold: a percentage and a dollar floor, where a line gets flagged only when it clears both.
The percentage alone is not enough, because a tiny account can swing fifty percent on a trivial dollar amount. The dollar floor alone is not enough either, because a large account can hide a real problem inside a small percentage. Requiring both keeps the flagged list short and meaningful.
Startups should size these thresholds to their own scale, not to the enterprise defaults published by large accounting vendors. A seed-stage SaaS company running a few hundred thousand dollars of monthly spend can start with a rule of ten percent and $2,500. Any line that moves more than ten percent and more than $2,500 gets a written explanation. Everything else is presumed normal.
Sensitive accounts deserve a tighter floor regardless of size. Cash, deferred revenue, payroll, and accrued liabilities carry more risk of a costly error, so a lower dollar floor of $1,000 is reasonable for those. As the company grows through Series A and monthly spend climbs, the floors move up with it. The percentage tends to hold steady while the dollar figures scale.
Explanations that survive audit and diligence
The number on a flux report matters less than the sentence next to it. A weak explanation restates the math. A strong one names the driver, quantifies it, and ties it to something a reader can verify.
The finance office at Harvard, whose flux guidance is widely cited, holds that a good explanation answers what changed, why it changed, and by how much, in business terms rather than accounting terms. "Expenses increased" is not an explanation. "Travel expense rose $45,000, or 150 percent, because two people attended a conference that did not occur last period" is one, because an auditor can match it to receipts.
Three habits make a flux note hold up under diligence. Name the specific event, not the category, so "three enterprise contracts went live" beats "revenue grew." Attach the dollar impact of each driver, so a reviewer can reconcile the drivers to the total variance. And keep the note in the working papers, dated and attributed, so the explanation exists in writing before the question is asked rather than after.
This is what turns flux from a monthly chore into an asset during a raise. When an investor asks why March gross margin dipped, a founder who ran flux has the answer on file: an infrastructure true-up that was corrected, not a structural decline. The question that would have stalled a call becomes a two-sentence reply.
MoM, QoQ, and YoY: choosing the comparison window
Flux can compare any two periods, and each window shows something different. The three common cuts are month over month, quarter over quarter, and year over year.
Month over month is the close's quality gate. It catches posting errors and miscategorizations fast, while the source documents are still fresh and the correcting entry is easy. This is the version a startup runs every close, and it is where most misbookings surface.
Quarter over quarter smooths the monthly noise and suits leadership and board reporting, where a single month's timing swing matters less than the trend. Year over year removes seasonality entirely, which makes it the honest read on underlying growth for any company whose revenue moves with the calendar. A strong close runs month over month every period and layers quarter and year comparisons in for the board and the annual audit.
Where Futureproof fits
Flux is only as good as the periods it compares. If the books are three weeks behind, the "prior period" is a guess, and the review flags lag instead of business events. That is why continuous close matters more than the flux tool itself.
Vic, the bookkeeping agent, keeps the ledger current so that every period flux runs against is already reconciled and trustworthy. Margo, the FP&A agent, runs the period-over-period comparison, applies the thresholds, and drafts the driver explanations for each flagged line, so the review arrives as a reviewed artifact rather than a blank spreadsheet. When the numbers reach the board, Nia carries the same explanations into the reporting pack, which is the foundation of board reporting that builds trust.
For teams still assembling the close by hand, financial close software and a fixed statement layout are the starting points. If the P&L is improvised each month, a stable template such as the pro forma income statement generator gives flux a consistent set of lines to compare across periods.
FAQ
What is the difference between flux analysis and variance analysis?
Variance analysis is the broad discipline of explaining any gap between two numbers. Flux analysis is the specific form that compares one actual period to another actual period, such as this month against last month. Budget vs actuals is the other main form, comparing actual results to the plan.
What materiality threshold should a startup use for flux analysis?
Use a dual threshold: a percentage and a dollar floor, flagging a line only when it clears both. A seed-stage company can start at ten percent and $2,500, with a lower floor of about $1,000 for sensitive accounts like cash, payroll, and deferred revenue. Raise the dollar floors as monthly spend grows.
How often should flux analysis be run?
Run month-over-month flux at every close, because it catches posting errors while they are still cheap to fix. Add quarter-over-quarter and year-over-year comparisons for board reporting and the annual audit, where trend and seasonality matter more than a single month's timing.
What makes a flux explanation strong enough for auditors?
A strong explanation names the specific business event, quantifies its dollar impact, and can be traced to a source document. "Expenses increased" fails. "Travel rose $45,000 because two people attended a conference that did not occur last period" holds, because a reviewer can match it to receipts and reconcile the drivers to the total change.
Flux analysis is the cheapest quality control a growing company owns. It turns the close from a data-entry exercise into a review that catches errors before they reach a board deck or a diligence room. Start your free trial and let your finance team run flux against books that are already closed.
