Flux Analysis: How to Explain Month-Over-Month Changes (With Template)

Flux analysis compares each account balance with the prior period, usually last month, flags the changes that cross a set dollar and percent threshold, and explains each one in writing. The change is current minus prior, in dollars and as a percent. It is part of the month-end close and answers what moved and why.

The pages that rank for this term come from close software vendors. They give the formula and one small example, and most stop there: no rule for a prior balance of zero, no answer on whether a line must pass the dollar floor and the percent floor or only one, no file to work in, and nothing about subscription businesses. A SaaS company's balance sheet swings every time annual renewals are invoiced, so it needs more than a percent column. This guide covers the formulas with Excel versions, how to set thresholds, a worked SaaS month with roll-forwards, how to separate seasonality from a real change, and a free template.

What is flux analysis?

Flux analysis is the review of how much each account changed between two periods, and why. "Flux" is short for fluctuation. You line up the closing balances for this period and the prior one, calculate the change, and write an explanation for every change large enough to matter.

It runs at the end of the month close, after reconciliations and before the numbers go to the CFO, the board or the auditors. Many teams skip the balance sheet half: FloQast reported in 2023 that fewer than 60% of the accounting teams on its platform run a balance sheet flux as part of their monthly close. It catches two kinds of problem:

• Errors. A missed accrual, a reversed entry, an invoice posted to the wrong account. A line that should be steady and moved 30% is often a booking mistake.
• Real changes that someone will ask about. A new hire class, an annual event, a big renewal month. If finance explains these first, the review meeting is about decisions instead of questions.

The usual comparison is horizontal: the same account across periods (month over month, quarter over quarter, or the same month last year). Some guides also cover vertical analysis, where each line is a percent of a total within one period, such as cost of revenue as a percent of revenue. Vertical ratios are useful context in a flux explanation, but the flux itself is the period-to-period change.

Flux analysis vs variance analysis

Flux analysis compares actuals with earlier actuals. Variance analysis compares actuals with a plan. The words are often used as synonyms, even by the vendors that rank for both, but they answer different questions. The two work together. Flux tells you that sales and marketing rose $65,000 from last month. Variance tells you whether that $65,000 was in the plan. For plan vs actual, with the price, volume and mix split, see the variance analysis guide. In most companies the controller runs the flux and FP&A runs the variance.

Flux analysis formulas, with Excel versions

Every flux calculation comes from a dollar change, a percent change and a threshold test. The Excel column assumes the prior balance in F, the current balance in G, the dollar and percent thresholds in J and K, the same layout as the template's MoM flux tab. Two details trip up most spreadsheets. First, a percent change against a zero prior balance divides by zero, so show "n/m" and let the dollar test decide. Second, a percent change on a negative balance (an overdrawn account, a credit balance) flips sign unless you divide by the absolute value.

How to set materiality thresholds for flux analysis

Set a dollar floor and a percent floor for each group of accounts, decide whether a line must pass both or either, and do it before you see the month. No single number fits every company. The threshold should match the size of the account and how much time the team has to write.

Published examples show how wide the range is:

Three choices matter more than the exact numbers.

1. AND or OR. Under AND, a line needs a large dollar change and a large percent change, which filters out a 40% swing on a $3,000 account and a 1% wobble on your largest balance. Under OR, either is enough. Pick one and write it down: Ramp's example uses "whichever is triggered first", which is OR, while Trintech's needs both, which is AND. In the worked example below, switching from AND to OR adds two lines to explain.
2. Groups, not one number. Revenue gets the lowest floors because it is the line the board reads first. Large, steady balance sheet accounts get higher dollar floors.
3. An always-explain list. SAB 99 lists qualitative reasons a small number can matter, such as whether it hides a change in earnings trends or affects loan covenants or management pay. Put accounts like cash, revenue reserves, related-party balances and anything tied to a covenant on a list that is explained every month, whatever the size.

Set the thresholds before the results are in. If you set them after, they tend to move until the awkward lines fall under them.

How to do a flux analysis in 7 steps

Close and reconcile first, then calculate, filter, split each change into its causes, and write. In order:

1. Reconcile before you flux. A flux on unreconciled balances explains numbers that will change again. Bank, receivables, deferred revenue and accruals should tie to their subledgers.
2. Pull prior and current balances from the trial balance. Same account list, same mapping, both periods. If an account was split or renamed, map the prior period to the new structure first, or the flux shows a reclass as a business change.
3. Calculate dollar and percent change for every account. Use the formulas above. Flag favorable or unfavorable for income statement lines, where direction has a meaning.
4. Apply the thresholds and the always-explain list. Only material lines get a written explanation. Everything else stays in the table.
5. Split each material change into its causes. For balance sheet accounts, use a roll-forward: opening balance, what came in, what went out, closing balance. For income statement lines, list the two or three items that make up most of the change, with amounts.
6. Check the same month last year. A month-over-month drop that is still growth against the same month last year is seasonality. Say so, with both numbers.
7. Write, review and carry forward. One short paragraph per material line, reviewed by someone who did not write it. Keep last month's explanations beside this month's: a change explained as "timing" three months in a row is not timing.

Worked example: a SaaS company's October flux

In this illustrative month, a SaaS company at about $12M ARR has 14 accounts in its flux review, and 7 need an explanation under the AND rule. The thresholds are the template defaults above. Cash is on the always-explain list. The readback in plain English: revenue grew slightly and services dropped after a large project ended. Spend rose for two known reasons, an annual event and two new account executives in sales and marketing, plus the start of the audit accrual in G&A. The balance sheet moves are almost all one event: annual renewals invoiced on October 1, which raised receivables and deferred revenue at the same time and left cash lower until customers pay.

Under the OR rule, two more lines would need a sentence. Subscription revenue passes the $25,000 dollar floor at +$25,200, and cost of services hits the 10% percent floor at -10.0%. Whether that is useful depends on the team. Subscription revenue probably deserves a line every month anyway, which is an argument for adding it to the always-explain list rather than switching the whole review to OR.

Balance sheet accounts: explain them with a roll-forward

A balance sheet explanation should quote the movements behind the change, and a roll-forward gives you those numbers. Opening plus additions minus reductions equals closing. If it does not tie, an entry is missing or a balance was reclassed, and you have found an error before the auditor did. Each line checks: $4,100,000 + $1,727,700 - $1,007,700 = $4,820,000, and $1,400,000 + $1,788,700 - $838,700 = $2,350,000. Billings equal subscription revenue plus the increase in deferred revenue ($1,007,700 + $720,000 = $1,727,700), plus $61,000 of services billed as delivered. The deferred commissions asset follows ASC 340-40, where commissions paid to win a contract are capitalized and amortized. The three statement model guide shows how these same billings, deferred revenue and commission lines link the P&L, balance sheet and cash flow.

The accrued commissions line is a good example of why flux needs context. It fell 55.7% because September, the last month of Q3, earned $107,000 of commissions that were paid in October, while October, the first month of Q4, earned $29,000. Nothing is wrong. That is the rhythm of a sales quarter.

Month over month or the same month last year

Compare with last month to catch errors, and with the same month last year to judge the trend, because SaaS months are not equal. Across ORM's customers, the third month of a quarter is always the largest and the first month is always the smallest. Q4 is usually the largest quarter, followed by Q2, Q3 and Q1.

The example's ARR bridge shows what that does to a month-over-month view:

New customer ARR fell 57.1% from September, a number that would alarm anyone reading it alone. Against last October it is up 20.0%. Both numbers belong in the explanation. Subscription revenue tells the same story: up 2.6% on the month and 29.8% on the year, in line with ARR.

The bridge also checks itself. October's beginning ARR equals September's ending ARR, and subscription revenue booked ($1,007,700) is close to average ARR divided by 12 ($1,007,667). Pete calls this monthly reconciling waterfall, beginning ARR, churned and contracted ARR, new and expanded ARR, ending ARR, with beginning always equal to last month's ending, the one chart he would put in front of every SaaS CFO. A flux review that includes it ties the revenue line to the customers behind it. The worked version is in the ARR waterfall that reconciles.

How to write a flux explanation

A good flux explanation says what changed, why, how much each cause explains, and whether it will repeat, in words someone outside the team can follow. Harvard's Office of the Controller lists the components of a good flux explanation as what changed, why it changed, the amounts behind the change, and writing that a third party can understand. Rules that keep the commentary useful:

• Lead with the number. The change in dollars and percent, then the cause.
• Make the causes add up. If you name two causes, their amounts should explain most of the change. "$40,000 plus $25,000" explains $65,000. "Various items" explains nothing.
• Say timing or permanent. Timing reverses (an invoice collected next month, a renewal billed early). Permanent does not (a lost customer, a new monthly cost).
• Say what happens next. Will it repeat, reverse, or change the forecast?
• Keep the same words for the same thing. Pete's rule for boards applies to flux too: a number that shows up on one slide must show up the same way on the next, or come with a clear reason it changed. The same goes for definitions. If the account mapping or an assumption changed since last month, note what changed and why.

Which SaaS accounts move, and what usually moves them

In a SaaS company, most large month-over-month changes come from billing terms, the sales calendar and hiring, so check those first. Bookings themselves are not on the balance sheet, but they drive half the table. A strong last month of a quarter raises accrued commissions, deferred commissions and, for annual billing, receivables and deferred revenue, then the next month unwinds it. When you know the sales calendar, most of the balance sheet flux explains itself.

The flux analysis template (Excel)

The free template has six tabs and every number in the example is illustrative. Yellow cells are inputs, white cells are formulas, material rows turn pink, and check cells must read 0. It opens in Excel and in Google Sheets (File, Import, Upload). How to use it with the example:

1. On Thresholds, set the rule (the example uses AND) and the floors per group before you look at the month.
2. On MoM flux, paste prior and current balances from your trial balance. Mark which income statement lines are better when higher, and which accounts are always explained.
3. Read the count at the top: in the example, 7 material lines. A second counter shows material lines that still have no commentary.
4. On Roll-forwards, enter the additions and reductions for each balance sheet account. Four of the cells fill themselves from other tabs, and every check reads 0 in the example.
5. On YoY flux and ARR bridge, enter the same month last year. In the example, new customer ARR is down 57.1% on the month and up 20.0% on the year.
6. Write commentary for each material line, using the roll-forward numbers. Change the rule cell to OR to see the two extra lines it would flag.

Common flux analysis mistakes

Most weak flux reviews come from thresholds set after the fact, percent changes that break on small balances, and explanations with no numbers.

• Running flux before reconciliations. The balances change again and the explanations go stale.
• Using only a percent threshold. Small accounts flood the list. Using only a dollar threshold hides a large percent change on a mid-sized account.
• Dividing by a zero or negative prior balance. Show "n/m" for zero and use ABS for negatives.
• Explaining with words only. "Timing" with no amount is not an explanation. Quote the roll-forward.
• Comparing month 1 of a quarter with month 3 and calling it a decline. Check the same month last year.
• Treating flux as a variance report. If the question is whether you hit plan, use variance analysis and write commentary the way revenue variance commentary describes.
• Writing a new explanation from scratch every month. Keep the log. Repeating causes are the most useful thing a flux review finds.

Where ORM fits

Flux analysis works because it studies change, not only balances. The same idea drives ORM's forecasting. ORM snapshots each customer's pipeline every day, so it can see what changed and when: a deal that moved stage yesterday, an amount that dropped, a close date that slipped. Pete's view is that the movement in the pipeline often tells you more than the state of it. Those daily changes feed a revenue forecast by week, with a range, that targets 95% accuracy, so the revenue line in next month's flux holds fewer surprises. To see it on your own data, see how ORM works.