Your ad budget said $18,000 for the month. Your actual spend came in at $22,400. That $4,400 gap is a variance, and whether it’s a problem or the smartest money you spent all quarter depends entirely on what happened next: did revenue climb enough to justify it, or did you just pay more to get the same result?
That question, asked over and over across every line of a P&L, is the entire job of variance analysis. It’s the practice of comparing what you planned against what actually happened, then digging into why the two don’t match. Calculating the gap is arithmetic anyone can do in a spreadsheet. Figuring out what caused it and what to do about it is where finance actually earns its keep.
One quick note before we go further: if you searched “analysis of variance” hoping for ANOVA, the statistical test used to compare group averages in research and A/B testing, that’s a different tool with a similar name. This article is about budget versus actual analysis, the kind that shows up on a monthly close, not statistical testing.
Variance Analysis Formula: How to Calculate It
The basic version looks simple:
Variance = Actual − Budget (or Standard)
But a single number like “$4,400 over” rarely tells you enough to act on. The real value comes from splitting that number into what caused it: a price change, a volume change, or both.
Price Variance = (Actual Price − Standard Price) × Actual Quantity
Volume Variance = (Actual Quantity − Budget Quantity) × Standard Price
Here’s why the split matters more than the total. Say your COGS came in $9,000 over budget for the month. If your supplier raised unit costs from $12 to $12.90 and you sold roughly the volume you planned, that’s a pure price variance: 10,000 units × $0.90 = $9,000. You need a pricing or sourcing conversation, maybe a renegotiation or a new supplier quote.
But if your unit cost held steady and you simply sold more than forecast, that same $9,000 is a volume variance, and it’s actually good news wearing an unfavorable label. Same dollar figure, opposite conversation. This is the single most common mistake in variance reporting: treating every unfavorable number as a cost problem when half the time it’s a demand problem, and demand problems are usually worth having.
Favorable vs. Unfavorable Variance: What It Really Means
This trips up a lot of operators who are new to reading variance reports, so it’s worth being blunt about it.
A favorable variance means the actual number beat the plan: costs came in lower, or revenue came in higher. Sounds great by definition. It isn’t always.
If your fulfillment costs came in under budget because order volume dropped, that’s technically favorable and genuinely bad. You didn’t save money. You lost sales. The variance report will show a green number, and the P&L will tell a worse story underneath it.
Flip it around. If your marketing spend ran unfavorable because you leaned into a viral moment or a surprise demand spike, and CAC stayed reasonable while revenue jumped, that unfavorable variance was the correct call. The sign of the number is not the verdict. The cause is the verdict.
5 Variances Every Ecommerce Business Should Track
Most variance analysis content is written for factories: labor hours, material yield, and overhead absorption. Almost none of it applies directly to a business selling product online. Here’s what shows up in a real e-commerce finance stack, and more importantly, how a manager should respond to each one.
Sales variance. Actual revenue versus forecast, split into price (did you discount more or less than planned) and volume (did units sold beat or miss forecast). If the miss is on volume and it’s consistent across the catalog, look at traffic and conversion before you touch pricing. If it’s concentrated in a handful of SKUs, that’s an inventory or merchandising problem, not a demand problem.
COGS variance. Driven by supplier pricing shifts, freight rate changes, or currency swings if you’re sourcing internationally. Understanding exactly how COGS is calculated for e-commerce is essential before you try to explain a variance. A recurring unfavorable COGS variance from the same supplier over two or three months is a renegotiation trigger, not a one-time write-off. A single-month spike tied to a known freight surcharge is noise you can footnote and move past.
Inventory variance. The gap between what your system says you should have on hand and what a physical count shows. Because inventory affects both cash flow and financial reporting, it deserves more attention than a simple operational stock check.
This one deserves the fastest response of any variance on this list, because it directly threatens your ability to fulfill orders. A small, consistent shrinkage percentage across every SKU usually points to a process issue in receiving or fulfillment. A large variance isolated to one or two SKUs is far more likely to be theft, a mis-pick pattern, or a system integration bug between your WMS and your storefront, and it’s worth investigating that week, not that quarter.
Marketing spend and CAC variance. Spending exactly to the budget is not the goal. Watch CAC against the forecast even when total spend lines up perfectly, because a brand can hit its ad budget dollar for dollar while cost per acquisition quietly climbs 20 percent. If CAC variance is unfavorable but LTV or AOV moved up to compensate, that’s a channel mix shift worth understanding, not necessarily a problem to fix.
Fulfillment and shipping variance. Carrier rate increases, dimensional weight reclassifications, and return rate swings rarely make it into the original budget accurately. If shipping variance is creeping unfavorable quarter over quarter, that’s usually a rate negotiation or packaging redesign conversation, not something you’ll fix by tightening the forecast.
How Often Should You Run Variance Analysis?
Cadence should match how fast the number moves, not a fixed accounting calendar. Sales and ad spend variance are worth checking weekly for any DTC brand spending real money on paid acquisition, which is why weekly financial tracking matters for e-commerce founders. COGS and inventory variance are typically a monthly exercise tied to the close, since they move more slowly, and the data (supplier invoices, physical counts) usually isn’t available more often than that anyway.
The mistake to avoid is reviewing everything at the same frequency. That either means you’re checking slow-moving numbers too often and wasting time, or checking fast-moving numbers too rarely and missing the window to act.
How to Use Variance Analysis to Catch Problems Early
The finance teams who get real value from variance analysis don’t review every line every week. They set thresholds, maybe 5 percent or a fixed dollar amount depending on the line item, and only dig into what crosses it. If 40 of your 45 SKUs are tracking within a normal range and five are wildly off, that’s where the meeting time goes. A report that highlights every variance with equal weight trains people to stop reading it.
Variance Analysis Example: Breaking Down a $4,000 Gap
Budgeted revenue for a product line: $50,000. Actual: $46,000. Unfavorable variance of $4,000.
Break it down: unit price held at plan, but units sold came in 8 percent below forecast. That’s a volume variance, not a price variance, which immediately rules out discounting as the cause and points toward either a traffic drop, a conversion issue, or a stockout on a bestselling variant. Pull the sessions and conversion rate for that SKU before you touch the marketing budget, and check inventory levels before you touch anything else.
That’s the whole discipline compressed into one example. A number moved, you split it into its parts, and the split told you exactly where to look next instead of where to guess.
Turn Variance Analysis Into Better Ecommerce Decisions
Variance analysis isn’t a compliance step to get through before the board deck goes out. It’s part of a broader financial reporting process for ecommerce businesses that turns raw accounting data into something management can actually use. Run properly; it’s the fastest diagnostic tool an e-commerce finance team has for catching a margin problem before it shows up three months later as a cash problem. The math takes five minutes to learn. Knowing which variances deserve a same-day response and which ones are just noise from a slow month, that’s the part that actually separates a useful finance function from a reporting exercise nobody reads past the summary row.
