Your books close. The P&L lands in your inbox. Revenue was $200,000 last month.
Now what? Is that good? Without something to measure it against, $200,000 is a fact, not information. It could be a record month. It could be a 25% miss that just shortened your runway.
Budget vs. actual (BVA) analysis closes that gap. It puts the plan next to the result and shows the variance, line by line, every month. Most CPG brands that produce one get limited value from it because they treat it as a report card for the month that just ended. The value is in what the variances tell you about the months ahead.
Variance Analysis Starts With Rate and Volume
Splitting a variance is what turns a number into a decision. Most revenue variances break into two questions. Did the rate change (AOV, cost per order, units per order)? Or did the volume change (units, orders, customers)? Most misses are a mix, and the mix tells you what to do.
For example, if last month we:
Budgeted: 10K orders at a Net AOV of $50. Planned revenue was $500K.
Actually did: 9.5K orders at a Net AOV of $40. Actual revenue was $380K.
The miss was: $120K.
What do you do next?
Without knowing what drove most of the $120K miss, it can be difficult to know where to put your team's focus.
The story behind the $120K miss is that orders only drove $25K of the $120K. The remaining nearly $100K was from our 20% lower AOV.
Demand was there, and the team hit the volume target within 5%. The headline is that order value came in $10 below target.
That reframes the conversation. The instinct on a $120K miss can be to spend more on acquisition, but order volume only accounted for $25K of it. The real work is figuring out what happened to Net AOV. What were our discount rates? Did our mix shift toward lower-priced items? How did customers bundle items? Did we see a spike in returns?
From there, we know what to target: does anything surface in our return feedback, how should we market product bundles, did deeper discounting help us acquire new customers?
With only a certain amount of time, you and your team have to know the root cause of a miss and how to address it.
A CAC Miss Doesn’t Stay in One Month
Revenue variances are the easy ones to read. The ones that do the most damage sit in marketing, because they keep working on you after the month closes. Take a separate example:
Budgeted: $300K in paid media at a $50 CAC. That buys 6,000 new customers.
Actually did: $330K at a $66 CAC. That bought 5,000.
You spent 10% more and got 17% fewer customers. Now add retention. The plan assumed two orders per new customer in year one. If repeat behavior also runs light, at 1.7, the misses compound.
Plan: 6,000 customers x 2 orders x $50 AOV = $600K from that cohort.
Actual: 5,000 x 1.7 x $50 AOV = $425K.
A $175K revenue gap from one month of acquisition, or $88K of contribution at a 50% margin. The $30K is already gone, but most of the gap leaks out over the next year.
Run the same miss for six months and you’re $180K deeper in spend, 6,000 customers short, and over $500K of contribution behind. That’s the variance worth addressing.
How to Decide Which Budget Variances Matter
Every line shows some variance. Most don’t deserve attention, though. An 80% miss on a $1,000 line is an $800 problem. A 15% miss on a $100,000 line is a $15,000 problem. The second one does nearly nineteen times the damage and it looks calmer on a percentage report. Percentage variance is how you spot something behaving strangely. Dollar variance is how you decide the level of concern.
Apply both tests. For a brand doing $10M to $30M, a workable approach is to flag anything moving more than 10% or $25K, whichever trips first, then let the dollar size decide how much time it gets. Scale the dollar figure to your size. The percentage test catches small lines going sideways early. The dollar test catches large lines drifting quietly.
Not Every Variance Has a Fix
Before spending an hour on a variance, ask whether you can prevent it next time. Some misses are one-time and non-operational. A rush freight charge to hold a retailer’s delivery window. An unbudgeted legal bill. A cold storage unit that failed. These are real dollars, and they belong in the commentary. But they don’t need a corrective action plan, because there’s nothing to correct. That’s the cost of doing business.
A variance earns real attention when you can write down what you’ll do differently and measure whether it worked. CAC running 40% over plan is addressable through targeting, creative, or channel mix. A landed cost jump is addressable through renegotiation, a second supplier, or repricing. Those get owners and dates. Sorting material variances into addressable and cost-of-doing-business turns a long list of misses into an action-based plan.
Then Ask What It Means for Your Forecast
This is what separates a BVA package from a history lesson. The most valuable thing a BVA gives you is a read on where the company is heading, which few drivers decide whether you get there, and what cash looks like on the way. Every material variance should end with “so what we should do is.”
Run it forward. If last month's rate and volume hold for the next 6 months, what does that do to revenue? A $120K miss becomes a $720K miss over that period.
Translate to cash. Revenue misses don’t all hit cash the same way. A volume miss drops through at contribution margin, because you also skip the cost of those orders. A rate miss from discounting drops through almost dollar for dollar, because the product cost doesn’t change. So the cash hit depends on what drove the miss. Then state what that does to runway.
Check what it says about your assumptions. A volume miss concentrated in repeat purchases means cohort behavior is diverging from plan, so LTV is overstated and CAC payback is longer than planned. That’s a bigger problem than one soft month, and the BVA is where you catch it first.
Understand the qualitative theme. “Returns ran 4 points above plan.” That’s a number in the variance column, but a question for the team. Pull return reasons, read recent reviews, talk to customer service. A margin variance is sometimes a product problem in disguise.
Decide whether to reforecast. Reforecast when an assumption breaks, not when a month misses. A soft month from a late shipment changes nothing. A conversion rate running 30% below plan for a full quarter means every future month built on it is wrong. Most brands settle on a quarterly reforecast with the original budget preserved alongside it, so you can see both what you committed to and your current estimate.
What a Complete Budget vs. Actual Package Looks Like
Your monthly BVA should have three layers:
A full workbook is the detail layer. Full P&L, Balance Sheet, and KPI Report with actual, budget, dollar variance, and percent variance, monthly and year-to-date. Revenue by channel with drivers underneath. This is the foundation.
The commentary is a targeted snapshot. This should capture what happened, why, areas to review, and what this means.
The headline is four to five sentences at the top. This breaks down the detail into the pieces that matter. A means B, and we should do C.
Where This Breaks
- The month closes on Day 30 or later. The decisions those variances should have driven are already made. By then you’ve placed the next inventory order, approved the media budget, and made the hire.
- No commentary. A variance column with no words is just data.
- The budget becomes irrelevant, and there’s no reforecast. Every month misses, and the BVA becomes an exercise instead of a tool.
- No forward statement. The package explains last month and stops, leaving you to work out what this means for next month.
None of This Works If Your Month-End Close Is Slow
Everything above depends on numbers you can trust, which is why this starts with the books rather than the analysis. You can’t break down a variance you don’t believe. Aim to close and deliver the package on the same schedule every month, with reliable and accurate data.
But you also don’t have to wait for the close on everything. DTC and Amazon data can be used to generate demand data throughout the month and support a weekly flash: units, revenue, AOV, discount depth, repeat vs. new customers against plan. If promo depth is running 10 points higher in week two, you can pull it back in week three rather than reading about it next month. The flash catches the trend early and helps you know what to look out for with the BVA.
All of it exists so you can see where the business is heading while there’s still time to change it.
Crew’s FP&A team builds budgets, variance analysis, and board decks for startup consumer brands.