Variance Analysis for Food Manufacturing: How to Find Where Your COGS Is Leaking
Your standard COGS says one thing. Your actual COGS says another. Variance analysis tells you exactly why they differ — and which variances are worth fixing.
Variance analysis is the process of comparing what you expected to spend (standard cost) to what you actually spent (actual cost), and decomposing the difference into its root causes. For food brands, it is the primary tool for understanding why gross margin came in above or below plan.
Most food brands know their gross margin at the end of the month. Very few know why it moved. Variance analysis answers the "why."
The Four Types of COGS Variance
1. Purchase Price Variance (PPV)
PPV measures the difference between what you expected to pay for an ingredient and what you actually paid. It is calculated per ingredient, per purchase order:
PPV = (Standard Price − Actual Price) × Actual Quantity Purchased
A positive PPV means you paid less than expected (favorable). A negative PPV means you paid more (unfavorable). PPV is the most common variance for food brands because ingredient prices fluctuate constantly — commodity grains, oils, and sweeteners can move 10–30% in a quarter.
2. Usage Variance (Yield Variance)
Usage variance measures the difference between how much of an ingredient your BOM says you should have used and how much you actually used. It is the production efficiency measure:
Usage Variance = (Standard Qty − Actual Qty Used) × Standard Price
Unfavorable usage variance means you used more ingredient than the BOM specified. Common causes: production yield below standard, ingredient waste, over-portioning, or a BOM that does not accurately reflect actual usage.
3. Labor Efficiency Variance
For brands with direct labor in their production cost, labor efficiency variance measures whether the production run took more or fewer labor hours than standard:
Labor Efficiency Variance = (Standard Hours − Actual Hours) × Standard Rate
Most co-packed brands do not track labor separately — it is bundled into the co-packer's per-unit fee. Labor efficiency variance is more relevant for brands with their own production facility.
4. Overhead Absorption Variance
Overhead absorption variance occurs when actual production volume differs from the volume used to set the overhead rate. If you budgeted to produce 100,000 units per month and only produced 80,000, the fixed overhead cost per unit is higher than standard — an unfavorable volume variance.
A Worked Example: Granola Bar Production Run
| Ingredient | Standard Cost | Actual Cost | Variance | Type |
|---|---|---|---|---|
| Rolled oats | $0.18/unit | $0.21/unit | −$0.03 (unfavorable) | PPV — oat price increased |
| Honey | $0.12/unit | $0.12/unit | $0.00 | No variance |
| Almonds | $0.09/unit | $0.11/unit | −$0.02 (unfavorable) | PPV — almond price up |
| Packaging | $0.15/unit | $0.15/unit | $0.00 | No variance |
| Yield loss | 2% standard | 3.5% actual | −$0.02 (unfavorable) | Usage variance — higher waste |
| Total COGS | $0.54/unit | $0.61/unit | −$0.07 (unfavorable) |
In this example, actual COGS is $0.07/unit higher than standard. The breakdown tells you that $0.05 is from ingredient price increases (outside your control in the short term) and $0.02 is from higher-than-expected yield loss (potentially fixable). Without variance analysis, you just know your margin was lower than expected. With it, you know exactly where to focus.
How Often to Run Variance Analysis
At minimum, run variance analysis monthly — after each accounting close. For high-volume production or volatile ingredient costs, weekly variance analysis allows you to catch problems before they compound. The most valuable cadence for most food brands is per-production-run variance analysis: every time a batch closes, compare standard to actual immediately while the production data is fresh.
What to Do With the Variances
Not all variances require action. The decision framework is:
- Favorable variance, recurring: Update your standard cost downward — your BOM or price assumptions are too conservative
- Unfavorable PPV, one-time: Note it, monitor it, but do not change the standard unless the price change is permanent
- Unfavorable PPV, recurring: Renegotiate supplier pricing, find alternative suppliers, or update your standard cost and reprice accordingly
- Unfavorable usage variance, recurring: Investigate production process — yield loss, over-portioning, or an inaccurate BOM
How Guidance Automates Variance Analysis
Guidance calculates variance automatically at the lot level. Every production run generates a variance report comparing standard BOM cost to actual ingredient consumption and purchase prices. When a variance exceeds a configurable threshold, the system flags it for review. Over time, the system learns from recurring variances and updates its Confidence Score for affected cost calculations — so you always know which COGS figures are based on current data and which are based on assumptions that may be stale.
Frequently Asked Questions
What is a "standard cost" and how do I set it?
Standard cost is your expected cost per unit based on your BOM quantities and the prices you expect to pay for each ingredient. It is set at the beginning of a period (typically annually or quarterly) and held constant for that period, so variances reflect actual performance against the plan rather than just price changes.
How do I handle variance analysis if I use a co-packer?
With a co-packer, your primary variance is between your standard ingredient cost (based on your BOM) and the actual ingredient cost reflected in your co-packer invoices. You may also have a co-packer fee variance if the per-unit fee differs from what you budgeted. Yield variance is harder to track with a co-packer unless they report actual ingredient usage.
What variance threshold should I investigate?
A common rule of thumb is to investigate any variance greater than 5% of standard cost or $500 in absolute terms, whichever is smaller. For high-volume SKUs, even a 2% variance can represent significant dollars and warrant investigation.
Automatic variance analysis at the lot level
Guidance compares standard to actual cost on every production run and flags variances automatically — so you always know where your COGS is leaking without building another spreadsheet.
Get Early Access →