What is Variance Analysis?
Variance analysis compares actual results to budgeted or forecasted amounts and investigates the reasons for differences. It answers: did we hit our plan? If not, why not?
Full guide → Budget vs Actuals: Variance Analysis for Startups
Variances can be favorable (better than expected) or unfavorable (worse than expected). Understanding the root cause matters more than the number.
Why Variance Analysis Matters
Regular variance analysis improves forecasting accuracy over time. You learn which assumptions were wrong and adjust future forecasts. It also surfaces operational issues early.
A revenue miss might be timing (deals slipping to next month) or structural (market has changed). The response differs dramatically.
Running Variance Analysis
Compare actuals to budget monthly. Flag significant variances (typically >10%). Investigate root causes. Categorize as timing, one-time, or structural. Update forecasts based on learnings. Share findings with stakeholders. Start with a pro forma income statement as your baseline projection, then measure actual performance against it each month.
How to Run Variance Analysis Step by Step
Step 1: Pull your budget vs actuals for the period.
| Line Item | Budget | Actual | Variance | % |
|---|---|---|---|---|
| Revenue | $90,000 | $82,000 | -$8,000 | -8.9% |
| COGS | $18,000 | $16,500 | +$1,500 | +8.3% |
| S&M | $35,000 | $41,000 | -$6,000 | -17.1% |
| R&D | $55,000 | $53,000 | +$2,000 | +3.6% |
| G&A | $12,000 | $14,500 | -$2,500 | -20.8% |
Step 2: Flag material variances. Anything over 10% (positive or negative) needs investigation. Here: revenue (-8.9%), S&M (-17.1%), and G&A (-20.8%).
Step 3: Categorize each variance.
- Revenue miss: Structural. Two expected deals slipped to next month (timing) + one customer churned unexpectedly (structural)
- S&M over-spend: One-time. Unplanned conference sponsorship
- G&A over-spend: Structural. Legal fees for contract negotiations, likely recurring
Step 4: Update the forecast. Timing variances self-correct. One-time variances can be ignored going forward. Structural variances require adjusting future projections.
Step 5: Present to stakeholders. Board reports should highlight the 3-5 biggest variances, their causes, and what you're doing about the structural ones.
Common mistakes founders make:
- Not having a budget to compare against (you can't do variance analysis without a plan)
- Treating all variances the same (timing vs one-time vs structural require different responses)
- Only investigating negative variances (positive variances contain insights too; why did you beat budget?)
- Not updating the forecast based on variance learnings
Formula
Variance = Actual - Budget
Variance % = (Actual - Budget) ÷ Budget × 100
Favorable variance: actuals better than budget
Unfavorable: actuals worse than budget
Example
Your SaaS company reviews monthly budget variance:
- Budgeted revenue: $150,000
- Actual revenue: $135,000
- Variance: ($15,000) or -10%
Root cause: Two large deals slipped to next month ($20K). One unexpected churn ($5K). Early close on three small deals (+$10K).
Net variance: -$15K, but deal slip is timing, not lost revenue.