The formula
Margin of safety = (expected sales minus break even sales) / expected sales. Compute it in units or in revenue; both give the same percentage when price is constant. The answer is your cushion: the percentage by which sales can fall short of expectations before the business starts losing money.
Note the denominator is expected sales, not break even sales. This is deliberate: the margin of safety describes risk relative to your plan. A 48% margin means sales can come in 48% below plan and you still cover costs. It is a forecast-risk number, not a cost-structure number.
Worked example: units version
Expected sales: 800 units a month. Break even: 417 units (from $5,000 in fixed costs, an $8 variable cost, and a $20 price). Margin of safety in units = 800 minus 417 = 383 units. As a percentage: 383 / 800 = 0.479, or 47.9%.
Read it plainly: the business can sell 383 fewer units than planned, down to 417, and still cover all its costs. Only when sales fall more than 47.9% short of plan does the month go red. That is a wide cushion, and it is exactly what a lender or an owner wants to see.
Worked example: revenue version
Expected revenue = 800 units x $20 = $16,000. Break even revenue = 416.67 x $20 = $8,333.33 (the exact figure, before rounding units up). Margin of safety in revenue = $16,000 minus $8,333.33 = $7,666.67. As a percentage: $7,666.67 / $16,000 = 47.9%, identical to the units version.
The revenue version is the one to quote when you plan in money: cash flow, loan applications, revenue targets. Use the exact break even revenue ($8,333.33) for the percentage math and the rounded-up version ($8,340) for the operational target. The difference is a few dollars of rounding cushion, and both are honest as long as you say which you used.
What counts as a good margin of safety
There is no universal good number, because it depends on how volatile your sales are. A common comfort zone for stable businesses is 20 to 30%: sales can miss plan by a fifth or more and you still survive. Seasonal businesses, startups, and anyone dependent on a few large customers should aim higher, 40% or more, because their sales swings are bigger.
A thin margin of safety, say under 10%, is a warning even when the business is profitable. It means one bad month, one lost client, or one supply disruption flips you into a loss. Thin margins of safety are how profitable-looking businesses die: not from bad economics, but from no room for error.
A negative margin of safety is the clearest possible signal. It means expected sales are below break even: you are forecasting a loss, and the percentage tells you exactly how far short the forecast falls. Do not round it away or rename it. Fix the inputs.
Three ways to raise it
First, lower the break even point. Cut fixed costs (the numerator), cut variable cost per unit, or raise price (both widen the contribution margin, the denominator). Lowering break even is usually the most reliable lever because it does not depend on customers cooperating.
Second, increase expected sales, but only sales you can defend. Padding the forecast raises the margin of safety on paper and nowhere else. A margin of safety built on a fantasy forecast is worse than none, because it hides the risk it claims to measure.
Third, shift the mix toward higher-margin products. The calculator's break even assumes one product; with several, the weighted-average contribution margin decides break even. Moving sales toward the products with the widest margins lowers break even without touching any cost or price, and the margin of safety rises as a consequence.
Margin of safety vs profit: do not confuse them
At break even, profit is zero and the margin of safety is zero. Above break even, both grow, but they are different things. Profit is dollars earned; margin of safety is risk tolerated. A business can have high profit and a thin margin of safety (great economics, fragile plan) or modest profit and a wide margin of safety (modest economics, robust plan). Track both. Profit tells you how the month went; margin of safety tells you how bad a month you can survive.