Margin of Error Calculator — Know How Precise Your Survey Is
Enter your sample size, confidence level, and (optionally) population size. We'll compute the ± margin of error with finite population correction applied automatically, show the formula step-by-step, and visualize your confidence interval.
Drag the slider to see how adding (or removing) respondents changes your margin of error — all other inputs held constant.
Ready to run your survey?
GenZform turns your research plan into a branded, mobile-ready survey in minutes — AI-assisted questions, logic, and analytics included. Free forever plan.
What is Margin of Error?
The plus-or-minus number that tells you how much sampling could move your result.
Margin of error (MOE) is the range above and below a survey estimate that likely contains the true population value. If a political poll reports 52% support with a ±3% margin at 95% confidence, the true support is most likely between 49% and 55%. Run the same poll 100 times with 100 fresh samples, and roughly 95 of those intervals would capture the real answer.
MOE matters because you almost never survey the whole population. You survey a sample — and samples disagree with the truth purely by chance. The margin quantifies how much disagreement to expect. A headline that says "63% of customers prefer option A" is meaningless without a margin attached. ±2% and ±15% both fit that headline and imply wildly different decisions.
What MOE does not capture: it's strictly random sampling error. It can't detect a leading question, a biased panel, or a 5% response rate. A tight ±2% on a broken survey is still junk. MOE answers "how precise" — not "how accurate."
Why Margin of Error Matters for Surveys
Three reasons MOE should appear on every chart you publish.
The Formula, Broken Down
Standard textbook margin of error — with finite population correction when it applies.
Worked Example: A Customer Satisfaction Survey
Real-world context with numbers you can audit step-by-step.
You run a SaaS company with 8,000 active customers. You email a CSAT survey; 600 respond. 72% rate you "satisfied" or "very satisfied." Your CEO asks: "Can we confidently claim 70%+ satisfaction in the investor deck?"
The 95% confidence interval is 68.55% to 75.45%. The investor claim holds: "We have 95% confidence that customer satisfaction is between 68.5% and 75.5%." Note that 70% is inside the interval — so "70%+" is plausible but not guaranteed. A stricter phrasing ("at least 68%") would be defensible at 95% confidence.
6 Ways to Reduce Margin of Error
Tighten your survey precision without tripling your sample budget.
Popular Use Cases
Pre-survey planning: verify whether your target sample size will deliver decision-grade precision before you launch.
Results reporting: attach a credible ± range to every stat in your post-survey report or stakeholder deck.
Academic research: meet IRB and publication requirements for documented sampling error in methods sections.
Customer insights: quantify uncertainty in NPS, CSAT, or product-fit surveys before making roadmap decisions.
Political & market polling: benchmark your MOE against industry standards (±3% at 95% confidence).
Employee surveys: apply finite population correction for small, bounded populations like a 500-person company.
Related Tools
Frequently Asked Questions
Everything about margin of error for surveys.
Margin of error (MOE) is the range above and below your survey result that likely contains the true population value. If 60% of respondents say yes and your MOE is ±3% at 95% confidence, the true value is most likely between 57% and 63%. It quantifies random sampling error — not bias from bad questions or non-response.
Most professional surveys target ±3% to ±5% at 95% confidence. Academic and political polling usually aims for ±3% (requires ~1,100 responses). Internal business surveys often accept ±5% (~385 responses). Margins above ±10% are considered directional only — useful for exploration, not decisions.
Four levers: (1) increase sample size — biggest practical lever, though with diminishing returns past n=1,000; (2) lower your confidence level (95% → 90% shrinks MOE ~16%); (3) use stratified sampling if your population has distinct subgroups; (4) when you genuinely know the expected split isn't 50/50, use the actual proportion — p=0.1 or p=0.9 yields a smaller MOE than p=0.5.
FPC shrinks the margin of error when your sample is a large fraction of the population. It's the factor √((N−n)/(N−1)). For infinite or very large populations, FPC ≈ 1 and is usually ignored. But for a 500-person company, a sample of 200 gives FPC ≈ 0.775 — cutting the MOE by 22.5%. This calculator applies FPC automatically whenever you enter a population size.
A 50/50 split is the most conservative estimate — it produces the largest possible margin of error. If you don't know what distribution of answers to expect, using 50% guarantees your sample is big enough. Once you have pilot data or prior knowledge, plug in the actual expected proportion for a tighter (smaller) MOE.
No — MOE only captures random sampling error. It ignores coverage error (who you could reach), non-response error (who chose to answer), and measurement error (badly worded questions, social desirability bias). A tight ±3% MOE on a biased sample is still misleading. Margin of error answers 'how precise' — not 'how accurate.'
Yes — higher confidence means a wider interval. 90% confidence uses z = 1.645, 95% uses z = 1.96, and 99% uses z = 2.576. Jumping from 95% to 99% increases the margin of error by about 31%, so you'd need ~70% more respondents to keep the same MOE. For most business decisions, 95% is the sweet spot.
Now that you know your sample plan, run the survey
GenZform turns your research plan into a branded, mobile-ready survey in minutes. AI-assisted questions, conditional logic, and built-in analytics so you can spot whether movement is real or just inside your margin of error. Free forever plan.