How many survey responses do you actually need? Enter your population size, desired confidence level, and margin of error to find out instantly.
The industry standard. You can be 95% sure results reflect the true population value.
Required Sample Size
Small Sample
You need 370 responses from a population of 10,000 to achieve 95% confidence with a Β±5% margin of error.
How sample size changes with different confidence levels and margins of error for a population of 10,000:
| Margin | 90% | 95% | 99% |
|---|---|---|---|
| Β±1% | 4,036 | 4,900 | 6,240 |
| Β±3% | 700 | 965 | 1,557 |
| Β±5% | 264 | 370 | 623 |
| Β±7% | 137 | 193 | 328 |
| Β±10% | 68 | 96 | 164 |
Sample size is the number of responses you need to collect from a survey in order for the results to accurately represent your entire population. If you survey too few people, your results may be unreliable. If you survey too many, you waste time and resources without meaningfully improving accuracy.
For example, if you have 10,000 customers and want to understand their satisfaction level, you do not need to survey all 10,000. A carefully calculated sample of 370 responses (at 95% confidence with 5% margin of error) will give you statistically reliable results that represent the whole group.
Sample size calculations are essential for any organization that collects feedback through surveys, whether that is customer satisfaction (CSAT), Net Promoter Score (NPS), employee engagement, or market research.
This is the total number of people in the group you want to learn about. It could be your total customer base, the number of patients at a clinic, or the number of employees in your company. If you are unsure, use your best estimate -- the calculator adjusts for different population sizes.
The confidence level represents how sure you want to be that your results reflect reality. A 95% confidence level (the industry standard) means that if you repeated the survey 100 times, 95 of those surveys would produce results within your margin of error. Use 90% for quick internal checks and 99% for high-stakes decisions.
The margin of error is the range within which the true value is likely to fall. A 5% margin of error means if your survey result is 75%, the true value is likely between 70% and 80%. Smaller margins give more precision but require more responses. For most business surveys, 3-5% is the sweet spot.
The calculator instantly shows the minimum number of responses you need. It also shows the response rate required and a comparison table so you can quickly see how different settings affect the sample size.
These two parameters work together to determine how reliable your survey results are. Understanding the tradeoff between them is key to designing effective feedback programs.
How sure you are that results reflect the true population value.
90% -- Good for internal estimates and low-stakes decisions.
95% -- The standard for most business and academic research.
99% -- Required for regulatory, medical, or high-stakes decisions.
The range of accuracy around your survey results.
1-2% -- Very precise. Requires large sample sizes.
3-5% -- The sweet spot for most business surveys.
6-10% -- Acceptable for quick directional insights.
The Sample Size Formula
n = (Z² × p × (1 - p)) / e²
Infinite population formula (p = 0.5 for maximum variability)
nadj = n / (1 + (n - 1) / N)
Finite population correction (N = population size)
The key tradeoff: higher confidence and lower margin of error both require more responses. A 99% confidence level with a 1% margin of error might require surveying nearly your entire population, while 90% confidence with 10% margin of error might need fewer than 100 responses. Choose settings that match the importance of the decision you are making.
Sample size is not just an academic exercise. Getting it right has real business implications across every industry.
Knowing your required sample size before launching a survey helps you plan realistic timelines and budgets. If you need 400 responses and your typical response rate is 20%, you know you need to invite at least 2,000 people. This prevents the common mistake of collecting too few responses and drawing unreliable conclusions.
When measuring NPS or CSAT across multiple locations, you need enough responses per location for each score to be meaningful. A restaurant chain with 50 locations might need 100+ responses per location to compare satisfaction scores reliably. Without proper sample sizes, location-level comparisons become noise rather than signal.
Healthcare organizations often need to report patient satisfaction scores to regulators and accrediting bodies. These scores must meet statistical standards for sample size and confidence. Under-sampling can lead to inaccurate reporting and potential compliance issues.
Before committing development resources to a new feature, you want to be confident that customer demand is real. A quick survey of 50 people might suggest 80% interest, but with a margin of error exceeding 10%, the true interest could be anywhere from 70% to 90%. For high-investment decisions, you need tighter margins.
For organizations with thousands of employees, surveying everyone annually is expensive and disruptive. A properly calculated sample lets you measure engagement with the same reliability while reducing survey fatigue and improving response quality.
Calculating sample size is step one. The harder part is actually collecting enough responses. Customer Echo automates feedback collection so you consistently hit your sample size targets without manual effort.
Customer Echo automates feedback collection across every channel so you always have enough responses for statistically reliable insights.