How Large Should a B2B Survey Sample Really Be?
There is no single sample size (N) that can be considered definitive or statistically sufficient for B2B surveys.
There is no single sample size (N) that can be considered definitive or statistically sufficient for B2B surveys.
Unlike B2C, B2B companies sell to a relatively small pool of potential customers. In many cases, the total addressable market consists of only a few thousand organizations. Surveying 20–30% of that population would surely reduce the margin of error but this percentage is neither feasible nor required.
Moreover, B2B research often requires input from senior decision-makers. Recruiting such executives is expensive, time consuming, and frequently difficult due to limited availability and organizational gatekeeping. As a result, pursuing a very large N is rarely feasible in practice.
At the same time, if N is too small, variance increases and, consequently, the margin of error widens, which can make the results statistically weak and less useful for decision-making.
To illustrate the relationship between sample size, margin of error, and cost, consider a hypothetical example in which the average survey cost is roughly $100 per response across three population sizes and three commonly used margins of error. The data in the table below shows how increasing the sample size improves precision but also causes costs to escalate significantly.
| Population | Margin of Error ±3% Sample (N) | Cost | Margin of Error ±5% Sample (N) | Cost | Margin of Error ±10% Sample (N) | Cost |
|---|---|---|---|---|---|---|
| 500 | 345 | $34,500 | 220 | $22,000 | 80 | $8,000 |
| 1,000 | 525 | $52,500 | 285 | $28,500 | 90 | $9,000 |
| 10,000 | 965 | $96,500 | 370 | $37,000 | 100 | $10,000 |
Consider what this means in practice. If a survey of around 100 respondents (about ±10% margin of error) shows that 70% report a specific challenge, the true proportion in the population could realistically lie anywhere between 60% and 80%.
Increasing the sample to roughly 300–400 respondents (about ±5% margin of error) narrows that range to approximately 65%–75%. While this improves confidence, it typically requires three to four times as many respondents — and therefore several times the cost.
In effect, organizations may double or triple their research budget to reduce uncertainty by only a few percentage points on either side of the estimate. The improvement in precision is real, but modest relative to the additional investment.
This dynamic holds across populations of 500, 1,000, or even 10,000, because once populations reach a moderate size, accuracy is driven far more by sample size than by the total population itself. The result is diminishing returns: substantially higher spending yields only incremental gains in certainty.
In practice, the goal of B2B research is not to maximize sample size but to reach the right decision-makers within the relevant ICP, even if they are few in number.
