How to Conduct a Survey – Deploying the Survey
With your objectives and questions in hand, it’s time to choose how you’ll reach respondents. Every deployment method has trade-offs in cost, reach, and response rate, and the right choice often comes down to what contact information you have available — email addresses, mailing addresses, phone numbers — and your budget.
Paper Surveys
Paper remains the most universal deployment method globally, reaching anyone who can read and write. It can be hand-delivered, mailed, or left for pickup, but completed surveys must still be collected and entered or scanned into a digital format for analysis. Paper typically produces lower response rates for customer surveys (roughly 10–15%) than for employee surveys, where response rates of 75% or higher are common.
Online Surveys
Online deployment collects data automatically in digital form, launches quickly, and is generally the lowest cost per response of any method. It works through invitation emails followed by reminders, and it supports features like branching to different question sets for different respondents.
Mobile & Text (SMS) Surveys
A shift since the earlier version of this series: short, mobile-first surveys sent by text message have become a mainstream deployment channel, not a novelty. Because a text survey sits directly in a conversation the respondent already has open, it consistently outperforms email — current industry benchmarks put SMS survey response rates in the 40–60% range, well above the roughly 15–25% typical of email-linked surveys. Mobile deployment works best for short, transactional feedback (a handful of questions, sent close to the moment of the experience) rather than a lengthy engagement survey.
Telephone Surveys
Telephone surveys ask very little effort of the respondent, which can drive strong response rates on customer research. The trade-off is cost and time: live interviewers must place every call. Employee surveys are rarely deployed by phone.
In-Person Interviews
In-person interviews are generally the most expensive deployment method once you account for interviewer training and travel, and they are generally not recommended unless no other option reaches your population.
Anonymity and Confidentiality
Respondents need to trust that honest answers won’t come back on them. This is especially true of employee surveys, where fear of a supervisor’s reaction can quietly bias results. Many organizations use a third-party firm specifically to protect that anonymity.
Representative Sample
Because you’re surveying a sample rather than every employee or customer, you need enough responses for that sample to represent the whole population reliably. The standard target is a 95% confidence level with no more than a 5% margin of error. As a rule of thumb: a 100-employee company needs about 80 responses to hit that target (an 80% response rate); a 1,000-customer population needs about 278 responses (a 27.8% response rate). Reaching a representative sample is never simply a matter of hitting a flat percentage of your population — the required number changes with population size.
Incentives
Incentives are rarely needed for employee surveys, since most employees want to be heard. For customer populations under 500, a $10–$50 incentive per respondent may be needed to hit a representative sample, since a high response rate is required from a small pool. For larger customer populations, a smaller incentive — or none at all — is usually sufficient.
NBRI Research Team
National Business Research Institute
























By submitting this form you agree to our