DatumForms.
Join the panel Sign in Create a survey

For research agencies & corporate teams

Sample size for a business survey: start with the decision

Plan a business-survey response target around the audience, sampling design, comparisons and feasible recruitment.

TL;DR

  • Define who and what the study is intended to describe.
  • Plan subgroup needs and recruitment losses before choosing a target.
  • More opt-in responses do not create a probability sample or remove bias.
Plan the sample first. Define who needs a voice. Match the target to the design.

“How many responses do we need?” is a planning question, not a number to copy from the last project. A useful answer connects the decision, target population, sample source, analysis and resources.

For an agency preparing a client proposal or a corporate team scoping a study, separate the desired number of completed responses from the number of people who need to be approached. Eligibility, contactability, participation and completion all affect the latter.

Define the population and unit

Write who belongs in the population, where they are, the relevant period and what one record represents. A survey of branch customers, a survey of businesses and a survey of individual purchasing managers have different units even when they concern the same market.

Consider an original hypothetical project comparing collection experiences at three service locations. Decide whether the study describes visits or unique customers. A frequent visitor should not accidentally be treated as several independent customers because their visits appear several times in the invitation list.

Identify how people will be selected

A probability sample uses a defined selection process with known, nonzero inclusion probabilities. An open link or an opt-in panel has different selection properties. State which approach applies and how the sample source covers the intended population.

Classical margins of sampling error depend on assumptions about design. They cannot simply be attached to an opt-in sample because it reaches a familiar size. Quotas may help manage sample composition, but do not by themselves establish those assumptions or remove coverage and self-selection issues.

Work backwards from the analysis

List the comparisons the report must support. A total of several hundred responses can still leave little information about a small subgroup. Decide which groups matter before fieldwork, and which differences will remain descriptive or exploratory.

Planning question What to write down
Main decision The choice or investigation the study supports
Priority groups The comparisons genuinely needed
Sample source How eligible respondents can be reached
Analysis limits What small bases or selection effects prevent
Recruitment plan Expected invitations, eligibility and completion assumptions

If the study requires formal estimation or hypothesis testing, involve a statistician with the actual design, effect or precision requirements, variation and clustering assumptions. A generic online calculator cannot resolve a mismatch between the sampling method and the intended claim.

Build a recruitment scenario

Use separate assumptions for qualifying, participating and completing. In a purely illustrative scenario, a team seeking 100 completed responses from a contact list might plan around 50% participation and 80% completion among participants: 100 divided by 0.5 divided by 0.8 gives 250 contacts, before any separate eligibility or delivery adjustment.

These are hypothetical arithmetic inputs, not Ghana benchmarks or a DatumForms performance estimate. Replace them with relevant project evidence and keep the assumptions visible. Do not multiply several percentages without checking whether each is conditional on the previous stage.

Check feasibility and cost

Ask whether enough eligible people are available within the country, time window, sample source and consent conditions. A desired target is not evidence that recruitment can deliver it. Include questionnaire length, review time and fieldwork constraints in the plan.

For panel research, an audience estimate is separate from recruiting people and accepting valid responses. Confirm actual study terms and available supply through the normal process; do not advertise a country or audience as available merely because it appears in a planning worksheet.

Agree what happens if the target is missed

Define the end date and what decisions the team can still make with a smaller achieved sample. Consider extending fieldwork, narrowing comparisons or reporting a limited exploratory study. Do not quietly combine incompatible sources or change eligibility to reach a round number.

In the report, disclose planned and achieved sample, selection approach, field dates, important exclusions and limitations. A transparent smaller study can be more useful than an impressive-looking total whose origins are unclear.

Continue the research workflow

Explore current plans , discuss your study or create a business account . For business inquiries: info@datumforms.com .

Sources and further reading

Tracking preferences

Essential functions stay on. Choose whether to allow optional tracking on this website. You can change your choices in the footer at any time.

Read our privacy policy