Телефонный опрос
Telephone survey services on DitWork can collect customer feedback, test hypotheses, measure satisfaction, update records and conduct structured research interviews. Unlike a telephone sales campaign, the primary objective is to obtain comparable answers without hidden promotion or pressure to support the client preferred conclusion.
Research quality depends on the sample, question wording, option order, interviewer instructions and coding rules. Before fieldwork starts, the client should define the research decision, target population, contact source, permitted attempts and the exact structure of the final dataset.
Suitable telephone survey projects
- customer satisfaction interviews after a purchase, support request or completed service;
- product use interviews covering choice, expectations, barriers and unmet needs;
- awareness or concept testing without turning the research call into a sales pitch;
- updating non-payment contact and profile fields under an approved questionnaire;
- surveys of employees, partners or event participants where contact is permitted;
- a short hypothesis check before a larger qualitative or quantitative study.
What the research brief should contain
| Element | Define | Why it matters |
|---|---|---|
| Objective | The decision that will use the findings | Prevents collection of interesting but unusable data |
| Sample | Inclusion, exclusion, quotas and contact source | Keeps the completed interviews relevant |
| Questionnaire | Exact wording, options, routing and neutral probes | Ensures consistent administration |
| Attempts | Calling window, limit, interval and final status | Avoids excessive repeat contact |
| Dataset | Codes, comment fields, identifiers and anonymisation rules | Makes the delivery auditable and analysable |
How to create a neutral questionnaire
- Start with the decision. Every question should test a defined hypothesis or provide information for a planned action.
- Separate concepts. Do not combine speed, quality and price in one rating when respondents may judge them differently.
- Avoid leading language. Wording must not praise the offer, exaggerate a problem or imply the expected answer.
- Check response options. Include other, do not know or not applicable when those responses are genuinely possible.
- Design routing. Respondents should skip sections that do not apply to their experience.
- Pilot the survey. A few interviews reveal unclear terms, excessive length and coding problems before full fieldwork.
Interviewer conduct
The interviewer identifies the organisation, explains the purpose and expected duration, follows the approved order and does not argue with an answer. Open comments should be captured close to the respondent wording. Any interviewer interpretation must be stored separately and clearly labelled.
- do not convert the research interview into concealed selling;
- do not replace a refusal with the nearest response option;
- do not read answer options for questions designed to be unaided;
- do not reveal another participant responses;
- end the call when requested and record any instruction not to contact the person again.
Fieldwork and dataset controls
The field report should show the original list size, unique contacts, attempts, completed interviews, partial interviews, refusals, unreachable numbers and exclusions. Without this disposition funnel, percentages can appear precise while the achieved sample is small or systematically biased.
Quality checks can include duration patterns, routing logic, missing required items, repeated answer strings and a sample of recordings or notes. Data cleaning must be documented, including removed rows, recoded values and reasons. The client should receive both the original export and the cleaned analysis file.
Privacy and data handling
The client confirms that the contact source may be used for the project and identifies the minimum personal data required. When an anonymous respondent code is sufficient, unnecessary identity details should not be shared with the interviewer. Recording, access, retention and deletion rules must be agreed before fieldwork.
Expected deliverables
A complete delivery includes the coded response table, variable dictionary, fieldwork report, exclusion log, verbatim comments and a short method note. Analytical interpretation should be separated from observed data: a distribution describes the achieved sample but does not automatically establish causation or represent the entire market.
Post the task on DitWork with the questionnaire, sample criteria, contact source, quotas, languages, attempt rules, recording requirements, target duration and dataset template. State whether neutral probes are allowed and how verbatim comments should be captured and reviewed.
