이 페이지는 아직 번역되지 않아 영어로 표시됩니다.
CAWI stands for computer-assisted web interviewing: an online survey that respondents fill in on their phone, tablet or computer. It is the fastest and most cost-effective way to reach large samples. The hard part is quality. Online samples need careful checks to keep out bots, duplicates and people who click through without reading. That is where most of our work goes.
When to choose CAWI
- Large samples quickly: hundreds or thousands of completes in days, not weeks.
- Several countries at once: the same survey in each language, fielded at the same time.
- Concept, ad and pack tests: show images and videos and measure the reaction.
- Trackers: brand health, awareness and usage, every month or quarter.
- Pricing and choice studies: conjoint, MaxDiff and price sensitivity questions.
How a CAWI project runs
- Brief and quote. You send the questionnaire or the objectives, the countries, the target group, the sample size and the length. We reply with a quote, including the expected incidence and timing.
- Scripting or your link. We script and host the survey, or field your own survey link with the redirects set up.
- Testing. You get a test link to check the routing, the wording and the look on a phone.
- Soft launch. We open the survey to a small group first and check the data before the full launch.
- Fieldwork. Quotas are tracked live. You get a daily progress update.
- Data delivery. You get the clean data, tables if ordered, and a report on the sample.
How we check quality
- Location and duplicate checks. Respondents must be in the right country and can answer only once. We check their network for VPNs and proxies.
- Attention checks. Simple questions that show whether people read before they answer.
- Speed and pattern checks. Responses that are far too fast, or that pick the same answer down a grid, are taken out.
- Open answer checks. Empty, copied or nonsense open answers are flagged and removed.
- Replacement. Every removed response is replaced, so you still get the sample you ordered.
What you get
- Data in SPSS, Excel or CSV, with a codebook.
- Tables with the banners you need, if ordered.
- Coding of open answers, if ordered.
- A sample report with the completes by quota, the incidence, the length and the responses removed.

