Reliable field data within 48 hours
Donors expect indicators. Program teams need data to adjust their actions. But between deploying fieldworkers, entering data from paper questionnaires, and difficulties accessing certain areas, traditional data collection requires significant resources, takes several weeks, and sometimes results in gaps in the data. The SOLIS Bot changes this by allowing affected individuals to respond directly from their phones, making it possible to quickly obtain reliable information without the need for fieldwork.
How to design a survey that gets completed
The rule learned in the field is clear: it is recommended that you create a survey with fewer than ten questions, using closed-ended responses as much as possible, and that takes less than five minutes to complete. Every extra question drops the completion rate. An effective post-distribution survey fits into three blocks:
- Identification (2 questions): area, type of aid received.
- Verification (4 questions): was the assistance actually received, did it match what was announced.
- Quality and satisfaction (3 questions): rating of how it went, fit to needs, free comment.
The person involved answers directly in the conversation, at their own pace:
SOLIDARITÉS INTERNATIONAL Thank you for taking part. Question 3 of 8: did you receive your assistance?
- Yes
- No
- Partially
Responses feed the dashboard in real time and export to Excel, ready for your reporting.
Where enumerators cannot go
In Ukraine, SOLIDARITÉS INTERNATIONAL collects its donor indicators in difficult to acces front-line areas. There, the design answers one non-negotiable principle, the respondent's safety first:
- anonymous by default, no personal identification questions,
- no risky questions, none about movements or positions,
- explicit consent at the start of the survey,
- the option to stop at any moment, with a simple "STOP".
In practice
The multiplier. In Lebanon, a MEAL team of three covers 300 distribution sites. Conventional monitoring had become increasingly difficult at that scale. With the bot, the same team collected 700 responses in 48 hours, for €8 in broadcast costs, where the traditional method would have cost over €1,700. The average response time went from more than four hours to three minutes and twenty seconds.
The uptake. Because the surveys are short and completed inside the conversation, they reach a 70% response rate at CARE Lebanon, an exceptional figure for remote monitoring. Across the network, 12,500 surveys were completed over the last period.
What it changes
The average cost works out at under €0.20 per response, 10 to 50 times cheaper than field collection. But the saving is not the only gain. By making collection faster and lighter, the bot lets you measure more often, react sooner, and concentrate field visits where a human presence genuinely adds something.
For quality data
- Test your survey with ten people concerned before the official launch. Ask them how long it took and which questions felt unclear. An ambiguous question across 300 sites costs you weeks of reprocessing, thirty minutes of testing saves them.
- Have sensitive surveys reviewed by your protection focal point, especially in a conflict context.
- Always triangulate the bot's data with other sources before an operational decision. The tool speeds up measurement, it does not replace judgement.