What is a survey

A survey is not a one-time questionnaire, but the foundation of the Ask phase in the Ask → See → Discuss → Verify cycle. Through it, the organization regularly collects team feedback, and then compares results between measurements.

What it consists of

A survey is assembled from categories, questions, and answer options.

Categories group related questions and help interpret the results later, not only across the entire survey but also by individual blocks. For example, one block might be about process transparency, another — about interaction within the team.

A question sets a specific focus for the participant. They answer it on a scale or from a list, can add a comment, and the AI facilitator asks a follow‑up question when needed. This helps preserve not only the score but also the reason behind it.

Answer options provide a numerical basis for scores. Usually, a question has five options: from the weakest signal to the strongest. The additional textual feedback is no less important than the score, because it is from this that recurring themes, growth points, and ideas emerge.

Template and manual survey

A survey can be created from a template or from scratch.

In a template survey, the structure is already assembled: categories, questions, answer options, and analysis topics are preset. This option is better suited for regular diagnostics and repeated measurements, because results are easier to compare over time.

A manual survey is useful when you need to check a local topic or situation that is not in ready‑made templates. You set the structure yourself, and analysis topics are formed when summarizing based on the participants’ real answers.

Size and publication

For publication, a survey must contain from 3 to 36 questions. A practical guideline for regular diagnostics is 8–16 questions: enough to see the picture without overloading participants.

Too short a survey provides little material for analysis. Too long a survey reduces the quality of answers: people get tired faster, tend to answer more formally, and leave fewer useful comments.

Before launching, it’s worth checking three things:

• categories come across as understandable blocks, not as a technical grouping;
• each question asks about a single clear focus;
• answer options match the question wording and do not nudge toward a particular answer.

What a survey doesn’t solve on its own

A survey only collects honest feedback. It doesn’t assign blame, replace discussion, or turn into a task automatically.

The next steps in the cycle are important: close or summarize the measurement, look at the results, discuss growth points, and verify changes with a new measurement. Without this, the survey remains a set of answers, not a tool for improvement.