Tag: Survey Design

Key Takeaways Branching exists to remove questions, not add them. The goal of every branch is a respondent who never sees an irrelevant question — not a more elaborate survey. Design the “no” path as carefully as the “yes” path. Most abandoned branching surveys fail on the answer the designer didn’t plan for, not the one they did. Shallow beats clever. Two or three branch levels cover almost every real use case; anything deeper usually means the survey is trying to do too much. Every path needs to be walked, not just the main one. A branching survey that’s only ever tested top-to-bottom will have broken paths nobody’s found yet. A branching survey shows or hides questions based on how someone answers the ones before it — so a respondent who says “no” to owning a car never sees three follow-up questions about their commute. Done well, it feels like the survey is paying attention. Done poorly, it feels like a maze: dead ends, questions that don’t make sense given what was already answered, or a “back” button that undoes more than intended.

Imagine you ask a friend, “Do you like the new office coffee?” If they only have two choices—Yes or No—they might struggle to answer. What if the coffee is “just okay”? Or what if they absolutely hate it?

Hey, form builders and workflow wranglers! We’ve all opened a form that looks like a wall of boxes and thought, nope, not today. Long forms are boss fights. No save point. Most people quit halfway.

Surveys are a powerful way to gather feedback and data, but guess what? They can go horribly wrong if you don’t get the details right. A little bias here, a confusing question there, and suddenly, your data is about as trustworthy as a random internet quiz telling you what kind of sandwich you are.