Quiz Logic Jump
A quiz logic jump is a branching rule that sends a respondent to a different question or page based on the answer they just gave.
Key takeaways
- The rule sits on the answer, so one question can lead to several destinations.
- Unmapped answers fall back to the default order, which keeps partial branching safe.
- Every extra fork multiplies paths, so tracing effort grows faster than the benefit.
- A shared tail of common questions keeps scores comparable across branches.
- Branch-specific questions shrink the per-question sample available for analysis.
In depth
A logic jump attaches a destination to an answer. When the respondent selects that option and moves on, the funnel evaluates the rule and loads the named question or page instead of the next one in sequence. The rule sits on the answer rather than on the target, so a single question can send its options to several different destinations. Any answer without a rule falls back to the default order, which is what stops a partly branched quiz from breaking.
Branching pays off when the questions after the fork genuinely differ by path. If both branches end up asking similar things in different words, the tree adds testing work without shortening anyone's journey. The cost also grows faster than the benefit, because two forks of two options give four paths to trace and three give eight. Depth is the thing to control, so most funnels branch once or twice near the start and then converge back into a shared tail.
The usual pattern is a segmenting question early, a short branch of path-specific questions, then a common set everyone answers so the scores stay comparable. In a scorecard quiz that shared tail matters, because respondents who answered different questions still need totals that mean the same thing. Give each branch the same number of scorable items, or normalise the score per path, so a shorter route does not read as a weaker lead when sales opens the record.
Branching splits your data as well as your respondents. Each path collects its own answers, so per-question sample sizes shrink and comparing completion between branches takes longer to become reliable. Deep trees also become effectively untestable by hand, and a broken path can run for weeks because nobody happened to take that route while testing. When the goal is only to skip irrelevant questions rather than to ask different ones, skip logic is the simpler tool.
Example in practice
How to measure it
Measure completion per path rather than only overall. Divide the respondents who finish each branch by those who entered it, then compare branches against each other. A branch that completes noticeably worse is usually asking questions that do not fit the segment it serves. Count how many respondents enter each branch as well, because a path almost nobody takes may not be worth maintaining.
Then check average path length and score comparability. Count the scorable questions each branch actually asks and compare the resulting score distributions. If one branch scores systematically lower, the branches are unbalanced rather than the segment being weaker. Watch drop-off at the fork itself too, since a segmenting question people cannot answer confidently loses respondents at exactly that step.
Common mistakes
The frequent failure is branching before the segmenting question is reliable. If the fork depends on a self-reported role or channel that people interpret differently, a share of respondents takes the wrong path and answers questions that do not apply to them at all. Fix the question first, with fewer options, plainer wording, and an explicit option for the case that does not fit, before attaching any routing to it.
The second is testing only the paths you expect. Teams walk the main branch, ship, and discover months later that one option routes to a question somebody deleted. List every branch, walk each one through to the result page, and repeat that walk whenever a question is added, moved or removed, since reordering questions is what most often breaks a jump target without any visible error.
Frequently asked questions
What is the difference between a logic jump and skip logic?
A logic jump actively routes a respondent to a specific next question based on their answer, while skip logic hides or bypasses questions that no longer apply. Jumps choose a path; skip logic removes steps from it.
Do logic jumps hurt quiz completion rates?
Done well, they improve completion by shortening the path and showing only relevant questions. Problems arise only when branches are untested and create dead ends, so trace every path before you launch.
Can a logic jump be based on more than one answer?
Yes. Advanced funnels support compound conditions, such as jumping only when a respondent selects a channel AND reports a budget above a threshold. This lets you build precise, intent-driven branches.
When should I use a logic jump instead of a longer quiz?
Use one when different segments need genuinely different questions, not merely fewer of them. If everyone would answer the same items and some are simply irrelevant, skip logic does the job with far less structure. Jumps earn their complexity when a branch asks about something the other path never covers at all.
How deep should branching go?
One or two forks near the start, then converge. Each fork multiplies the paths you have to trace and test, so a third level usually costs more in maintenance than it returns in relevance. If you find yourself needing deeper trees, that is often a sign the quiz is serving audiences that should have separate funnels.
How do I keep scores fair when branches differ?
Give each branch the same number of scorable questions, or normalise by dividing the achieved score by the maximum available on that path. Without one of the two, a shorter branch produces systematically lower totals, and its respondents land in lower tiers for a reason that has nothing to do with how good a fit they are.
What happens if an answer has no jump rule?
It follows the default order and moves to the next question in sequence. That fallback is what lets you branch only where it matters and leave the rest linear. It also means a deleted or renamed target can fail quietly, so confirm after every edit that each rule still points at a question that actually exists.
Do logic jumps make a quiz harder to analyse?
Yes, moderately. Each branch collects its own answers, so per-question samples shrink and comparisons take longer to reach a size you can trust. Report completion and drop-off per path rather than pooled, and keep a shared tail of common questions so at least some items are answered by every respondent regardless of route.
Can a jump send someone back to an earlier question?
Most builders technically allow it, and it is nearly always a mistake. A backwards jump can create a loop where a respondent repeats a section, and it makes the progress indicator move the wrong way, which reads as a fault rather than a design choice. Route forward to a new question that asks what you need instead.