Understand your audience's answers and behavior
You'll see what visitors answer, which segments they belong to, and how their conversion rates differ, with the limits of each comparison explained.
Use your audience reports to see what people need and decide what to investigate next. Questions show their responses, segments show who belongs to each group, and Audience breakdown compares conversion rates among eligible segments.
These reports count different things. A typed answer doesn't automatically create a segment group. Collecting an email address, assigning a segment, and updating your email platform are separate steps.
Read answers to a question
The Questions library summarizes responses, visitors segmented, and visitors identified by email questions. It counts each visitor's response to each question separately, rather than counting a person once across the library.
A choice-question card shows its answer distribution and available performance measures. Someone who answers several questions can appear on several cards.
Open a question's Results tab for the selected date range and flow scope. All flows includes uses across the project. Choosing one flow narrows the report.
The funnel counts distinct visitors within your dates and flow scope. It shows who saw, answered, skipped, or closed the question where that data was captured, along with the answer rate. If data wasn't captured, it explains what's unavailable. Historical answers remain, and unknown views aren't treated as zero.
Choice questions show an Answers distribution. People can pick several answers on multi-answer questions, so the shares don't have to total 100%. Input questions show What happened with answers. Typed values aren't automatically answer segments.
Responses per day charts respondents by day. A recurring visitor can appear on several daily points, so their sum need not equal the distinct period total.
Read NPS scores and write-in themes
For an NPS question, Results shows the score, respondent count, detractor, passive, and promoter groups, and a 0-to-10 histogram. Each visitor counts once, using their latest score in your selected range.
For choice questions with Other write-ins, What "Other" answers say groups recent write-ins into themes. You can see the sample size, people counts, and example text. A theme summarizes that sample. It doesn't assign a segment to everyone who wrote an answer.
Add as answer turns a useful theme into an ordinary choice. An incompatible running test can lock this action, and you'll see an error if it fails. Review the wording and the places that use the shared question before changing it. The link to all write-ins opens the individual responses.
For filters, exports, and deletion, see manage and export responses. The question's Activity tab separately lists answers and captured skips or closes, newest first.
See who belongs to each segment
Open Segments, select a group, and open Results. Your date range controls visitor counts and shares. The pageviews-per-visitor comparison has its own limit. It looks back up to 30 days from your selected end date and samples up to 1,000 visitor IDs per segment.
Segments applied counts memberships. One person can belong to several groups, or to several segments in a group that allows multiple matches. That's why memberships can exceed the number of unique segmented visitors. For a group that allows multiple matches, distribution shares count memberships, not separate sets of people.
People can join a segment through rules or recognized contact data as well as answers. Segment totals aren't a count of past survey submissions. Different engagement rates also don't show that belonging to a segment caused someone to behave differently.
The engagement comparison uses the other segments in the same group as its baseline, not everyone else on your website. If the group allows multiple matches, one person can appear on both sides. See segment counts and engagement limits.
Compare conversion rates across your audience
From Dashboard, open Audience breakdown under More detail. This report starts at 30 days and has its own range, separate from the dashboard's Activity and Working selection.
The summary counts segment groups, groups ranked for comparison, groups with too little data, and groups with no visitors. Biggest differences highlights measurable differences within a group for your project's primary conversion measure.
| Control or section | Behavior |
|---|---|
| Rank by | Sorts by Conversion difference or Visitors. |
| Filter groups | Matches group names. A filter with no matches has its own empty state. |
| Ranked groups | The first two start expanded. Other groups can be opened individually. |
| Small differences | Groups without a large measured gap are kept separate from the leading differences. |
| Too little data | Groups without enough evidence for a useful comparison. |
| No visitors | Groups with no measured visitors in the period. Review in Segments opens their setup. There is no archive action here. |
Each comparison says Observation, not a test. It uses the group's combined visitors and converters as the reference, counting each person once. It doesn't use the site-wide average. Don't add overlapping segment rows to calculate a new total.
A group needs at least 30 conversions and two eligible segments before RightMessage ranks its conversion differences. Each eligible segment needs at least 200 visitors and either 10 actual conversions or enough visitors to expect 10 conversions at the group's average rate. Identity-only or email-platform-only audiences aren't equivalent to question-based comparisons. A missing ranking doesn't mean the segment has no value.
RightMessage leaves these comparisons out for routing measures. The quiz's answers decide where someone goes, so comparing those answers with their destination wouldn't tell you anything new.
When a report is empty or unavailable
No groups, no segmented visitors, no measurable differences, no filter matches, and unavailable data each mean something different. If your project has no primary conversion measure, Choose a primary outcome opens its settings.
For a question with No responses yet, use See where it's used to find its flow or form. If Results fails to load, Retry requests the data again. If an answer looks wrong or missing, check its question, flow, date range, and visitor identity before comparing the app with your email platform.
Use an observation to choose what to investigate. Use split testing or campaign experiment decisions when you need experimental evidence.