Docs / Analytics & optimization

Finding your best-converting answers and segments

When people take your quiz or survey, the answers they pick can tell you a lot. Some answers are chosen by people who go on to buy more often, spend more, or stay valuable for longer. SupaPop groups your respondents by the answer they picked and shows you how each group performed. This helps you spot which segments to lean into, what products to feature, and where people get stuck.

Before you start

  • You need a popup with at least one quiz or survey question (a question that has answer choices).
  • People need to have completed the quiz so there are respondents to group. New popups, or quiet date ranges, will not have enough data yet.

How to find your best-converting answers

  1. Open Analytics from the top app navigation.
  2. In the controls bar, find the selector labeled Popup. By default it shows All popups.
  3. Choose a single popup from the Popup list. The per-question cohort tables only appear when one popup is selected. They do not show under All popups.
  4. Pick a date range using the date range selector (Today, Yesterday, 7d, 30d, 90d, All, or a Custom range). This window controls which respondents and purchases are counted.
  5. Scroll down to the tables. You get one table per question, with the question text as the heading. Each row is one answer choice.

How to read the table

Each table has these columns, left to right:

  • Answer: the answer choice people picked. Answers with no recorded value show as (no answer).
  • Respondents: how many quiz completers picked this answer.
  • Purchases: how many of those people went on to buy, matched through Shopify orders.
  • Purchase %: the share of that group who bought (purchases divided by respondents).
  • vs avg: how this answer’s purchase rate compares to the question’s own average, in percentage points (for example, +3.2pp or -1.5pp). Green is above average, red is below.
  • AOV: average order value for the buyers in this group (revenue divided by purchases).
  • Rev / resp: revenue spread across the whole group, including people who did not buy.
  • LTV (90d): average 90-day value among the respondents in this group who left an email, including repeat orders.
  • Dropoff: the share of people who picked this answer but never moved past this step, with the sample size in parentheses (for example, 100% (3)).

Just under each question heading, the helper text shows that question’s baseline purchase rate. Every vs avg number and outlier flag is measured against that baseline, not a single number for the whole popup. So each question has its own average.

How to spot the strongest segments

  • Look for outlier+ (green) or outlier- (red) pills next to an answer. A pill means SupaPop is statistically confident the group really does convert above or below the question’s baseline, not just by chance.
  • Use the Findings panel near the top of the page. It surfaces the most actionable issues for the selected popup as plain-language cards, ranked by severity and how many visitors they affect. Strong answer outliers show up here (an outlier card reads something like “answer” converts at X% (higher than Y% baseline)), alongside other issues like drop-off hotspots and form friction. It shows the top eight.
  • Once you know which answers convert best, you decide what to do with that insight. You might feature those products more heavily, write copy that speaks to that group, or build routing so those people see a more relevant next step. SupaPop shows you the segments; acting on them is a manual choice you make in your store and your popup setup.

Tips

  • vs avg is a comparison, not a promise. It tells you the people who picked an answer converted more or less. It does not prove the answer caused the result. Treat strong segments as patterns to lean into, not guaranteed levers.
  • A big vs avg number with no outlier pill is probably noise, especially when Respondents is small. Trust the pills before you trust the percentage.
  • The Findings panel uses a stricter bar than the table. An answer can show a pill in the table but not appear in Findings if the group is too small or the gap is too narrow. Findings is your prioritized shortlist to look at first.
  • AOV and Rev / resp are not the same. AOV counts only buyers. Rev / resp spreads revenue across everyone, buyers and non-buyers. A group can have a high AOV but low Rev / resp if only a few people bought. Use Rev / resp when comparing whole segments.
  • LTV (90d) is averaged only over the respondents in a group who left an email. Respondents with no email on file are not included in that average, because there is no email to link their later orders to. A group where almost no one left an email can show an empty or near-zero LTV, even if some of those people bought.
  • The Dropoff column and the outlier flags rely on data collected from late April 2026 onward. Popups that ran before then, or low-traffic answers, will show a dash. A dash does not mean zero dropoff. It means there is not enough data yet.
  • A figure like 100% (3) in Dropoff is a tiny sample. The number in parentheses is there so you can sanity-check it. Do not act on a 100% built from a handful of people.

Why don’t I see the tables?

  • You are on “All popups.” The per-question cohort tables only render for a single popup. Open the Popup selector and pick one popup.
  • “No data yet.” This means the selected popup has no activity in your date range. Pick another popup, or widen the date range using the date range selector above the tables.
  • A column is full of dashes. The Dropoff column needs answer data that is only collected from late April 2026 onward, and it hides any group with fewer than five selectors to avoid misleading you with small samples.