Skip to content

Help centre / Understand results

Read your analytics and responses

What each Analytics section counts, how to change the date range, and how to inspect shopper responses on every plan.

Updated 2026-10-31

Open Analytics in the app navigation. The numbers come from what shoppers do on your storefront: views from bots, results restored from a link, previews and the theme editor's design mode are not counted.

Pick the period

Use Date range to choose a preset period or Custom dates, then click Apply. Dates use UTC, and both the first and the last day are included.

What you see

  • All quizzes: totals: quiz views, shoppers started, completions, recommendations served, and add-to-carts from recommendations.
  • Per-quiz funnel: the same steps for each quiz, with the rate of each step.
  • Where shoppers leave: for each question, how many shoppers reached it, answered, skipped or left. A shopper counts as having left after 30 minutes without activity and no completion. With fewer than 20 shoppers, treat the counts as an early signal.
  • Lead capture: for each actual email-form position, the number who reached it, successful captures and capture rate, explicit skips, shoppers who continued after a save error, and those who left at the form. Capture rate divides captures by shoppers who reached that position. Drop-off excludes successful captures, explicit skips and continuation after a save error; a recent visit may still be in progress. Repeat visits count once per quiz session. Read this alongside the quiz completion funnel when comparing early and late capture.
  • Answer choices: how shoppers answered each question. On a question with several answers the totals can exceed 100%.
  • Top recommended products and Top products added to cart.

All of these are available on every plan. Home also shows shop-wide totals.

Compare traffic sources

Each quiz's Traffic sources section compares starts, completions, completion rate, leads, add-to-carts and attributed sales. Completion rate here is completions divided by starts. Source, medium, campaign and referring host stay together, so two email campaigns remain separate. Use Traffic source beside the date controls and click Apply to filter the funnel, questions, outcomes, top products, answer suggestions and quiz sales. Store-wide sales comparisons keep the same whole-store denominator.

Both quiz-page capture and the optional app embed wait for Shopify to allow analytics processing through its Customer Privacy API before reading or saving the remembered source or sending source data. Enabling the embed does not grant analytics permission. Quinn follows Shopify's permission result, which combines your store's privacy settings, the visitor's region and their choice. A declined, unavailable or failed permission check collects no source. The quiz keeps working while permission is being checked or unavailable.

When allowed, Quinn records the first traffic source in the current browser tab visit. It reads utm_source, utm_medium and utm_campaign from the current page, then falls back to an external referring host. Enable the Traffic source capture app embed in your theme editor to remember a permitted landing-page source before the shopper reaches the quiz. Without that embed, collection starts on the quiz page after analytics processing is allowed. The visit is remembered in session storage and expires after 30 minutes of inactivity on pages where capture runs; each quiz session keeps its original recorded source.

If analytics permission is later denied or cannot be checked, Quinn clears the remembered source from the browser and stops further capture. A later grant starts capture from the current page; earlier pages visited without permission are not recovered. Source facts already attached to saved quiz records remain covered by customer data export and redaction.

Direct or unknown means permitted collection ran without a usable UTM or referring-host signal. Not recorded means that session has no source record, including new sessions where analytics permission was declined or unavailable, and sessions from before this feature shipped. Quinn does not infer historical sources. Blocked JavaScript or unavailable session storage can also limit collection.

Attributed sales use Quinn's existing quiz-session order attribution, period, currency and test-order rules. A source records where the quiz visit came from; it does not claim that the channel caused the purchase. Source fields also appear in leads, response and lead exports, Shopify Flow and email-provider events, so you can use them in conditions and segments. Quinn keeps only bounded source, medium, campaign and referring-host values, never full URLs, query strings or advertising click IDs.

What to review (every plan)

Each quiz has up to three ranked suggestions. Quinn can highlight an answer associated with more shoppers leaving before completion, or an answer whose takers mostly receive only closest matches. Each suggestion names the saved question and answer, shows the counts behind it, and links to that question in the builder. Review the wording and next steps for departures; review product rules, available catalog coverage and the Sold-out products check for closest matches. These are associations, not proof that an answer caused a departure or that a product is sold out.

Suggestions select sessions by their first recorded answer in the chosen UTC period. Quinn waits a full 24 hours from that answer, then freezes completion and recommendation outcomes at that deadline. Later returns do not rewrite a check. This differs from the live funnel and question drop-off counts above, which retain their existing definitions. Each session contributes only its last committed choice per question in that window. A later skip replaces an earlier answer; multiple-choice selections are counted once per answer. The comparison group answered the same question with the same saved prompt but did not select that answer.

The chosen traffic source filter also applies to suggestions. Every compared group needs at least 20 shoppers after that filter, plus conservative confidence bounds adjusted for all answer versions checked in that quiz and period. A departure suggestion needs a supported difference greater than 10 percentage points. A closest-match suggestion needs evidence that more than half of shoppers with known recommendations got only closest matches. One exact match removes a session from the closest-only count; unknown match types are excluded from that check. Twenty shoppers unlocks a check, not a guaranteed suggestion. When data is thin, Quinn shows the closest group to that minimum. When enough data has no clear issue, Quinn says so and shows how many checks had enough shoppers.

Saved prompt, answer ID and answer label define each answer version. Earlier or deleted wording stays readable and is marked as historical; a deleted question links to the quiz builder. Identical wording can span edits to product rules or branching, which receipts do not snapshot. Findings cover single-choice, multiple-choice and dropdown answers, including picture choices. Free-text and slider values are excluded. Previews and theme-editor traffic are excluded at recording time, as for the other analytics. Storefront tests on a development store count as shopper sessions.

Shopper responses (every plan)

  1. On the quiz's funnel, click Shopper responses.
  2. Each completed session lists the shopper's answers, the products recommended (marked Exact match or Closest match), what they added to cart, and the email if one was captured.
  3. Use Search answer text to find answers, and Export CSV to download the list.

Cart additions are recorded quiz actions, not orders. Question and answer text is saved as it was when the shopper took the quiz; older records may show the current text.

To send these events to Google Analytics or Meta, see Connect other tools.

Still stuck?

Email us at info+support@legitimateapps.com. Tell us your store address and what you expected to see, and we will help.