Analysing product variant conversion: Measure conversion using sessions where a variant was selected and shown.; Use stable variant keys in selection, basket and order records to ensure accurate matching.; Track eligible opportunities, matched orders and unavailable selections separately for clarity.
Image: Ecommerce Insight Desk

Funnel Analysis

Part of Product performance analytics

Analysing conversion by product variant

Define a variant conversion denominator using identified, available options and matched orders, with clear treatment of later purchases.

Desk Measurement · · 2 min read

Measure variant conversion against opportunities where the option was identified and available. A product-page view before a shopper chooses a size or colour belongs to the parent product. Assigning that view to every variant would inflate each option's denominator.

Choose the opportunity

For a selected-session-to-order rate, count sessions in which an available variant was selected and shown. Count how many of those same sessions contain a qualifying order for that variant. Divide the matched-session count by eligible selected sessions.

This same-session definition avoids crediting one later order to several earlier sessions. It also excludes purchases made in a later session, so label it accordingly.

A longer follow-up question needs a different cohort. Count each eligible shopper once for their first observed selection of that variant in the cohort period, then check for a qualifying purchase within a stated window.

Use this measure only when permitted identity and order linkage are reliable. If selection or linkage is missing, report selections, basket additions and variant orders as separate counts rather than publishing a conversion rate.

A basket-to-order comparison can be useful when selection data is unavailable, but it begins later in the journey. Define the eligible basket and whether the purchase must occur in the same session before calculating it.

Steps to Calculate Valid Variant Conversion Rates

  1. Match session to qualifying order within same session or defined windowEnsure identity and order linkage are reliable; use first selection in cohort if longer window is used
  2. Exclude later purchases from earlier sessionsDo not credit one purchase to multiple sessions; define 'same-session' or 'cohort-based' rules clearly
  3. Verify stable variant key mapping across systemsConfirm that product variant IDs or descriptors match consistently between selection, basket and order records
  4. Report counts and rates transparentlyShow eligible opportunities, matched orders, unmatched purchases, and unavailable selections separately

Match the saleable option

Use a stable variant key in selection, basket and order records. Shopify's analytics fields reference describes metrics and dimensions, and its sales reports can be grouped by product or channel, so check which stable key your own variant records expose. A GA4 ecommerce payload might include a variant descriptor, but do not assume that a store sends one or that a descriptive label maps to a saleable order variant. Check that mapping before joining the records.

Keep unavailable selections visible, but outside an available-variant conversion denominator. Note other differences between options, including price, promotion, preselection and stock. They may affect the comparison without establishing why anyone bought.

For each option, show the eligible opportunity count, matched qualifying orders, ordered units, rate definition and observation period. Show unmatched purchases and unavailable selections separately. A small denominator can produce a large percentage change from very few orders, so retain the counts.

If a shopper selects one colour and later buys another, preserve both the selection and the final order line. Inspect that path before interpreting it as lost demand for the first option. The result should point to a check of presentation, availability, price or variant mapping.

More from Funnel Analysis