Choose orders sessions or qualified demos explicitly
Orders per session and qualified demos per eligible visitor answer different questions. An event count can include repeated actions by one person, while a user-based rate may count people who completed the action. Label the measure precisely rather than assuming every report uses “conversion rate” in the same way.
Illustrative store: 20 orders from 1,000 sessions gives two orders per hundred sessions. Illustrative SaaS site: eight sales-accepted demo requests from 400 eligible visitors gives two qualified demos per hundred eligible visitors. Both ratios equal 2%, but they describe different populations and outcomes. Neither establishes that the two sites perform equally well or that their buyers face similar decisions.
For a session-based purchase rate, count sessions with at least one purchase. If those 20 orders occurred in 18 of the 1,000 sessions, 1.8% of sessions purchased, while orders per hundred sessions remains two. Google Analytics uses this session-based approach for its session key event rate.
Keep eligible audiences and time windows comparable
A rate can move because the traffic mix, eligibility rule, time window or tracking changed. Small counts can swing sharply. A higher rate also does not establish higher profit, more suitable customers or a measured effect from the latest design change.
Next, write the metric as a sentence naming the action, denominator, exclusions and dates. Compare periods only after checking that those definitions and the event implementation stayed consistent.
Website field guide · Choose a conversion rate baseline you can use · Qualified demo
Reference: Google Analytics: ecommerce events; Google Analytics: key events; Google Analytics: dimensions and metrics. Checked November 4, 2026.