Suggested reading:
How are "Impressions" calculated?
A product impression is logged per product, and only when at least 50% of the product card is visible in the user's viewport is it considered a valid impression.
Why is the Impression count lower than the Click count?
This can happen if a user clicks a product extremely quickly (before 50% of the card registers as visible) or if there is a specific integration issue where the visibility event fails to fire but the click event succeeds.
Why does the Control Group have zero impressions?
This is expected behavior in the A/B Testing View. The Control Group represents users who are eligible to see the campaign but are shown nothing. Since no widget is rendered, no "product impressions" are logged. Their activity is tracked at the session level to compare total site revenue against users who did see the widget.
What is the difference between Direct Revenue and Assisted Revenue?
Direct Revenue: The user clicked a recommended product and purchased that specific product (or its variant) within the attribution window.
Assisted Revenue: The user clicked a recommendation but purchased a different product in the same session. This measures the "discovery" value of the widget.
Why do I see a discrepancy between "Smart Recommender Analytics" and "A/B Testing View"?
They use different logging logic:
A/B Testing View: Logs 1 impression per session (to compare against a Control group that sees nothing). It tracks site-wide conversions.
Product Analytics: Logs impressions per product (50% visibility rule). It tracks conversions tied to specific product clicks.
Advice:
Use A/B Testing to measure total uplift vs. Control. Use Product Analytics to optimize which products to show.
Why does my product appear in Top 100 Product Analytics but not in individual campaign results?
If a product is recommended across multiple campaigns, its data is aggregated in the Top 100 list. It might not meet the threshold to appear in the reporting view of a single, lower-traffic campaign.
How do I choose the right Attribution Window?
You can set the window to Session, 1, 7, 14, or 30 days. Choose based on your sales cycle:
Fast-moving goods (Groceries): Use Session or 1 Day.
High-consideration goods (Furniture): Use 14 or 30 Days to capture delayed decisions.
What does the purple trophy icon mean in the results?
The purple trophy highlights the winning variant in an A/B test. The winner is determined primarily by Direct Revenue. If revenue is equal, it looks at Click-Through Rate (CTR), and finally Impressions.
Is it possible to see which products are top-performing for a campaign or a strategy?
Yes. You can check the top-performing products from the Product Metrics table by filtering the page by a campaign or a strategy.