Review sales associated with AI visits

Use Shopping > Sales to review available ecommerce outcomes associated with recorded AI visits. The view depends on the site's connected Google Analytics information. It is separate from the product appearances measured in AI answers.

1. Check the data source

Connect the correct Google Analytics property and make sure it records the ecommerce activity you want to review. A connected property with no purchase information cannot establish product revenue.

Switch between vs last month and vs last year to change what the cards compare with. The cards cover the last full calendar month. Keep the period and data source consistent when explaining a change.

2. Compare visits and revenue separately

Review AI visits against other visits and Revenue by assistant. A source with more visits may not produce the same share of recorded revenue.

Check whether the numbers cover a meaningful set of purchases before drawing conclusions. A single large order can change a small channel's result substantially.

3. Investigate the pages shoppers arrive on

Use Where AI shoppers land to review entry pages. Open the relevant page and assess whether a visitor can find accurate product information, availability and a clear next step.

Compare the page with the answer that named the product where that evidence is available. The tracked answer and the recorded visit are different datasets; do not assume every visit came from a prompt you ran.

4. Report the limits beside the outcome

Revenue associated with a recorded source is not proof that the source alone caused the purchase. Missing referral or purchase information limits what the report can show.

Revenue from AI is a floor, because some apps hide where a visit came from. Avoid treating a missing tracked outcome as proof that a shopper never purchased.

Continue with Products for answer-level accuracy or Landing pages for page-level search and visit information.