View analytics

This page describes metrics that can help you determine how Vertex AI Search for retail is affecting your business.

Data sources for analytics

Review the prerequisites described in this section before using the analytics dashboard.

The analytics dashboard uses ingested user events as data sources for metrics. You must complete user events integration to be able to see user event analytics.

User events

The following user events are required to view some metrics:

  • Search events. Required for any type of metrics.
  • Detail page view events. Required for page view metrics.
  • Add-to-cart events. Required for add-to-cart rate, conversion rate.
  • Purchase-complete events. Required for total revenue, recommender-engaged revenue, average order value, purchase rate, revenue rate, average unit value, purchase order rate, conversion rate.

The following information is also used for analytics:

  • Product details. To ensure accurate computation of search-related metrics, include UserInfo.productDetails within your user events. This information allows Vertex AI Search for retail to attribute searches with specific product interactions. If the UserInfo.ProductDetail.quantity field is omitted within user events, a default value of 1 will be assumed for calculations.
  • Revenue. The UserInfo.PurchaseTransaction.revenue field is crucial for generating accurate revenue metrics. Ensure this field is populated within your user events to enable meaningful revenue-related analysis.
  • Attribution tokens. Attribution tokens are unique IDs generated by Vertex AI Search for retail and returned with each search request. Make sure to include that attribution token as UserEvent.attributionToken with any user events resulting from a search. This is needed to identify if a search is served by Vertex AI Search for retail.
  • User agent. Include UserInfo.userAgent with user events resulting from a search so that you can filter on user event metrics by device type.

View recommendations analytics

  1. In the Google Cloud console, go to the Search for Retail page.

    Go to the Search for Retail console

  2. In the navigation menu, click Analytics.

  3. Click the Recommendation tab.

  4. Select a metric type from the Metric list to see its graph for the last month.

View search analytics

Metrics are refreshed about every six hours. So, after you create your search app, it can take several hours before you can use the Analytics page.

If there are no user events, default values are shown for all metrics.

To view search analytics:

  1. In the Google Cloud console, go to the Search for Retail page.

    Go to the Search for Retail console

  2. In the navigation menu, click Analytics.

  3. Click the Search tab.

  4. Click a tab to view that metric group:

    • Per Search. Metrics are grouped by searches.
    • Per Visit. Metrics are grouped by search visits.
  5. To filter your metrics, use the following filters:

    • Date range. Select a preset date range or, if available, enter a custom date range.
    • Device type. Select a device type that queries occurred on.

View browse analytics

Metrics are refreshed about every six hours, so it can take several hours before you can use the Analytics page after you create your search app.

If there are not yet any user events, default values are shown for all metrics.

To view browse analytics:

  1. In the Google Cloud console, go to the Search for Retail page.

    Go to the Search for Retail console

  2. In the navigation menu, click Analytics.

  3. Click the Browse tab.

  4. Click a tab to view that metric group:

    • Per Browse. Metrics are grouped by browses.
    • Per Visit. Metrics are grouped by browse visits.
  5. To filter your metrics, use the following filters:

    • Date range. Select a preset date range or, if available, enter a custom date range.
    • Device type. Select a device type that browsing occurred on.

View serving config analytics

  1. In the Google Cloud console, go to the Search for Retail page.

    Go to the Search for Retail console

  2. In the navigation menu, click Serving Configs.

  3. Click the name of the serving config that you want to view analytics for.

  4. Click the Analytics tab.

  5. Select a metric type from the Metric drop-down list to see its graph for the last month.

Metrics definitions

Summary metrics

The sections below provide definitions of the site-wide metrics displayed on the Analytics page:

Recommendations summary metrics

The following metrics are displayed for Recommendations on the Recommendations tab.

Select a metric type from the Metric drop-down list to see its graph for the last month.

Metric Description
Total revenue

The total revenue from all recorded purchase events.

This value includes shipping and tax.

Recommender-engaged revenue

The revenue for purchase events that include at least one product that was selected from a recommendation panel.

This value includes shipping and tax and any discount applied.

Average order value (AOV)

The average value of orders from all purchase events.

Total revenue divided by the number of orders.

Recommender-engaged AOV

The average value of orders that include at least one product selected from a recommendation panel.

Recommender-engaged revenue divided by the number of orders with at least one product that was selected from a recommendation panel.

The following metrics for search text searches are displayed on the Search tab.

Metric Description
Search count

The total number of searches performed site-wide.

No results rate

The number of searches without results divided by the total number of searches.

Click-through rate (CTR) per search

The number of clicks from search results divided by the total number searches.

Add-to-cart (ATC) rate per search

The number of add-to-cart events occurring from search results divided by the total number of searches.

Purchase rate per search

The number of units in purchase-complete events occurring from search results divided by the total number of searches.

Revenue rate per search

The amount of revenue generated from purchase-complete events divided by the total number of searches.

Average unit value (AUV) per search

The amount of revenue generated from purchase-complete events divided by the total unit quantity generated by purchase-complete events.

Calculated based solely on purchase-complete events attributed to searches.

Personalization rate

The number of personalized searches divided by the total number of searches.

Search visit count

The total number of search visits performed site-wide.

Page view per search visit

The total number of detail-page-view events occurring in the search visit divided by the total number of search visits.

Add-to-cart (ATC) rate per search visit

The number of add-to-cart events occurring in the search visit divided by the total number of search visits.

Calculated by including all add-to-cart events, regardless of whether they can be directly attributed to the search.

Purchase order rate per search visit

The number of purchase-complete events occurring in the search visit divided by the total number of search visits.

Calculated by including all purchase-complete events, regardless of whether they can be directly attributed to the search.

Revenue rate per search visit

The amount of revenue generated from purchase-complete events in search visit divided by the total number of search visits.

Calculated by including all purchase-complete events, regardless of whether they can be directly attributed to the search.

Average order value (AOV) per search visit

The amount of revenue generated from purchase-complete events in the search visit divided by the number of purchase-complete events occurring in the search visit.

Calculated by including all purchase-complete events, regardless of whether they can be directly attributed to the search.

Browse search summary metrics

The following metrics for search browse searches are displayed on the Browse tab.

Metric Description
Browse count

The total number of browse searches performed site-wide.

No results rate

The number of browse searches without results divided by the total number of browse searches.

Click-through rate (CTR) per browse

The number of clicks from browse results divided by the total number of site-wide browse searches

The total number of browse searches includes paginated queries.

Add to cart (ATC) rate per browse

The number of add-to-cart events occurring from browse results divided by the total number of browse searches.

Purchase rate per browse

The number of units in purchase-complete events occurring from browse results divided by the total number of browse searches.

The total number of browse searches includes paginated queries.

Revenue rate per browse

The amount of revenue generated from purchase-complete events divided by the total number of browse searches.

Average unit value (AUV) per browse

The amount of revenue generated from purchase-complete events divided by the total unit quantity generated by purchase-complete events.

Calculated based solely on purchase-complete events attributed to browse searches.

Personalization rate

The number of personalized browse searches divided by the total number of browse searches.

Browse visit count

The total number of browse search visits performed site-wide.

Page view per browse visit

The total number of detail-page-view events occurring in the browse search visit divided by the total number of browse search visits.

Add to cart (ATC) rate per browse visit

The number of add-to-cart events occurring in the browse search visit divided by the total number of browse search visits.

Calculated by including all add-to-cart events, regardless of whether they can be directly attributed to the browse search.

Purchase order rate per browse visit

The number of purchase-complete events occurring in the browse search visit divided by the total number of browse search visits.

Calculated by including all purchase-complete events, regardless of whether they can be directly attributed to the browse search.

Revenue rate per browse visit

The amount of revenue generated from purchase-complete events in the browse search visit divided by the total number of browse search visits.

Calculated by including all purchase-complete events, regardless of whether they can be directly attributed to the browse search.

Average order value (AOV) per browse visit

The amount of revenue generated from purchase-complete events in the browse search visit divided by the number of purchase-complete events occurring in the browse search visit.

Calculated by including all purchase-complete events, regardless of whether they can be directly attributed to the browse search.

Configuration-specific metrics

You can see metrics for a specific serving config on the Serving Configs page. For metric graphs, click a serving config name to go to its details page, then select the Analytics tab.

The table below provides definitions for configuration-specific metrics.

Metric Description Details
Click-through rate (CTR) The number of product detail views from a serving config's recommendation panel, divided by the total number of predict queries for this serving config. For example, if the serving config points to a Frequently Bought Together model, then the CTR would be the number of product detail pages viewed from the shopping cart recommendation panel divided by the number of predict queries on the shopping cart page.
Conversion rate (CVR) The number of add-to-cart events from a serving config's recommendation panel divided by the total number of predict queries for this serving config. Similar to CTR, except that instead of product detail views, conversion rate uses add-to-cart events.
Recommender-engaged revenue The total revenue from the recommendations for this serving config. Similar to recommender-engaged revenue for summary metrics, but for this serving config only.

To track clicks from a serving config, Vertex AI Search for retail aligns the results in predict and search responses with ingested user events. If a clicked item appears in the responses for the same visitor ID within a one hour time window, the click/purchase is treated as a result of the Vertex AI Search for retail.

This process is fully automatic; you do not need to set anything up. However, when you configure your prediction and search requests for the first time, you should confirm that:

  • Visitor IDs in the request are the same as the visitor IDs you used in event ingestions.
  • The timestamp in the response roughly match the timestamp for that event.

When metrics are compared to the ideal expected result, or ground truth, the values might be lower, but the trends align.

A more direct alternative to the method above is to use attribution tokens. This requires significant instrumentation and is only recommended as an advanced tracking use case.