Table of Contents
Before you start using Convert's reports, ensure that you've finished setting up your account. To unlock the full functionality of Convert's reports, you'll need the following:
A Google Analytics data integration provides the behavioral data that populates Convert's reports. For more information on setting up a data integration, click here.
Convert creates a set of reports for each Site in your account. A Site is tied to a data integration that it uses as its data source. For more information on setting up a site, click here.
A Rank Model uses machine learning to generate product rankings based on your behavioral data. For more information on setting up a rank model, click here.
Note: To preview rankings in Convert reports, a rank model version will need to have finished the creation process.
Solution and Campaign
A Solution defines how rank models should be applied to a product list. A Campaign makes rank models and solutions available for ranking. For more information on setting up a solution and campaign, click here.
Note: To preview rankings in Convert reports, a solution version will need to have finished the creation process.
A Recommendation model allows you to view recommendations for users or products. You can request this by navigating to the Recommendations report and clicking Enable Recommendations.
Once you've met the requirements above, Convert's reports will be available for use. Please note the following key controls:
Reports can display data from one data source at a time. Use the Site Selector dropdown to toggle between different data sources.
Reports display data from the last 7 days by default. Use the Date Picker to change the date range to view data from.
Note: When previewing Convert's product rankings, ranks will be based on all data available to the rank model, regardless of date range selected.
The All Products report provides insight into product performance independent of context. In other words, all product metrics reflect gross performance across the site as a whole—including interactions from categories, search results, recommendation carousels, and more.
In addition to viewing product performance metrics, the All Product report can be used to view sitewide relevance ranks. In a nutshell, these rankings answer a basic question: What products are likely to be most relevant to customers across all contexts?* This information can be used to improve your merchandising across numerous use cases, such as PPC advertising, email campaigns, inventory management, and more.
*Note: Remember that "relevance" is determined by the goal of the rank model. If you view ranks using a "Purchase" rank model, you're answering the question "What products are most likely to be purchased by customers across all contexts?" Likewise, if you view ranks using a "Revenue" rank model, you're answering the question "What products are likely to generate the most revenue across all contexts?"
To view sitewide relevance rankings, select a Campaign, and click Show Ranks.
The Product Lists report provides insight into the performance of individual products lists on your site, such as categories, search results, and recommendation carousels. Additionally, you can view the performance of individual products within these lists.
The Product Lists report is also intended to be used for previewing the product rankings Convert will display on your site once implemented. To view Convert's rankings for a given category or search term, follow these steps:
- Click on the Product List you want to view product rankings for.
- Select a Campaign and click Show Ranks.
Lastly, the Product Lists report allows you to compare how a product performs within a given list to how it performs across your site as a whole. To enable this functionality, click the Tools button and select Show Global Performance.
The Recommendations report is intended to be used for previewing the recommendations Convert will provide for a given user or product. To preview recommendations, follow these steps:
- Select a Recommendation Model.
- Enter a Product ID or User ID and click Show Recommendations.
Note: When viewing recommendations for a You May Also Like recommendation model, you will also see a column labeled "Probability". This value is intended to quantify the relative likelihood that the user will be interested in the product.
|Product List Views||All Products, Product Lists||The total number of times a product has been viewed in a product list.|
|Product List Clicks||All Products, Product Lists||The total number of times a product was clicked when it appeared in a product list.|
|Product Detail Views||All Products, Product Lists||
The total number of times the product detail page was viewed for a product.
|Carts||All Products, Product Lists||The total number of times a product was added to the shopper's cart. Does not reflect the total quantity of units added.|
|Checkouts||All Products, Product Lists||
The total number of times a product entered the checkout flow. Does not reflect the total quantity of units checked out.
|Purchases||All Products, Product Lists||
The total number of times a product was purchased. Does not reflect the total quantity of units purchased.
|List Clickthrough Rate||All Products, Product Lists||The rate at which a product viewed in a product list was clicked.|
|Purchase to List View Rate||All Products, Product Lists||The rate at which a product viewed in a product list was purchased.|
|Purchase to Detail View Rate||All Products, Product Lists||
The rate at which a product with a detail view was purchased.
|Probability||Recommendations||The likelihood that a user will be interested in the recommended product.|