Product recommendation reports
Analytic reports for Optimizely Product Recommendations let you see the success of your campaigns and which Product Recommendations efforts work or do not work. You can track and view graphs for the following:
- Total orders, revenue, and products sold
- Order, revenue, and products sold through recommendations
- Average order value and units per order uplift with product recommendations
- Best performing pages
- Best performing widgets
Attribution model
The Optimizely Product Recommendations attribution model is the industry standard. If someone visits your site and clicks on a product recommendation, and then returns to the site within 30 days and buys the product they clicked and converted on, then full attribution for the conversion is given to Product Recommendations.
Example
If a visitor engages with a product recommendation, returns to the site, and purchases the item, plus other items they were not recommended, Product Recommendations are only attributed to the revenue generated from the recommended item they engaged with and clicked—not the entire basket value.
For example, Product Recommendations recommends a t-shirt for $5.50 to a visitor while browsing the site. The visitor clicks the recommended t-shirt but leaves the site without purchasing. The visitor returns later that day and purchases two of the t-shirts that Product Recommendations recommended to them and they engaged with. They also purchase a jumper for $14 that Product Recommendations did not recommend. The order total is $25; the total revenue attributed to Product Recommendations would be $11 (2 t-shirts at $5.50 each). If the visitor buys multiple units of the same item that Product Recommendations recommended to them and they click, this collective sum is attributed to Product Recommendations. Equally, if it was just one item, the sole value of that one item is attributed to Product Recommendations.
Product Recommendations can produce a graph that plots attribution against time. This shows when most people convert after being shown a product recommendation. The metrics used are cost per click (CPC) against time. On average, 85% of conversions occur within 24 hours.
Orders
The Orders report shows orders from online channels that Product Recommendations tracks.
- Time range – Edit the time range for the reports. Week is the default selection, which shows information from the past 7 days. You can also choose Month, 3 Months, or a custom time range. When you change the time range, click Apply to refresh the results.
- Currency – Select the currency you want to display in the report.
- Language – Filter by language to view product recommendation performance for a specific site locale.
- Export CSV – Export any reports to a CSV file.
- % of Orders from Optimizely – The percentage of total orders in which an Optimizely product recommendation was purchased.
- % of Revenue from Optimizely – The percentage of total revenue generated from Optimizely product recommendations purchases.
- % of Product sold via Optimizely – The percentage of total products sold that came from Optimizely product recommendations.
- Average order value – The average value of orders that contained an Optimizely product recommendation compared to those that did not over the time range.
- Average units/order – The average number of units purchased in orders that contained an Optimizely product recommendation compared to those that did not.
- Orders, Revenue, and Products sold tables – Displays the order data in table format. You can organize orders, revenue, and products sold using the following columns.
- Date – The date of the order.
- Total – Total number of tracked transactions, revenue, or units sold for that date.
- Recommendations – Total number of orders, revenue, or units sold that contain or were sold through an Optimizely product recommendation.
- Recommendations in Percent – The percentage of orders, revenue, or products sold that contain an Optimizely product recommendation.
Site Engagement
The Site Engagement report displays page impressions, clicks, click-through rate (CTR), and click-to-purchase conversion for product recommendations across your site.
- Total impressions – The number of page impressions where product recommendations were served.
- Total clicks – The total number of clicks on product recommendations.
- Click-through rate – A percentage worked out on the sum of clicks divided by the sum of impressions.
- Click-to-purchase conversion – A percentage calculated on the sum of purchased recommended products divided by the sum of clicks.
- Impressions versus click-through rates – A graphical representation of this information. The bars show the daily number of page impressions with product recommendations, and the line shows the daily click-through rate for product recommendations.
- Site Engagement table – Displays the site engagement data in table format. You can organize the data by the following columns.
- Date – The date of the engagement event.
- Impressions – Total number of impressions.
- Clicks – Total number of clicks.
- CTR – Click-through rate. Percentage of product recommendations visitors clicked.
- CTP conversion – Click-to-purchase. Percentage of product recommendations visitors purchased after clicking.
Products
The Products report shows the best-selling products on the site based on the revenue generated from Product Recommendations.
- Total units sold – The total units sold.
- Recommended units sold – The total units of Optimizely product recommendations sold.
- Total revenue – The total revenue.
- Recommendation revenue – The total revenue generated from Optimizely product recommendations.
- Best-selling products data table – Displays a full list of your products, sorted by the products that generate the most revenue from Optimizely product recommendations.
Page Types
The Page Types report shows page types on the site where product recommendations are served and how much Product Recommendations revenue is generated from each page.
- Page type report by revenue – A pie chart that shows which page types generate the most product recommendation revenue.
- Page Type data table – You can organize the data using the following columns.
- Page type names where product recommendations are present
- Number of impressions for that page type
- Number of clicks on product recommendations for that page type
- Click-through rate (CTR) for product recommendations, which is shown as a percentage, calculated by the sum of clicks divided by the sum of impressions
- Items bought shows the number of items purchased through product recommendations on that page type
- Click-to-purchase (CTP) rate for product recommendations shows a percentage calculated by the sum of purchased recommended products divided by the sum of clicks
- Revenue generated from product recommendations on each page type
Widgets
The Widgets report is similar to the Page Types report but with finer detail for individual widgets because some pages may have more than one widget.
- Widget report by revenue – The pie chart that shows which widgets generate the most revenue.
- Widget data table – Shows similar information as the Page Types report, except based on the widget level. Click a widget name to see the individual position performance details within a widget. You can organize the data by the following columns:
- Widget
- Impressions
- Position
- Clicks
- Click-through rate (CTR)
- Items bought
- Click-to-purchase (CTP) rate
- Revenue
Categories
The Categories report shows your top ten most performant categories.
- Category report by revenue – A pie chart that shows which categories generate the most product recommendation revenue. The revenue is not the total revenue for that category, but instead is the revenue from a widget that recommended an item in that specific category. You can filter by different category levels, which are determined by your product catalog.
- Category data table – You can organize the data by the following columns:
- Category
- Widgets
- Page Types
- Revenue
Server Metrics
The Server Metrics report shows two graphs:
- Number of product recommendations served per day
- Average time it takes to generate product recommendations and return them to a site (after a tracking request is sent)
Conversion
The Conversion report shows the ranking of products based on their overall conversion and page views on a site. It does not focus on product recommendations but can provide accessible, actionable insights based on your whole site, visitors, and orders.
Conversion definitions
- High converting – Any product that is in the top 10 of the highest 10% of conversions
- Low converting – Any product in the lowest 10 of the lowest 10% of conversions.
Traffic definitions
- High traffic – Any product in the top 30% of page views.
- Low traffic – Any product that is in the lower 40% of page views and with the lowest 10% excluded.
The following list shows the various combination views of conversion data, which are displayed in the tabs on the Conversion Summary view.
- High converting, high traffic – The 10 best-converting products on the site from the top 30% of products with the most page views.
- High converting, low traffic – The 10 best-converting products on the site from the 40% of products with the least page views, excluding the 10% of products with the lowest number of page views.
- Low converting, high traffic – The 10 worst-converting products on the site from the top 30% of products with the most page views.
- Low converting, low traffic – The 10 worst-converting products on the site from the 40% of products with the least page views, excluding the 10% of products with the lowest number of page views.
Rank & Trend view
The Conversion Rank & Trend view lets you generate a top 5 report of products, categories, or brands sold on your site based on a combination of criteria; best or worst performing by page views, revenue, or conversion. The generated report shows the total page views, units bought, conversion, revenue per page view, and total revenue from sales on your site.
Search (legacy)
The Search report (legacy) shows the terms used in a site search that returned no results.
You can sort by the following criteria:
- The term that was searched
- The number of times a term was searched
- The number of unique visitors who searched for a term
AB tests
The AB tests report shows the results from the latest Product Recommendations A/B test on a site and past tests.
You can do the following in the AB tests view.
- Filter the results to specific dates.
- Set the control group.
Note
A/B testing supports up to 5 groups, so you can view A-to-E testing.
- Select a specific currency that you want to examine.
- In the table, you can view and compare key metrics on the site between test groups, such as:
- Visitor sessions
- Site revenue
- Conversion rate
- Average order value through average selling price and units per order
- Recommendation engagement and revenue
Daily charts
The Daily Charts view shows graphs for revenue generated, visitor-to-order conversion rates, average order value, average selling price, average number of units sold, and number of sessions per day on the site.
Uplift charts
The Uplift charts view shows graphs displaying the revenue, conversion, and average order uplift value for a chosen test group versus the control group for the selected date range (which is cumulative) and over the past 7 days.
Summary
Algorithm performance
The Algorithm Performance report displays detailed performance metrics for each algorithm configuration across widgets and pages. Use this report to compare how individual algorithms contribute to clicks, purchases, and revenue, and to identify which configurations perform best on each page type.
You can filter the report by the following options:
- Time range – Edit the time range for the report.
- Currency – Select the currency in which revenue displays.
- Widget – Restrict the report to a specific widget. All Widgets is the default selection.
You can organize the data by the following columns:
- Algorithm – The algorithm configuration that generated the recommendations.
- Served – The number of times recommendations from the algorithm were served to visitors.
- Clicks – The number of clicks on recommendations from the algorithm.
- Items bought – The number of recommended items purchased.
- CTR – Click-through rate. A percentage worked out on the sum of clicks divided by the sum of impressions.
- CTP – Click-to-purchase conversion. A percentage calculated on the sum of purchased recommended products divided by the sum of clicks.
- Revenue – The revenue generated from purchases of recommended products served by the algorithm.
- Page – The page type where the algorithm serves recommendations, such as Home Page, Product, Category, or Search Results.
- Custom page – The custom page where the algorithm serves recommendations, if applicable.
Download the report
Click Export CSV to download the report data as two CSV files, one containing Impressions and Clicks, and one containing Revenue broken out by Currency.
Note
The exported CSV files might include additional columns, such as Hints and Expressions, that do not display in the report. The report shows only the most relevant columns. The CTP column appears only in the report, not in either exported file.