Build a Recommendations experiment

  • Updated
Optimizely Web Recommendations is not currently included in the packaging for new customers. If you are looking for personalization solutions to use with Optimizely Web Experimentation, the Optimizely Customer Success Managers can help review the portfolio of tools to find the right fit for you.

You can create a recommendations experiment and add or edit your site's content, product, or category recommendations.

Optimizely Adaptive Recommendations is a recommendation engine that lets you introduce visitors to products and content based on browsing behavior. 

You can test different recommendation algorithms and measure the impact in real-time with Stats Engine.

Recommendations work with other Optimizely Web Experimentation products for better experimentation and targeting.

You can use recommendations in your experiments by:

  • Creating an experiment that targets the pages on your site where you deliver recommendations.

  • Adding an audience to specify who sees those recommendations.

  • Verifying your recommenders are working properly.

See Optimizely Recommendations to learn more or speak to your account manager.

Create an experiment in Optimizely Web Experimentation

To create a recommendations experiment:

  1. Go to the Experiments and click Create New Experiment.

  2. Select A/B Test and name your experiment.

  3. In the Target by drop-down list, select Saved Pages.

  4. To add an existing page where you want your recommendations to display, search for pages in the Browse for Pages field or select a page from the list of recently created pages. If you want your recommendations experiment to run on a new page instead, click Create New Page to add one.

  5. Click Create Experiment.

Add an audience, metrics, and traffic distribution

  1. If you want to show your recommendations only to a specific group of visitors, you must add an audience. Click Audiences to begin the process. Learn how to create a new audience in Optimizely or add an existing audience.

  2. Choose metrics to measure the success of your recommendations. Click Metrics to begin the process. Learn how to create a new metric in Optimizely or add an existing metric.

  3. Use traffic distribution to determine what percentage of visitors see your recommendations. By default, Optimizely shows the variation with recommendations to 50% of your audience and the original variation to the rest. Click Traffic Allocation to adjust the distribution to show recommendations to a higher or lower percentage of visitors. Learn how traffic allocation works in Optimizely.

Next, choose an algorithm and decide where to show your recommendations.

Configure recommendations

Use the editor to choose a recommender and place the recommendations on your page.

  1. From Experiments, select your experiment.

  2. Select a variation.

  3. Click Create to make changes to the variation.

  4. Find and select the Recommended Products extension. This is built during the recommendations onboarding process.

  5. Modify the recommendations to your preference.

    • Click Header Text to modify the header.

    • Click Maximum Products to select the number of recommendations that display at one time.

  6. Click the Algorithm drop-down list to choose an algorithm for your recommendations.

  7. Decide where you want your recommendations to display. Click Insert After Selector to change the position of your recommendations module and add the selector that displays below the module.

  8. Choose the timing for loading your recommendations on the page:
    • Asynchronous changes – Let the rest of the page load while your recommendations load.
    • Synchronous changes – Let the recommendations load first before loading the rest of the page.

  9. Click Save.

  10. QA your recommendations experiment with the Preview tool. When it looks and works the way you prefer, publish your experiment.

Your recommendations are now live to the world. Refer to the Experimentation Results page to measure the impact on your key metrics.

Verifying Recommendations Setup

Before proceeding too deep into building recommenders, you should verify that the data you collect is what you anticipated. There are two ways to do this:

  • Look at the stats in the inspector interface.
  • Download the SummaryStats CSV file, which has more detail.

For example, a customer has 927,000 view events in a week. If a week's worth of traffic typically includes about two million view events, the customer loses about half their usual views with the recommender in place.

In this example, custom events are any user-defined events other than a view. These are usually things like add-to-cart clicks or social shares.

The item count represents the unique items included in last week's event. In this example, that number is 542. Determining whether this is a problem depends on context. If the customer's catalog contains 3000 items, less than 20% of the catalog was viewed during the past week. If, however, the catalog contains closer to 650 items, the item count is likely an error.

Visitor count is the number of unique visitors who visit the customer's site. Validation failures list the number of unavailable catalog items when Optimizely Experimentation tries to visit their pages.

Check these stats for anything that stands out or does not seem right when you set up your collection.

Ideas for recommendations

There are many ways to use recommendations on your site. Here are a few retail and B2B use cases.

Retail sites

  • In the checkout funnel, show accessories that complement the items a visitor purchases using a co-buy algorithm applied to add-to-cart events.

  • Product detail pages show alternative items related to the product a visitor is browsing using a co-browse algorithm applied to page view events.

  • Highlight crowd favorites on the homepage using the popular algorithm on view and buy events.

If you use Optimizely Experimentation to test on a checkout page, you might need to configure your site for PCI compliance.

B2B or lead generation sites

  • Show visitors whitepapers, infographics, blog posts, and other content based on browsing behavior.

  • Suggest knowledge-based articles or community posts.

Experimentation ideas

  • Collect different events, for example, add-to-cart, social share, interaction with a color picker, or description expansion. 

  • Experiment with the layout, positioning, and appearance of your recommendations and the number of recommendations shown.

  • Experiment with the events that go into the recommenders. A popularity algorithm based on view events delivers different results from one based on add-to-cart events.

  • Apply different filters. For example, you can compare an unfiltered recommender with a similar price range filter, where the item viewed sets a range for the recommended item price.

  • Experiment with the effects of different algorithms. Ensure that the input types match. You can compare the different results delivered by a co-browse, co-buy, or semantic algorithm.