We would like to start integrating Recommendations. We went throught the documentation but nothing mentions how to send products to Optimizely. We know that we can use the events API to do this by tagging our Product Details page, but I don't know id there are specific tags we should use.
Solved! Go to Solution.
What you could do is set-up the different recommendations for the products as different variants, this is the most likely case I guess for what you're trying to achieve. This data is then accessed later by the experiment itself to make sure you can show the accurate product recommendations that you want to. The Optimizely REST and X API doesn't have the ability to specifically send information for products, what you can do is use the Optimizely X API for creating + updating experiments: https://developers.optimizely.com/x/rest/v2/#create-an-experiment
I think were not understanding ourselves. I'm reffering to get this up and running:
Specifically, the algorithms talk about co-buy. How does the stat-engin what products were bought? How does he distinguish a product from a regular content page?
Also, this document: https://help.optimizely.com/Build_Campaigns_and_Experiments/Build_a_Recommendations_experiment talks about a recommendend product extension. How does he know what is a product if there is no reserved tag or even for products and buying? How can we access it?
Optimizely X Recommendations is a separate product that you would need to purchase before the catalog can be connected. When you add it on we do a one-time setup to connect the events you tagged to the product catalog and fetch them in the extension. If you're already using Optimizely Experimentation you can add Recommendations by contacting your Customer Success Manager