A Recommendation Algorithm for Every Situation

Personyze recommendations utilize a wide variety of algorithms to show visitors what they’re most likely to be interested in reading or buying, by combining crowd data with tracked individual visitor interests, as well as other user data such as demographic and location.

Visitor interests are derived from how each visitor has interacted with the content or products in your catalog/feed. Each time a visitor views, clicks, likes, comments upon, adds to cart, or purchases a product or piece of content, this feeds into Personyze’s data and contributes to recommendations that increase engagement across your website, email, and app.

Personyze recommendations and all other Personyze features work on ANY ecommerce/website platform, including WordPress, Shopify, WooCommerce, Magento, and every other platform in existence.

Note: This list does not include every algorithm that is available with Personyze, but rather those which are most popular with our clients. If there is an algorithm you are interested in that is not found here, feel free to contact us to inquire whether it’s available in our system.

Recommendation Types

Product Recommendations On Website

Content Recommendations On Website

Product Email Recommendations

Content Email Recommendations

App

Recommendation Algorithms FAQ

What product data is required in my catalog for my recommendations to work properly?

The more data your feed contains, the more rich and dynamic your product recommendations will be. The most essential things to include are:

Product Title
Product Listing URL
Product Image URL
Price
Product Description
Interest Categories for each product (click here to read more about interests)
Product ID (SKU, or similar unique ID)

Some additional product data that are helpful to have are:

Sale Price
Cross-Sale SKUs
Up-Sale SKUs
Brand
Secondary Interest Categories
Seasonal product associations (sells more in the Summer/Winter)
Availability (In Stock or not)

For content, the requirements are fewer, and the key essentials are:

Content Title
Content Page URL
Content Image URL
Author
Content Description
Interest Categories for each article (click here to read more about interests)
Content ID (SKU, or similar unique ID), if applicable

You’ll also need to tell Personyze which column corresponds to which value, in your Personyze product feed.

Will my product/content feed or catalog work?

Yes, if your product inventory or content catalog exists in the form of any type of feed or spreadsheet, it’s possible to integrate it with Personyze. The default formats our system works best with are .CSV, RSS feed, .XLS, and similar.

Can I filter for demographic and other factors?

Yes, most recommendation algorithms allow you to add additional filtering, based on any of the 70+ data points which Personyze has on visitors, including demographics, location, CRM attributes, and more.

Can I show different algorithms on the same page?

Yes, you can show as many display widgets with as many designs and algorithms as you want, throughout your site, emails, and apps.

Can my recommendations be targeted, and seen only by certain visitors?

Yes, every recommendation widget can be targeted to individual audience segments, such as only returning visitors for Buy It Again recommendations.

How long will it take for Personyze recommendations to gather enough interaction data to be accurate?

This will depend on a few different factors, including your site traffic, volume of products or content, and whether or not Personyze is starting from scratch.

It’s possible to upload past interactions data, such as past ecommerce transactions of users, to give the algorithms a head-start, and if such data is available this is highly advisable. The Personyze team will help you with this process.