The Power of Our Product Recommendations System

Product Recommendations are a must-have in today’s online market place, and their power to drive conversion is proven; so naturally, you want them on your site. That’s why Personyze employs the latest recommendations technology in our completely customizable recommendation widgets, and can display them anywhere on your website, and even in your re-marketing emails. Each customizable and responsive widget allows you to choose a different algorithm, so you can show what makes the most sense for the page where you’re using it.

Some Key Advantages of Personyze Recommendation Software:

  • Setup: Integration of your Google Product Feed (or other) is included in our setup kit, when you sign up.

  • Analytics: All product interactions can be set to send to Google Analytics, or other systems, in addition to our own internal analytics

  • Integrate with Other Systems: Trigger emails or other business logic within Personyze, or in 3rd-party providers based on visitors’ interactions with recommendations

  • Quick and Easy: Can be live and running in one day

  • Social Dimension: Can display how many viewed or bought each product in the recommendation

  • Variety: View again, buy again, cross-sales, up-sales, many algorithms are available

  • CRM Data: Use CRM data, to know interests from off-line sales and other sources not otherwise available

  • Social Data: Filter based on demographic and other data from Facebook

  • No Limitations: Our recommendation software works on any platform or website. If your site displays in a browser, we can display recommendations on it.

Personyze uses a “hybrid” recommendation engine, which means that it combines crowd wisdom (what users in general are doing) with knowledge about the individual customer’s ranked interests, gained from previous interactions with your site, your products, and even offline transactions from your integrated CRM data. Exactly what recipe you use to turn this data into brilliant recommendations is entirely up to you, as you select the algorithm, and the variables.

After that, simply customize the look and feel with our code-free editor to match your website, determine which page or page-groups to show it on, and your recommendations can be ready to deploy in minutes. With our recommendation software’s testing capabilities, you can also test different algorithms, to ensure you use only those with the highest performance metrics.

We make personalization easy. In many cases, we may set your recommendations up for you, as part of a free POC. We can do this because the process is so easy, once we have the Google Product Feed, we can have a recommendation widget up and running in an hour or less. With a free trial and no commitment, why not try it out?

More About Our Recommendations System

Recommendations are an extremely versatile functionality, and one of the most popular on the Personyze platform. You can use them to show products that are cheaper, products that are popular, products that match visitor interests, and very often a combination of these and other factors.

The key to effective recommendations is having as much data as possible to determine user interests, and then choosing the right algorithm for the particular application. Typically, each type of page, such as product pages, home page, and cart page, each require a different algorithm, for different purposes.

A typical use case is as follows:

  • When visitor arrives at the home page: You may want to display recommendations based on their most recent interest, combined with what is most popular overall on the site, or among their same demographic.
  • On product pages: You may want to show products that are similar to the item currently being viewed, or perhaps the cheapest items of the same category.
  • On your cart page: You may want to show items that are typically bought with whatever is in their cart, for cross-selling purposes. This could include accessories, or insurance packages for given products.
  • In your re-marketing emails: You may want to recommend products similar to what they had in their cart, before they abandoned it. If they didn’t abandon, then you can recommend things accessories and other cross-sells.

Whatever type(s) of recommendation you need, Personyze recommendations software has the data and the algorithm that will drive engagement and conversion on your company’s site.

PERSONYZE RECOMMENDATIONS IN DIFFERENT SCENARIOS

Slider 7_Wisdom of the Crowd RecommendationsProduct Recommendations Engine on E-commerce Website: Personyze has a variety of algorithms for any need you may have, your recommendation system widgets can show the most bought or viewed items based on visitor interest, what was most bought together based on product currently viewed, what is most bought together with an item just bought, on a thank you page for cross-sales. There are more algorithms for accessories for an item you bought, new stock for item you bought that typically is refilled, such as makeup, coffee, printer paper, and other household items. You can also include these and more in your re-marketing emails, such as items similar to what you left in cart.

 

Content recommendation_articles to read

Content Recommendations Engine on a Content Website: Our powerful content recommendations can show the most popular content based on visitor’s demographic, most read content based on category in which visitor has shown interest, what to read next after this article, and people who read this also read this. You can also show new content from the last time you visited the site, new content just for you based on most recent interest, and new content from your favorite author or reporter.  All of these recommendations can be shown in widgets on any page or group of pages on your site, or in your marketing emails.

 

Product Recommendations Engine in Re-Marketing Emails: Once your product catalogue is integrated with our system, Personyze can automatically send re-marketing emails containing recommendations based on abandoned carts. Other Re-marketing recommendations might include re-stocking for products you should buy again like makeup, coffee and other household items. They can also show things on clearance, accessories to items recently bought such as a new mouse for a computer, or a case for your new tablet.

Content recommendation_video to watchProduct Recommendations Engine On Video-Based Sites or Vlogs: Interest widgets take note of categories that account for the maximum interactions (shares, likes, comments), videos previously viewed, and most watched sections to dynamically recommend videos that make sense, given the past history of the visitor. Recommendations can be based on actors the visitor likes, their most-watched genres, combined with what others with similar interests like to watch, most-watched movie from each genre, and also what others of their same gender, age, and other demographics are interested in.

The Back End

The equation to create great recommendations requires not just a powerful recommendation engine, but also a product feed that ideally has interest tags or description keywords that allows our system to easily know what interest category a product represents, and what other products it’s connected to. We can always work with the bare minimum, but the better the data is organized, the better the recommendations will be.

Another important way to optimize your recommendations is by setting up product interactions, and setting your site to update Personyze about “interest events,” so that our recommendation system knows as much as possible about the user’s interests. You can read more about setting up product interactions and other technical facets of recommendations on our wiki site.

The good news about these aspects is that they generally only have to be done once, and we help you complete the process as part of our standard on-boarding procedure.

Beyond these background elements, implementing recommendations with Personyze is as simple as choosing an algorithm, it’s variables and filters, and altering its look and feel. Sometimes, the customization process can be simplified by uploading the CSS file being used on your site. Even if you can’t do that, our template editor allows you to create any type of recommendation you want.