How to Surface Products of Interest in Key Category Pages for Returning Visitors

In this article, we explore how merchants can use Nosto’s Category Merchandising product on their category pages and incorporate an added layer of personalization to their merchandising strategy. To illustrate the impact, we will focus on a segment of Returning Visitors who have yet to convert to customers and who have only left behavioral signals at this stage of their customer lifecycle.



When connecting and engaging with a retailer, 98% of all signals left by shoppers are based on their behavior, while only 2% are transactional. This is valid for all segments of existing customers – but if we isolate segments of non-converters, 100% of all signals are behavioral-based.

With a commerce experience platform for digital retail (like Nosto) that are geared to capture and leverage both transactional and behavioral signals, it’s possible to create relevant shopping experiences using product recommendations or content personalization across an entire store at any stage of the customer lifecycle.

However, until now, there were limitations with optimizing or personalizing category listing pages due to complexity of implementation and lack of technical solutions. These pages were, ultimately, limited to actioning merchandising strategies.

In this article, we explore how merchants can use Nosto’s Category Merchandising product on their category pages and incorporate an added layer of personalization to their merchandising strategy. To illustrate the impact, we will focus on a segment of Returning Visitors who have yet to convert to customers and who have only left behavioral signals at this stage of their customer lifecycle.

Enhancing Category Pages: Best Practices Using Personalization and Merchandising

Before we personalize the sorting of category pages, let’s explore some of the best practices we’ve seen in other corners of digital retail. To start, let’s look at Amazon – a company that leverages massive amounts of data and behavioral signals to create a clearly visual personalized experience.

Until now, Amazon has decided to split their personalization and merchandising strategies, and it remains quite common to see dedicated placeholders for personalized product recommendations above a category page or at the bottom of a list page when users are done scrolling.

In one example, a product recommendation placeholder invites the users to discover the most gifted products of the explored category, resulting in a diverse selection of cooking books, ranging from ’50 Ways To Eat Cock’, ‘The Meateater‘ to ‘What the Fxxk Should I Make for Dinner’? The title itself sheds some light on the algorithmic logic running behind the scenes.

Let’s see how this evolves when a given user might not be keen to follow the recommendations they’re shown and, instead, look for very specific items in this category, (such as vegan cooking books). The sorting of the category page itself lists various items, using product popularity as an attribute for sorting as indicated with an orange ‘Best Sellers’ badges hover beside the image thumbnails.



Incorporating a Layer of Personalization in Your Existing Merchandising Strategy

If users start browsing and eventually revisits the same page after exploring a few product detail pages for Vegan cooking books during their initial visit, the highlighted items will now be tailored to that user’s personal interests along with an adjusted title, ‘Related to items you’ve viewed’.

Note how the product listing itself, in both cases, doesn’t seem to differ with The Meat Eater still listed first.



Combining Personalization and Merchandising

While the above implementation can be executed by using separately Nosto’s Onsite Product Recommendations & Category Personalization, we will focus on combining these two approaches and create personalized listing page. For the purpose, we will create a sorting rule to promote product by popularity and blending a layer of personalization into it for a truly unique experience for each shopper.

First, create a new sorting rule, choose Browsing History in the Personalization panel and give it a weight of 100%.

This will enable Nosto to display already viewed products in relevant categories when users with a Browsing History will visit those category pages.

Second, add an attribute to sort the selection of products under Configure Sorting.

For example, you can select Conversion in the list of all attributes under Performance Metrics, which will offer a way to promote or demote products based on their conversion rate and give it a weight of 100%.

When set up as such, the end result blends both the personalization & the merchandising sorting rule for each shopper with a browsing history. For example, if a user has viewed multiple products in a given category in her or his first visit, she or he can be exposed to those products that will be prioritized whenever she or he visits one of the concerned categories. The magic trick lies on Nosto’s ability to also sort these previously considered products in order of priority by respecting the merchandising sorting rule applied to the desired categories.



Category Selection and Segment Targeting

You can select one (or multiple) categories and set the rule for the segment of Returning Visitors by creating a new Sequence.

In this example, I create a new sequence and choose Returning Visitors as a segment and Personalized Most Converting Products:

And we’re all set! All returning visitors who have a browsing history will be served by a unique selection of products, tailored to sort products that convert the best while prioritizing already viewed products.



Ready to Optimize for the Best Performance? Time to A/B Test

Unsure if you should mplement such a sequence on a given category for your Returning Visitors? You can use Nosto’s A/B Testing & Optimization to test multiple sequences against one another.

Under a targeted category, select the relevant segment in the dropdown menu and add any variation by clicking + Add Variation on the top right menu.

Once the variations are created, you can select them in the table and click “Go to test”, which will redirect you to the test creation wizard, pre-selecting the right category,

Choose Sequence as Test Type, a specific optimization goal such as Conversion Rate and evaluate if this resonates better than a different custom rule or a more generic Sort by price ascending/descending or Sort by New Releases, for example.