Product Recommendations — Filters

When you configure product recommendations, whether it is on your webshop, in Triggered Emails or Newsletter Content, Filters allows you to select specific groups of products based on brand, colour, category, price range, and more depending on the data available about your products.

Filtering the product recommendations

The product recommendation configuration can be divided into two concepts:

• Filters
• Strategy

The strategy defines if the recommendations should be retargeted products, top products, most bought, and so on. Based on the chosen strategy, Hello Retail automatically finds products that are relevant depending on if the recommendations are on the front page, a product page, etc. You can easily stick with only choosing a strategy and leave out any filters, and you are guaranteed relevant recommendations leading to more and larger sales. 

But if you want to fine-tune the recommendations a bit more, you can apply filters. When you apply a filter, the chosen strategy will still define the overall selection of products to be recommended, but the filter will narrow the selection down.

The concept explained. Let's take an example:

Recommendations on a product page will usually be configured with the Related products strategy (often it's a bit more complex, but this is just an example). This strategy finds products that Hello Retail consider related based on customer interaction, for example, products that are often viewed shortly after or before the current product. So, if a customer is looking at a shirt, related products will to a high degree be other shirts, but it could also be a pair of pants that are often viewed together with the shirt. Since the products are based on customer data, the strategy will always find products that are related and will get the customers browsing more relevant products.

But you might want to narrow the recommendations down a bit more so that related products for a shirt are only shirts. Or maybe even only shirts from the same brand. For this, you can use Filters.

Filter example:

A filter is configured in three steps, as you can see in the screenshot below.

In this example, we have applied a filter by brand. We have then chosen that the brand should match the brand. To be more specific, the brand of the recommended product should match the brand of the currently viewed product. So, if the customer is viewing Hugo Boss shirts, Hello Retail will find related products from Hugo Boss.

Let's add another filter:

Here the recommended products will both match brand and hierarchies (which is the definition Hello Retail uses for the category). So, with the same example of a Hugo Boss shirt, Hello Retail will recommend other Hugo Boss (matching the brand) shirts (matching the hierarchy).

You can apply as many filters as you want, but of course, adding more filters will restrict the available choices for the Hello Retail algorithm and at some point, it might be too restrictive. Have in mind, that the Hello Retail recommendations are finetuned to find relevant products based on data gathered from all customers and visitors, and we do that pretty well, if we may say so ourselves. What one thinks is relevant might not match the data, but we also appreciate the insight a shop owner has from pure experience — therefore, a healthy mix of data-driven automated product recommendations sprinkled with some guidance through filters might just be the magic harmony that brings it all to the next level.

Global and Source Filters

You can add Filters at two different levels for your Product Recommendations. Either you can add an overall Filter covering all products that will be shown in a row of product recommendations; Global Filter. Or you can add a filter only covering a specific step, or source, in the recommendation strategy; Source Filter.

With the two levels of filters, you can really customize which products will be shown to the customer. Maybe you would like all recommended products to be of the same brand, but only the first two, to be the same hierarchy — for example, if the customer is viewing a Hugo Boss shirt, the first two recommended products could be other Hugo Boss shirts, but the next could then be any Hugo Boss product. If so, you set a Global Filter to match on brand, the first source (in the strategy) to match hierarchy and set the max. products to 2. Then you add another source and apply a filter to not match the hierarchy.

In general, Filters, are not a requirement to get relevant product recommendations. Just setting the right strategy, depending on the location of the recommendations, will get you relevant product recommendations right away.  Remember, Hello Retail constantly keeps track of all customer interaction with your webshop, so that product recommendations automatically are relevant and based on the statistics of all the available data. Filters are a way for you to tailor the recommendations towards your specific wishes based on your insights and experience.