Brand Matching: How to Use AI to Manually Set Merchandising Rules in Fashion and Cosmetics

A few years back, when setting up product recommendations on sites that sold particular brands (especially on beauty & cosmetics sites), we came across the same problem: Some brands have strict guidelines for retailers which limit the cross-selling of products to the same brand viewed on a product detail page (PDP). For example, if a customer is viewing a fragrance by Acme Cosmetics, the site is only permitted to recommend other Acme Cosmetics products as cross-sellers. However, when a customer views other brands like Watercooler Makeups, the site can recommend any product or brand, without restrictions.

One particular brand was so adamant about this rule that I remember once sitting in a meeting with a client of ours and advising them to use Advanced Dynamic Filters (also known as “If This Then That” rules) to avoid any friction with the brand’s representatives. The dialogue was pretty much:

Nosto Client: “How did you know about this?”
Me: “Well, let’s just say we’ve come across it before.”

Enabling Advanced Dynamic Filters in Brand Matching

While brand-matching is sometimes a must-have, dynamic filters also work as an effective merchandizing tool regardless of the vertical. This is because they allow for adjusting the algorithm and the output (recommended products) by adding the merchandizer’s business strategy into the Artificial Intelligence. Let’s call this: Holding the reigns of Nosto’s Ecommerce Intelligence Engine.

This can be taken to the next level with advanced filters that also support advanced merchandising and dynamic up-selling. An example of this is when a viewed product falls into a certain price range. As seen in the example below, recommended products (which fall into a specifically designated price range) never feature the cheapest budget alternatives.

In yet another example, our friends at Costo use the same rules to support a relatively straightforward and important business target: Purchase profitability nicely combined with their brand.

Some of their product ranges have detachable and changeable bobbles and every beanie comes with one. However, as one of Costo’s brand values is simple customization, they have a huge range of add-on bobbles, each with a relatively nice margin compared to the main product. When a shopper is viewing a product which come with a bobble, Costo is able to promote bobbles as the primary or only recommendation. However, when viewing a different product range, the recommendation automatically falls back to recommend alternative products (not bobbles).

To take it even further, Nosto’s Ecommerce Intelligence Engine picks the most relevant product options for each shopper by also removing unavailable items, if and when they are sold out or removed from the inventory. As a result, there is a drastic reduction of manual pinning and matching which saves a huge amount of time & effort.

As a third and final example, The Sports Edit and their merchandizers know that some ranges are always a perfect match for each other. By tagging ranges such as “workout tops” and “tights” together, they ensure that key inventories are always matched while elsewhere, Nosto’s Artificial Intelligence is in full and autonomous control.

Dynamic Filters = Unlimited Brand Matching Opportunities

In a nutshell, advanced dynamic filters allow merchandizers to set up rules that fit an almost infinite number of scenarios. Read more about how to get started on our support pages or ask us if we support a scenario you have in mind.

Watch a 4 minute demo
Let us demonstrate why over 2,600 leading retailers worldwide have chosen Nosto’s AI-based personalization platform.