Badges are visual labels on item tiles that encourage purchasing by highlighting important information about the item. They help shoppers quickly identify popular, new, or trending items, enhancing their shopping experience.

Merchandisers often struggle with badges due to the need for consistent application, timely updates, and identifying which badges effectively drive sales. Certain badges also require specific data calculations, which many existing tools cannot automate.

Constructor's AI-powered badges algorithm automates this process, ensuring badges are applied to the right items at the right time, ultimately boosting KPIs and aligning with a company’s strategy.

ℹ️

An important note on badges

At this time, Constructor only supports badges on item listing pages (PLP) for Search and Browse. Badges for Recommendations and on item description pages (PDP) will come at a later date.

Key benefits

• Encourage purchasing.
  • Badges highlight specific item information, such as popularity, newness, or trending status, which can motivate users to purchase.
• Enhance the shopping experience.
  • Badges help users quickly identify key items, improving their overall shopping experience by making it easier to find popular or relevant items.
• Automate badge management.
  • For merchandisers, badges simplify the challenging task of managing badge application. Constructor automates consistent and timely badge updates.
• Boost key performance indicators (KPIs).
  • By effectively highlighting items, badges can contribute to increased sales and align with editorial strategies, positively impacting KPIs.

Types of badges

Constructor calculates and applies two different types of badges: global and context-dependent.

Global badges

Global badges are calculated based on item performance across all search queries and browse categories.

For example, an item might earn a Bestseller badge because it has the highest sales across a company’s site, regardless of the specific query or category. For instance, a particular sneaker style might be the top seller across all categories, making it eligible for a global Bestseller badge.

Context-dependent badges

Context-dependent badges are calculated based on specific queries or categories.

For example, an item might earn a Trending Now badge in the "Mobile phones" category because it’s performing exceptionally well in that category, even if it’s not a top seller across the entire site. For instance, a mid-range phone has recently seen a surge in purchases and clicks within the "Mobile phones" category. The item would be eligible for a query-dependent Trending Now badge.

How Badges works

Constructor’s badges algorithm ensures that the most relevant items receive the appropriate badges. The process can be broken down into three main stages: selecting candidates, ranking candidates, and combining badges.

Stage 1: Selecting candidates

First, Constructor identifies items eligible for badges based on specific criteria tailored to each company.

For instance, to qualify for a Bestseller badge, an item must exceed a certain purchase threshold. This threshold is unique to each company and aligns with the company’s KPIs and business strategy.

Here’s an example: A luxury retailer wants to incorporate Bestseller badges on their search and browse pages (categories and collections); however, to avoid too many items qualifying for the badge and potentially cluttering a page, the company sets a high threshold to ensure items maintain an exclusivity when receiving the badge. This means that only items with the highest number of purchases would be identified by Constructor as eligible for the Bestseller badge.

Stage 2: Ranking candidates

Once Constructor identifies eligible items, they are ranked based on several factors, including recent purchase frequency, price, and search result position. This ranking helps determine which items should be prioritized to receive the badge.

Here’s an example: Consider a scenario where two items meet the criteria for a Bestseller badge. If Item A has been purchased 50 times in the past week and appears in the first position on the first search result page, while Item B has been purchased 30 times in the past week and appears on the third page, then Item A would be ranked higher and would likely receive the badge.

Stage 3: Combining badges

In cases where an item qualifies for multiple badges, Constructor applies a priority list to determine which badge should be displayed. This ensures that only the most effective badge is shown to the shopper.

Here’s an example: If an item qualifies for both the Bestseller and Trending Now badges, the algorithm may prioritize the Bestseller badge because it carries more weight in influencing shopper purchase decisions.

Example of Badges in action

Let’s look at an example of how Constructor determines the application of a Bestseller badge.

First, Constructor and the example company set a threshold for the number of purchases (or add-to-carts) needed to qualify as a bestseller. In this example, the focus is on the Men’s Shirts category, and the company identifies the threshold based on add-to-carts and purchased items.

Based on this, Constructor identifies five (5) candidates for the Bestseller badge on the first page of results.

Next, Constructor identifies which candidates ranked highest in results. These items will receive the Bestseller badge.

Now the Bestseller badge will appear for the final candidates on the company’s website. Here is how it might look:

How to get started

Constructor’s standard badges—Bestseller, Trending Now, and New—are available to all companies using Search and Browse.

The Constructor team is hard at work developing additional badges, such as social proof (e.g., In 17 Carts, 24 Bought Today, etc.) and attribute-based badges (e.g., Eco-Friendly, Premium Materials, etc.) which will be available for those customers using Attribute Enrichment.

To get started with badges, please reach out to your Customer Success Manager (CSM). The Constructor team will set up your Badges pipeline to analyze your data and calculate thresholds for your badges. This can be tested in your testing or development index before being enabled for production.