The Big Picture:

eCommerce success for luxury brands revolves around how effective they are in bringing the offline shopping experience to online shoppers. A big part of delivering this online experience lies in the nature and quality of product data on-site. An AI-powered solution must go beyond generic tags and basic image descriptions to provide business-specific intelligence with rich, consistent, and accurate data.

Diesel started using Vue.ai’s AI-powered automated product tagging solution, Vuetag, in 2020 to streamline onboarding, tag their products, build custom solutions, and save time in the process.

22,300

Tags generated

130

Batches uploaded

The Problem:

Diesel had new products coming in every season that needed to be sorted, styled, photographed, and then sent to the team for attributes to be tagged manually. The challenges that the manual process threw up ranged from slow onboarding time to increased effort to inconsistency in the catalog from a lack of standardization.

To solve these issues, Diesel needed a tool that could:

Automate the
tagging process
with the capability
to integrate with
their existing
PIM tool

Recognize and
identify specific
attributes like
'iconic' logos that
are present in the
products

Standardize
the values and
attributes for
use across the
organization

Save time by
streamlining
the process
end-to-end

Diesel chose VueTag since it catered to their needs to optimize process efficiency and build customer-specific taxonomies while enhancing the quality and reliability of product data.

The Solution:

Summary: AI-Powered Tagging Solution

Automated tagging for faster product onboarding
AI-powered tagging for rich data and a scalable process
Standardized tags for clean, consistent catalogs

Automated tagging for faster product onboarding

VueTag saves 30 hrs of time per week per person.

Diesel implemented AI-powered instant tagging to get their products to market faster. VueTag ensured:

Product tagging is automated across attributes and categories, thereby reducing the time teams spend on tagging.

Products are tagged as soon as the images are made available, rather than waiting for a significant portion of the collection.

Review processes and feedback on tags is faster and easily done through the tool - helping reduce go-to-market time significantly.

VueTag generates tags with standardized information - reconciling and generating content in the process. Product data is generated with a combination of computer vision and NLP.

Images are from the VueTag Dashboard and may not represent the Diesel catalog

AI-powered tagging for rich data and a scalable process

450M product tags and attributes served with VueTag

Tags extracted by the tool helped Diesel enrich their catalog data. Downstream benefits included:

Powering filters on the website to assist user journeys

Enrichment of catalogs from other vendors, and

Using the extracted tags for analysis and forecasting

By shifting efforts from manually extracting product data to a streamlined workflow - teams now focus on informed decision-making. Diesel teams can review predicted tags and send their feedback back to VueTag networks, which improves the accuracy of predictions over subsequent iterations of uploads.

Images are from the VueTag Dashboard and may not represent the Diesel catalog

Standardized tags for clean, consistent catalogs

121 types of attributes recognized; 1043 different attribute values available with VueTag

Standardizing data across the catalog is essential for unifying the shopper experience across channels. VueTag:

Identifies missing data, incomplete metadata, or duplication of images across sellers.

Helps reconcile visually and textually extracted attributes.

Highlights inconsistencies and boosts the confidence and accuracy of attributes based on contextual cues from different sources.

VueTag takes into consideration the information based on historical product data or data uploaded from various sources. This helps standardize the information stored in the database about that product.

Images are from the VueTag Dashboard and may not represent the Diesel catalog

Summary: Customer-Specific Solutions

Integration
Custom
Taxonomy

Integration

VueTag can be integrated by retailers in a number of ways - from the dashboard, to APIs, to even integrating with their existing PIM system.

Diesel decided to integrate with VueTag via API with their incumbent catalog management & PIM system. A critical aspect of the integration was the standardization of product data.

Once integration & testing was completed, Diesel was able to automate tagging of all images in a seamless, hassle-free manner.

The integration process included the ability to customize & map the values of attributes to match exactly what the Diesel team wanted.

This included training the Vue.ai networks to recognize and identify logos & patterns which the Diesel team considered to be 'iconic'.

Custom Taxonomy

VueTag’s A.I. can be customized with new tags and rules defined by the retailer based on their business goals and priorities.

As a globally recognized brand - Diesel has many iconic logos. A key part of the automation process was detecting these logos on products automatically by the system.

Diesel used VueTag to build their own custom taxonomy and trained networks to identify the highly specific tags they required.

The system understands both text and image-based input from Diesel - identifies the specificities and tags the products in a consistent & easy-to export format.

A simple QA process allows Diesel to correct any incorrect tags while forming a feedback loop which in turn improves the network in an iterative manner.

Key Takeaways

Post-implementation of Vuetag, Diesel observed:

  • Faster product digitization process with less time wastage & shorter go-to-market period.
  • Decrease in the effort taken to manually tag products, and improvement in the efficiency of product onboarding.
  • Enhanced search and detailed filters for better product discovery as a result of the depth of product data.
  • Reduced upfront data cleaning work that is often needed to leverage and gain deeper insights from structured and unstructured data.
  • Better understanding of the volumes of data in the catalog and shoppers' interaction with the products with granular tags.
  • Optimized pages on the site that help teams make better-merchandising decisions and inventory planning.

Request Demo

Automated product tagging powered by A.I.

Custom demos typically last for 30 minutes. In this time, you can expect to:

  1. Get a in-depth understanding of automated product tagging solution and its capabilities
  2. Understand how other retailers are using the solution to achieve their business goals.
  3. Discuss your unique needs and challenges.

One of our experts will get in touch with you on your work email. Rest assured, we respect your privacy.

×

Your best source for all things A.I. & retail

News, ideas and updates for all your retail teams