The Art of AI-Powered, Customer-Centric Item Set-Up

For the retail merchandiser, the process of onboarding a new product so that it is set up to sell effectively (and rapidly) can be a complicated one. There’s a tremendous amount of depth that a retailer has to consider that may not be obvious right off the bat. For instance, how will the retailer index for the search of this product?; what is the product assortment strategy?; which page should this product show up on online?; how will filters and facets be deployed for better site navigation?; and so on. 

Although the process of accurate and optimized item set-up for a merchandiser’s new products may appear to be tedious – and with good reason (we know there’s usually a lot of products) – there are ways this onboarding can be improved and made more simple – with the power of AI. 

In this guide, we’ll take a look at some of the main culprits that can cause issues during item set-up, how to improve upon the process, the benefits of correct item set-up for retailers, and how Lily AI approaches the art of AI-powered item set-up with a focus on customer-centric merchandising. Plus, we’ll share results from Lily AI’s partnership with JOOR, a premier wholesale management platform for brands and retailers.    

As a merchandiser, we know you’re likely to have your own unique process for climbing the mountain that is item set-up. But this climb doesn’t have to be a slow one. Perhaps it’s best to avoid this altogether. Why? It’s important to note that a slow, manual, legacy process has implications that go beyond just feeling like it takes forever. 

If a product can’t get on the shelves quick enough due to slow item set-up (manually attributing products) then you might miss out the opportunity to capitalize on the current trends. If you can’t get the item out there until that trend has long passed, then time, materials, and money are most likely going to go to waste. Mistakes at item set-up impact the entire retail value chain.  

Why not make the climb easier? With the help of AI and automation, the process can move along far quicker when products aren’t needing to be attributed manually. You can breathe that sigh of relief knowing that the products are being attributed not just more quickly, but accurately to boot.

When products are attributed with generic, legacy, or industry-focused language, as opposed to customer-centric attributes that speak in the language of the consumer, these products are far less likely to reach their potential to sell at full-margin. 

Generic product attributes determined by manufacturers or distributors make it harder for shoppers to actually find what they’re looking for, because these out-of-the-box terms don’t typically match up with how a consumer actually searches for products. In fact, according to The Harris Poll on behalf of Google Cloud, more than $300 billion is lost to search abandonment – when a consumer searches for a product on a retailer’s website but does not find what they’re looking for – each year in the United States alone. Thin and inconsistent product details that start at item set-up have clear and cascading effects down the line.

AI-powered, customer-centric product attributes are a key and important part of building a taxonomy that converts and creates a strong item set-up process that gets products on the site at the right time and does what you want them to – sell at full price.

One of the best ways to effectively plan your assortments in less time is by implementing auto attribution at scale. When product attribution is automated, the onboarding process can be done rapidly, no matter how many items need to be attributed during the set-up. 

Auto attribution saves both money and time by giving time back to merchandisers so the focus can be put on other important areas – i.e., assortment planning, sales forecasting, trend analysis, and more – vs. the time-consuming and manual tasks of attribution when product onboarding isn’t automated at scale.

Some of the core benefits of auto attribution in the item set-up process are:  

• Increased Speed & Accuracy: Dramatically reduce time by taking the focus off of manual product tagging.
• Measurable Cost Savings: Better optimize your merchandising teams’ time and expertise away from cumbersome manual tagging.
• Reduced Errors: No more incorrect product listings that can hurt site search, SEO/SEM, or lead to inaccurate product analysis. 
• Optimized Assortment: Improve your decision making with greater tagging detail and anticipate gaps in your assortment earlier in the buying process. 
• Faster Time to Site: Accurately tagging products earlier in the process decreases the amount of time it takes for a new product to go live on your site.
• No More Delays: No longer rely solely on brands or distributors for product details, or delays in waiting for product information from them.

Lily AI, trusted by global retailers, brands, and industry leaders like Bloomingdale’s, the Gap, and thredUP, helps retailers and merchandisers with our AI-based technology to speed up the item set-up process and ensure the accuracy of product attributes.  

We enable merchandisers like yourself to work smarter, so they can spend their time on other things. By powering rapid product onboarding, Lily AI can help to eliminate days of manual work, while setting you up with a customer-centered product taxonomy that will drive better decision-making for years.

Lily AI works to take your current data to either improve or replace some of the existing processes or workflows. For example, we know as a retailer that you’re already going to have assets present – image and text – but this still may not be enough to offer something for sale online.

What we do to get you ready for the next step of actually selling those products online is take the data feed (the images and text) provided to us and break it into batches of similar products. After this, we will:

• Run the data through two different AI-powered processes:
  • Image recognition
  • AI-based classification

• Run the text through text recognition and Natural Language Processing (NLP) pipelines. 
• Combine and leverage the image and text together (image- and text-based predictions together get better results than they do separately). 
• Loop in the human element of our domain experts team, who may review a selection of the data for better accuracy.  

And most importantly, when this is all completed, we give you the most accurate and relevant product attributes back, often going from 4-5 attributes per product to 15-20 attributes per product – perfect for ensuring that shoppers can now find what they’re looking for, no matter how they search.  

JOOR, a leading digital wholesale platform and data exchange, works with Lily AI to help JOOR platform users significantly reduce the time they dedicate to manually tagging products. 

What was once a time-consuming process has now been simplified thanks to auto attribution. JOOR customers can have their entire assortments tagged with Lily AI’s AI-powered robust taxonomy of over 15,000 attributes, which is then mapped back to any of the retailer’s existing tags.

How Does Auto Attribution Work? 

The item set-up process for merchandisers through JOOR follows these steps:  

• All tagging takes place directly within the JOOR platform to allow for ease of use.
• Existing retailer attributes are mapped with Lily AI tags, providing a much more in-depth and accurate product taxonomy.
• Products are tagged in real time as they are being added to assortments within the platform.

Merchandisers like yourself can edit attributes within the platform – i.e., the ultimate amalgamation of art and science.

Take an In-depth Look 

From JOOR, to the retailer, to fully supercharged with Lily AI.  

From Joor…

1. All of the products purchased through the JOOR platform are quickly and accurately tagged with Lily AI’s auto attribution. This will apply to any product purchased through the JOOR platform from a subset of the Lily AI proprietary taxonomy. 
2. This enriched product data with Lily AI’s auto attribution will be ingested directly into your PIM systems when you export the product from JOOR. 

To the Retailer…

1. Merchants can now internally assign additional attributes based on your taxonomy. 
2. For many of you, this is often the time your photoshoots are taking place prior to uploading your imagery to your e-commerce platform. 

To Supercharged with Lily AI 

1. You can increase product attribution coverage for products purchased on JOOR and your entire remaining catalog with Lily AI Foundation and Enterprise packages. Product photography and merchant tags can be sent to Lily AI directly to be further attributed. 

With Lily AI Foundation or Enterprise your entire product catalog and historical data can be attributed to supercharge your existing tech stack to enhance site search, SEO/SEM, item set-up, personalization, and demand forecasting.