An explanation
YesPlz.AI, May 2023
Product tagging for eCommerce is the single most important factor for brands looking to enhance their personalization, offer advanced product filtering, and create a better search experience.
Fashion tagging is a key component of any eCommerce. More advanced tagging uses AI (instead of manual labeling) to recognize product attributes in an image and NLP for text, to create rich product tags that are then used to build personalization. Fashion tagging is the foundation for any personalization strategy.
With new fashion tagging tools available, it’s important to take a critical look at how you can use these new tools to improve your eCommerce. For example, how was each product tagging tool built? What features does it have? Does it use fashion-trained AI?
In this article, we’ll go over how YesPlz built its product tagging for eCommerce, and why it stands out from the rest:
Before training our AI, we sat down with real shoppers to define the product attributes that matter most to them. We captured the key product attributes that they value, such as:
We learned that shoppers are very clear about the body parts they want to flaunt or show off, and require silhouette-based product tags in order to quickly filter the products. In addition, shoppers valued attributes related to design such as color and pattern.
The human element behind AI technology is key to ensuring that we’re creating tools that accurately reflect shopper needs. First, we accurately assess user needs, then create technology that meets them.
We’re a team of engineers and fashion analysts with a deep background that’s both technical and fashion-based.
Our product tagging for eCommerce is trained to understand fashion datapoints, creating the foundation for building fashion-first personalization.
And, we double-check the quality of our AI’s fashion tagging with human-based analysis from experts in the fashion industry. Our fashion data annotators come from Parsons School of Design and Fashion Institute of Technology, and know all the latest fashion trends and terminology.
YesPlz’s Product Tagging for eCommerce Categories:
YesPlz can tag a wide range of product categories, including women’s tops, jeans, jackets, shoes, and more. Check out the full list below:
Let’s take a deeper look at the product attributes tagged:
YesPlz’s product tagging for eCommerce can tag:
Silhouette and Design: Depending on the product category, YesPlz will tag relevant silhouette attributes. For example, if we look at women’s tops, we see that YesPlz tags:
For women’s jeans, the product attributes tagged change to be relevant for jeans, such as:
Vibe/Occasion: Based on YesPlz user research, we learned that users also value vibe and occasion when shopping for products, making it a necessity for product tagging for eCommerce to provide accurate tags.
But, historically, tagging for vibe and occasion is difficult because of differing opinions on what constitutes a vibe. For example, a “romantic” dress could have different interpretations depending on different backgrounds.
However, YesPlz is able to overcome this problem by using user research to understand what exactly constitutes a vibe, and training our AI to fit the definitions.
YesPlz AI also tags for colors, patterns, and details, making it easy for shoppers to filter:
It’s easy to plug into the YesPlz tagging process, because all you need to do is provide us with your product feed, and we’ll handle the rest.
What actually happens when we tag product attributes?
We use a 4-step process:
1- The key product attributes are defined by user interviews (as explained above)
2- We train the AI with both image and text
3- The AI tagging is reviewed by fashion experts
4- We share the final tagging results with you, the fashion retailer
Product tagging for eCommerce doesn’t need to be complicated or expensive. With YesPlz AI tagging, your eCommerce can create the building blocks for personalization.
Are you curious to read more resources?
Check out:
Schedule a 20-min consultation to learn more