What is eCommerce product discovery?

Product discovery for eCommerce targets shoppers across the customer journey, providing solutions for all types of shoppers, including different types of shopper intent and search preferences. 

A complete product discovery for eCommerce solution includes: filter search, text search, product recommendations, and personalization.

The goal: to help shoppers effectively discover a product, without getting lost in the online store. With increasingly complex search journeys, a discovery solution can make the shopping process less daunting and guide shoppers to the products they love. 

The added value of product discovery solutions is proven: higher conversions, happier shoppers, and more repeat shoppers.

But, the market is currently fragmented, forcing retailers to purchase multiple discovery products from different vendors to retrofit a solution. 

The result?

• An expensive, non-cost effective solution
• No cohesive experience throughout the discovery solutions, for both retailers and shoppers

The 3 Biggest Problems with Current eCommerce Product Discovery Tools

1. Not customized enough for retailer-specific needs

Every fashion retailer has different needs–and requires a product discovery solution that’s completely customizable to their business. Other solutions lack customization, locking retailers into products that don’t serve them or their customers, and leaving retailers with a blanket, one-size-fits-all solution.

2. Fragmented search and recommendation products 

Which is more important to shoppers: visual or text search? Every shopper is different–some shoppers want to engage text, others with visual search–or begin with a faceted filter search. 

When eCommerce product discovery solutions are too narrowly focused, retailers miss the opportunity to spark discovery for every type of shopper during every part of the journey.

3. Complicated integration

Other eCommerce product discovery solutions require even more work from busy retailers to integrate. Each different product requires a different type of integration process–and managing multiple integrations is no easy task for any busy retailer.

But–what if there was a way to design a discovery system that covers visual search, text search, and filter search, as well as complete shopper data that could be used to optimize the discovery experience?

Schedule a free 20-min consultation

How does YesPlz compare to existing eCommerce product discovery solutions?

Deep Product Tagging

Deep product tagging is the process of using artificial intelligence to quickly and accurately label product attributes on an image. A single clothing category can contain 20 to 60 different attributes, which takes significant time to tag. By going beyond traditional tagging, we can use artificial intelligence to create precise product tags in milliseconds. 

Deep tagging is the foundation for product discovery for eCommerce: product tags are the building blocks for search filters, text search, recommendations, and personalization. 

The YesPlz Approach to Deep Product Tagging

YesPlz uses computer vision and natural language processing to create accurate product tags in milliseconds. Our AI-powered tagging can identify product attributes in:

• all types of image quality
• in both user-generated content and professional photos
• in photos with and without humans
• text information, such as product titles and meta data

And, YesPlz product tagging is based on real shopper preferences, making it easier to quickly define thematic attributes.

Fashion Filters

A fashion filter helps shoppers narrow down their clothing preferences based on size, color, silhouette, fit, or theme, making the shopping experience less overwhelming.

But, retailers are often stuck working with separate vendors for fashion filters, and end up with piece-meal, scattered filters that are confusing and inaccurate.

💡 Did you know?
YesPlz combines fashion filters with deep tagging, so shoppers always see on-trend, accurate filter options.

The YesPlz Approach to Fashion Filters

YesPlz AI has an all-in-one fashion filter (all powered by AI), that includes:

A.  Virtual mannequin filter: Intuitively design silhouette options on a virtual mannequin, and get instant results.

B. Thematic Filters: Shoppers filter by mood, vibe, occasion, or theme with over 20 thematic filters. 

C. Faceted Filters: Easy-to-use faceted filters, along with visual cues. Shoppers select multiple filters to build search results that reflect her preferences.

Product Recommendations:

Product recommendations are at the core of any successful discovery solution. With the power to inspire shoppers, getting product recommendations right is a must. 

A good recommendation should be relevant, take design and aesthetics into consideration, and show products across categories.

The YesPlz Approach to Product Recommendations:

• Similar Recommendations
• Frequently Bought Together (a.k.a Collaborative Filtering)
• Complete the Look

We make Similar Recommendations based on vibe and design to curate products that are similar–but different enough to spark engagement and interest. 

Vibe-based recommendations are particularly important, because they can make the difference between a relevant and irrelevant recommendation.

For example, if a shopper is browsing two different black t-shirts: one with a “workout” vibe, and another in a “party” vibe, the workout vibe will generate similar vibe recommendations, as well as similar silhouettes (loose, cropped). 

Frequently Bought Together (aka “collaborative filtering”) is based on the idea that similar shoppers have the same interests. By matching shoppers with similar interests, based on buyer behavior, and then generating recommendations, shoppers are placed into “cohorts.” 

But, a common problem with Frequently Bought Together recommendations is the “cold start” issue: without data about the product or shopper to begin with, it’s difficult to generate recommendations at first.

The YesPlz Approach to Frequently Bought Together Recommendations

YesPlz solves for the “cold start” problem by combining frequently bought together with similar recommendations. 

The power of a blended recommendation system allows retailers to use each of the product recommendation types to enhance the other’s qualities and create a product recommendation system that is fast, accurate, and shows off the depth of product catalogs.

YesPlz recommendations also include “Complete the Look,” an AI-styling tool that builds outfits for shoppers. We taught AI basic styling rules, based on real user interviews, then, based on the product category recognized by AI, these rules are applied to create unique, relevant outfits. 

Text Search:

Text search is an essential part of eCommerce product discovery. When shoppers type in their best guesses at search queries that will bring relevant search results, they’re often met with poor search results. Text search can often feel one-dimensional and frustrating, as shoppers are stuck typing in repeated search phrases as they attempt to get closer to the keyword combination that will get them results. 

The YesPlz Approach to Text Search

We created an enhanced text search that keeps shoppers engaged, by designing a system that constantly refreshes popular keywords, using AI. The YesPlz Enhanced Text Search System also includes AI autocomplete with product preview images, making discovery seamless.

YesPlz combines deep product tagging with text search to create more accurate search results that are tailored to retailer’s product catalogs.

The TL;DR answer: Why is YesPlz different from other discovery software solutions?

1. You don’t need to supplement our eCommerce product discovery with any other products.

2. Our core AI technology is trained by fashion experts, as well as shoppers, so we’re always up to date on the latest, most popular trends and keywords

3. YesPlz combined technologies create the most optimized discovery solution for your website, strengthening the overall quality of search, recommendations, and filter accuracy. We use valuable data from one area, such as recommendations, to create better quality search, for example.

Learn more about YesPlz complete product discovery 

Why YesPlz AI Is the Best eCommerce Product Discovery Solution

With the large amount of solutions available, how can you be sure that you’re choosing the best one? 

We believe that YesPlz AI is the best product discovery for eCommerce solution because:

1. Our integration is simple.

Single Data Ingestion: With single data ingestion, any Commerce can simply plug in through our API or dump their data feed to YesPlz, and get access to multiple solutions. Since YesPlz takes care of the rest, integration is as simple as possible. There’s no need for sharing data over and over.

No more vendor fatigue: Working with 3 different vendors means 3 different integrations. But, when working with YesPlz, eCommerce can find multiple solutions in only 1 vendor, making integration that much easier.

Auto Product Tagging: YesPlz fashion tagging runs on a fashion-trained algorithm that can accurately identify product attributes from any quality photo in milliseconds. By using YesPlz auto product tagging, personalization becomes easily scalable.

2. Easy customization for every client need.

YesPlz customizes product tagging, UI, and product discovery algorithms for their best performance. Every fashion retailer has different needs, that range from product tagging to widget integration–and YesPlz technology is set up to support customization.

Examples of YesPlz Customization

• UI Customization:  By collaborating with retail partners to customize design and features, we can enhance overall performance. For example, we know that men and women expect different results, even from the same search term. YesPlz uses machine learning to learn the most popular, trending keywords, identify similar search terms, and make relevant autocomplete suggestions. 

• Hyper Recommendations: YesPlz creates different weights for recommendations, calibrating the boost by with design, vibe, popularity, and more. 

• Occasion Filter: We create powerful occasion filters using product data from micro-tagging, allowing retailers to access modern thematic filters that shoppers love to use.

3. YesPlz creates the most innovative products:

From basic search and recommendations that are enhanced for nitty gritty details by our machine learning to a truly one-of-a-kind mannequin filter (patented) and fashion quiz, we’re dedicated to innovating.

We spend most of our time talking to end users and testing new solutions to make the search and discovery experience more delightful. Our clients love our fashion-focused technology and its many unique applications. 

The virtual mannequin filter is a revolutionary way for shoppers to search, and doesn’t require any knowledge of previous fashion jargon, allowing shoppers to customize their exact style and silhouette preferences. 

We also took an innovative approach to solve the inherent text search problems, like lack of contextual and missing product data in text. By adding product tags to a search library, we bring even richer search results.

And lastly, all of the YesPlz discovery solutions are built to mutually reinforce each other. With product tagging data, we can build unique, customized personalized solutions. When combining collaborative and similar search recommendations, we can overcome the cold-start problem, one of the most frequent problems in recommendations. 

How does YesPlz AI create such sophisticated solutions?

With fashion-trained technology and a laser focus on the search and discovery experience, YesPlz is capable of building discovery solutions for any size eCommerce. We specialize in machine learning, computer vision, and NLP, and have a deep understanding of both technology and fashion. 

Our process always involves collaboration with retail partners to build the highest quality discovery solutions.

And, most of all, we’re committed to making the shopping experience delightful and efficient, through innovative discovery solutions. That commitment drives us to create unique solutions to some of the oldest problems in eCommerce.