Constructor, the leading AI-powered product discovery and search platform for enterprise ecommerce companies, today announced Attribute Enrichment, a solution that improves online product catalog quality using deep learning and machine vision, both forms of AI. Attribute Enrichment combines raw product catalog data with buyers’ behavioral clickstream data across touchpoints — correcting and auto-enriching product attributes to help buyers discover the items they’ll find most attractive, while improving personalization across their customer journey.
Attribute Enrichment addresses one of the most common pain points today among both ecommerce companies and their consumers. For ecommerce companies, needing to work around bad or missing product data is a problem as old as ecommerce itself, with site merchants either forgoing that data altogether or needing to rely on poorly tagged data from third parties. For B2C shoppers and B2B buyers, this leads to time-consuming and frustrating product discovery and search experiences. For instance, less than 3 in 10 buyers say their product-finding experiences are “quick,” and 62% will leave an ecommerce site if they can’t find what they’re looking for.
Ecommerce companies seek to remove this friction and provide visitors with highly contextual, hyper-relevant experiences that increase engagement and conversions. Consistent, comprehensive, well-categorized product data helps power these experiences… but getting it is often easier said than done. Relying on vendor attributes — the product information and categorization from third-party suppliers — is a gamble that often leaves “holes” hindering product discovery and personalization. And do-it-yourself efforts among ecommerce teams — including manual enrichment, categorization, retagging and harmonizing third-party data into a consistent catalog — is time-intensive and can be error-prone.
How it works
Constructor’s Attribute Enrichment uses deep learning techniques: that is, training AI models to identify products, categorize them and generate new product attributes to enrich product data and increase discoverability. The newly enriched data is delivered straight to ecommerce companies’ websites, without any effort on their behalf — giving visitors immediate access to better and more complete product information that drives purchases.
Attribute Enrichment provides ecommerce companies — including B2C retailers, D2C brands, distributors, manufacturers, wholesalers and other B2B organizations — with:
- Out-of-the-box enrichment capabilities to address common retail use cases, including category enrichment, color enrichment, style enrichment, flavor enrichment and many more. For example, if a shopper is searching for “Spanish red wine” on a grocery or beverage site, Attribute Enrichment can augment existing product catalog data to populate flavor profiles and regions to help the shopper better find what they’re looking for. And as an example of category enrichment, an impact driver on a construction supply site might be classified in the “tools” category. Using everything it knows about the item and others like it, Attribute Enrichment can automatically associate additional categories (e.g., “power tools,” “drills and drivers,” “cordless drills and drivers”) with the product, making it easier for buyers to find it as they search, browse and filter.
- Custom attribute enrichment capabilities tailored to ecommerce companies’ needs and catalogs.
- Personalized enrichment experiences, predicting attributes that are best aligned to buyers’ affinities, such as the style of dress they like, whether they prefer organic products, whether they prefer scented or unscented cleaning products, whether they typically buy cat food or dog food, and many more.
- The ability to automatically generate and associate new attributes with products, based on trends in customer behavior (e.g., if searches for “keto” or “vegan” become prevalent, tying those attributes to various items).
- Reporting capabilities, including reports that display the most important attributes for buyers and calculate Attribute Enrichment ROI.
Benefits of Attribute Enrichment
By incorporating Attribute Enrichment into their product discovery strategies, ecommerce companies can:
- Increase revenue per visitor and average order value by displaying more attractive and strategic products to buyers.
- Improve the customer experience and increase customer satisfaction, delivering experiences that enable buyers to find the right products faster and easier.
- Increase consistency in item classification and categorization across complex and extensive product catalogs.
- Optimize product discoverability and relevance.
- Increase operational efficiency by automating previously manual tasks and harmonizing supplier product data to enable field sales teams, customer support teams and distributors.
- Improve search engine optimization (SEO) through accurate and more complete product descriptions, contributing to higher search engine rankings and lower bounce rates.
- Extend the value of their product data by sharing the newly enriched data with their upstream systems and their other partners for use in improving discoverability of products in third-party marketplaces, improving ad placement and more.
“Constructor helps ecommerce companies turn their websites and apps into personal shopping assistants, and our new Attribute Enrichment solution is an important part of this,” said Eli Finkelshteyn, CEO, Constructor. “For too long, an excuse among search and discovery vendors for bad results was that the data they were given was bad. Attribute Enrichment aims to take away that excuse. We want to guarantee we’ll give the most compelling results to our customers and their shoppers, even in cases where their product catalogs aren’t perfect. Our goal is to not give excuses for poor results, but to give solutions. And Attribute Enrichment is just that.
“Using modern data science methods like machine vision, mixed with customer feedback via clickstream and behavioral data, we’re able to generate and auto-enrich product attributes, categories and metadata, all at the speed of AI. It’s a win-win for both buyers and ecommerce companies alike: fueling the best product discovery experiences for buyers, while driving sales and giving ecommerce teams time back in their days to focus on strategic initiatives.”
SOURCE: PRNewswire