The modern consumer seeks a personalized online shopping experience. Within e-commerce, AI and Big Data have pampered customers by improving product recommendations, adding discounts, providing swift customer service, and forecasting demand in retail based on factors such as upcoming festivities, even the weather.
E-commerce businesses have understood the importance of incorporating such emerging tech and Swoop is adopting the same to provide e-commerce solutions to retailers on a global scale.
What is Swoop?
Swoop is a SaaS company that provides an e-commerce platform to retailers who are looking to sell products and merchandise internationally. Their expertise lies in creating localized versions of e-commerce sites that calculate duty, tax, and shipping for customers in those specific countries. Swoop also helps retailers by updating product prices based on the exchange rate at the time and offers the payment methods preferred locally.
The complexity of raw data
With a wide range of clients, Swoop’s offering starts with understanding each retailer’s product catalog and categorizing them appropriately. These products come with information such as product type, product descriptions, gender, apparel, type of weave, etc. This was across six retailers and more than 6000 product types.
Swoop holds this information in an uncategorized format, and they need to label all products from multiple retailers with a consistent labeling mechanism.
The issue is that converting unstructured data into a readable format can be a mammoth task for a team, considering the sheer volume of product data and the lack of the right annotation resources.
Swoop turned to Bridged to annotate their data.
Converting raw data into readable information
Swoop needed customs descriptions for their clients’ products to help their unique system of labeling, which would enable their business solutions. Swoop provided Bridged with complex unstructured data rows that contained the following product information:
|Information provided by Swoop||Details|
|Product ID||ID for identifying each product|
|Retailer Name||Gucci, Adidas|
|Category||Apparel, bags, jewelry|
|Item||Belt, backpack, necklace|
|Materials||Leather, plastic, gold|
|Gender||Women, men, unisex, infants|
For example, here’s a portion of the unstructured data rows Bridged was assigned, and their structured counterparts:
Swoop made use of Bridged’s resources to convert this raw data into information that could be read, deciphered, understood and made sense of by NLP. Using a combination of trained annotators and in-house technology, Bridged was able to deliver high-quality, NLP compatible data rows.
With the help of these data rows, Swoop has an efficient way to manage and locate the data they possess. This helps in various cases such as filtering items based on brand, gender, apparel type, etc. The data rows can also be used to train AI models to create insights from e-commerce numbers such as product clicks, purchases, etc.
The future of data in e-commerce
The e-commerce space is a strong market for data aggregators to tap into. Big shopping and product data are extremely useful for businesses to maintain a healthy relationship with their customers and also optimize their product line. But, the gateway to achieving this is data sets that can provide strong insights for effective business strategy and consumer analysis.