We’ve seen the fancy data visualizations on Facebook and the legendary algorithms of Netflix. But what can datascience do for a nuts-and-bolts e-commerce website?
The thrill and buzz around datascience has been electrifying. Bushytailed academics transitioning from consulting to tech have been drawn in by tales of Netflix’s mythical datascience algorithms that predict which movie you’re gonna watch next.
But ask about e-commerce, and the eyes grow a bit puzzled. How exactly would datascience help?
You could broadly categorize that into 2, maybe 3 categories; namely competition, affiliate sales, and analytics
If you’re selling something online, it makes perfect sense to figure out what your competition is selling the same product for. That’s pretty easy to do if you have a couple of dozen products to sell. Not quite, if you have a million products. What if you wanted to match your top selling products to their website and see which ones sell better? What if you wanted to figure out if your products are not priced optimally?
Not that easy to do for a million products.
Datascience helps here by pulling together product and pricing information from thousands of online retailers and putting it into a database, like this one. Using an API, you can pull price and product datasets from your competition, find which ones that match your products and price accordingly. It also helps to find out which of your competition’s products are trending.
Affiliate marketing is a great way for online retailers to generate sales without spending too much on marketing efforts. However, in many cases, affiliate marketers often struggle with poorly updated product feeds, outdated prices, or poorly formatted feeds. Having access to a frequently updated database with a normalized product feed is very useful, especially when you’re developing a front-facing retail app that uses affiliate links to sell
E-commerce analytics companies find product datafeeds very useful especially when they’re coming up with great data visualizations, and generating key metrics like inflation, etc. It also helps when you’re trying to dynamically optimize your pricing
I run business growth at Semantics3, where we help many of our customers with similar issues. We primarily focus on growing our ecommerce product and price database though information extraction from over 25,000 online retailers. We now have over 46 million unique products in our database.