The retail world is in turmoil recently after a batch of brick-and-mortar earnings reports were met with sudden selloffs. As we’ve discussed before, contemporary B&M companies are obligated to devote assets towards modern tech. Having an online presence isn’t enough–Amazon, Walmart, and other retail giants are aggressively pursuing machine learning strategies to target — consumers more intelligently and manage supply chains with greater efficiency.
Even when it comes to something as straightforward as a product search experience, there’s plenty of room for innovation. Check out a few of the ways that intelligent data gathering and semantic classification can lower the hurdles between a customer and an online sale.
Descriptive Search
Sure. An iPhone is an iPhone. If you want one, you can go into a store (real or virtual) and say, “I’d like an iPhone.” But what about that cool, futuristic-looking washer and dryer, with the brushed steel and the touchscreen on the front? Things like household appliances tend to have less memorable product names, and though it may seem trivial, any ambiguity equals an opportunity for competitors.
That’s why it’s becoming more and more crucial to incorporate product features and specs into a natural-language search experience. Retailers that allow users to search for “washer with touchpad”, for example, put less of a burden on consumers. Bonus points if you have a semantic thesaurus that can read “touchscreen” and automatically search for “touch-screen”, “touchpad”, and “tablet”.
Prescriptive Search
On the other hand, some users may have all their vocab down, but aren’t as sure about what they’re looking for. Have you ever been to a retail website and typed in something like “gifts for mothers day”? Try that on a few different sites and you’ll quickly learn about the limitations of text-based search. Some sellers even include phrases like “mother’s day” in the product title itself, hoping to capitalize on literal-language search technology in a natural-language world.
But there’s a better way to cater to indecisive shoppers. Aided by machine learning, retailers can flag nebulous searches like “gifts for grads” or “cheap productivity tools” and consider the products those shoppers actually end up buying. The more data ingested, the more intelligent recommendations can become, until you’re giving your customers the products they literally didn’t know they wanted.
Of course, at the end of the day, many of these suffering brick-and-mortar companies simply don’t have the capital to launch some shiny new technology initiative. That’s why companies like eContext, who have already don’t the hard work of gathering and organizing large-scale data, are a life saver for any business going through innovation growing pains.