Back to Blog List

Topics/Previous Posts

How Voice Assistants Change the Way We Shop

Even five years ago, the idea of a personal assistant was something reserved for the rich and famous, the wealthy or top-end business professionals. Today, however, every smartphone owner gets the experience of a personal assistant in their pocket. Thanks to Apple’s Siri, Microsoft’s Cortana and Amazon’s Alexa (and, to some extent, Google Assistant), being able to speak a command and get results instantly now possible.

More than giving you the weather report, scheduling events on your calendar or calling mom, these digital personal assistants are also changing the way people shop. No longer are people spending time sifting through pages of products or clicking through refinement options to find what they’re looking for—instead, they’re asking the disembodied assistants in their pockets to find the items they’re looking for.

While it might seem like magic to ask Alexa to buy you a product without ever having to navigate to a website, the process behind this ask-to-order scenario is powered by Natural Language Processing (NLP).

Tracing commands to results

Without going into the nuances of exactly how digital personal assistants function, it’s important to understand how they take spoken words and translate them into results. Many assistants use pre-programmed preferences and specifications to process customer requests. This makes it easy to give direct commands, however, when commands deviate from the normal, NLP takes over to provide logical function.

As an example: if a customer says, “hey, Siri, re-order my mascara from Makeup Depot,” there’s a good chance this isn’t a stock command that either Siri nor Makeup Depot are equipped to process. However, thanks to NLP, it could be translated into one. Here’s how:

The term “re-order” could be interpreted as a “repeat order” command by Siri before being given to Makeup Depot. This will trigger a re-order executable. If a customer profile is linked to their smartphone, “my” will instantly signal the owner of the smartphone, allowing Siri to provide customer information to Makeup Depot, which will trigger an automatic customer lookup. Once the customer profile is pinpointed, the digital assistant can look for past instances of “mascara” products being ordered. Even if the product isn’t titled with “mascara,” NLP allows for comprehensive searching that can distinguish this product. With all of these naturally processed words and commands, the interpretation becomes something  on which both Siri (or another digital assistant) and an eCommerce platform can collaborate.

Recognizing actionable commands

Perhaps the most essential characteristic of NLP as it applies to digital personal assistants and modern eCommerce is the ability to recognize what people want, even if they’re not saying it. Instead of looking exclusively for static information, NLP enables a true understanding of what’s being asked, in an effort to provide true results. Some examples include:

“I need a cheap pair of pants” can be understood as “return results on men’s pants that are on sale or discounted below their normal price point.”
“Find a high quality winter coat” might be recognized as “show results for women’s winter coats that offer many features and characteristics, which also contain mention of the term ‘high-quality’ in the description or in customer reviews.”
“Let me see a good laptop” can spur results of “laptops that have comparatively high ratings versus others of a similar make.”

In each of these cases, generalizations and colloquialisms are turned into actionable commands, which is essential in utilizing natural language search.

Digital personal assistants like Siri, Cortana and Alexa are all leveraging NLP to better serve smartphone users and streamline the online shopping experience. Is your eCommerce store helping or hindering the process? Consider updating your site search to Celebros to help welcome the new age of digital personal shopping!


Leave a Comment

Only comments approved by post author will be displayed

Back to Blog List

Close