Talking To The Future: Voice Powered Search & Ecommerce
Most everyone who has a smartphone has pressed their device’s main button and verbally asked for information to be retrieved. We are so accustomed to this technology today that it only seems logical that we should also be able to use it to shop for things on the internet. Voice powered ecommerce is extremely appealing and does exist today, but it currently has rather severe limitations that will have to be overcome in order for the technology to be put into wide spread use and also meet consumers’ demanding expectations.
Let’s take a look at some of the challenges surrounding voice powered ecommerce that the industry is working to resolve.
Natural Language Processing
Natural language processing (NLP) is the foundation of voice powered ecommerce. This topic has actually been an area of research and development for over 60 years. In essence, NLP involves the vocal interactions people have with computers and a machine’s ability to understand and interpret information based on the natural way that humans talk.
The ultimate goal of NLP is to develop a way for computers to meaningfully interact with people based on the warm, flowing conversational style that humans use to communicate. A fundamental component of NLP is machine learning, a computer’s ability to “learn” without being additionally programmed. Machine learning is based on artificial intelligence, that analyzes large data sets and utilizes patterns to continually improve the ability to understand speech.
Even though there have been incredible advances in NLP, the technology where it stands today has not been able to deliver the same level of satisfactory and effective ecommerce experiences that typing into a search box currently affords us.
The Stats Don’t Justify The Fanfare
There has been so much talk in recent years about voice powered ecommerce that it would easy to conclude that there’s a huge demand for it in the marketplace. However, the actual statistics on the matter tell a different story. According to The Information, a digital media company, only 2% of people who own a smart speaker device have ever made a purchase with their voice.
Data also shows that 90% of people who did make a voice command purchase did not do it again.
It is further interesting to note that most items today purchased via voice are very impersonal, such as paper towels, laundry detergent and other household goods. Products that require more research or involve somebody’s personal preferences are hardly ever bought today via a voice powered, smart speaker.
“Buy Me A Cheap, Red Shirt"
With current onsite search technology, shoppers are able to describe or accurately pinpoint what it is they are looking to purchase from typing into the search box. Once search results are returned, facets and categories can be used to refine the relevancy and personalization of the website visitor’s experience.
At this point in the voice search world, the technology is not able to accomplish this same level of search satisfaction. Imagine telling your smart speaker today to buy a cheap, red shirt. What does cheap mean? What shade of red? Is this for a male or female? What kind of shirt? Spoken words are very often open to interpretation and even humans sometimes need clarification about what others mean when they use potentially ambiguous words.
Voice powered ecommerce is an exciting technology that will be sure to advance in years to come and breakthroughs in Natural Language Processing (NLP) will continue to be the wind beneath the wings of voice based, online purchasing. For the time being, smart speaker devices will help shoppers buy more standardized products like tissue and cleaning supplies, but not other items that are highly personalized. For now, stay close to the search box and your trusty keyboard.
Want to learn more? The HawkSearch platform offers a powerful set of tools and functionlaity that enables search across your ecosystem and provide your visitors with the best search experience. Schedule a demo today and see HawkSearch in action.