With more people using smart phones than ever before, the ability for search engines to handle natural language and understand the intent of the user has never been greater. In 2015, voice search went from zero to 10% of all search volume.
Soleo announced the launch of its latest version of the Local Search API at Street Fight Summit 2016, opening it up to developers to start using in their own projects. This latest version together natural language processing and its database of over 20 million business listings.
Jason Nordhaus, Chief Scientist at Soleo, showed that ComScore estimates that 50 percent of all searches will be by voice by 2020. The reality is that, for nearly two-thirds of people, the ability to ask their question while their hands or eyes are occupied is an important feature.
One might imagine that the primary setting for using voice search is while on the go, but in reality, 43% of people use it in their home, which makes sense considering the rapid rise of products like the Amazon Echo.
“We’ll see more interaction with voice search in different settings that we don’t even know about yet,” Nordhaus said.
Nordhaus showed natural search is changing the traditional way that people are used to engaging with search engines. A traditional search query would be “dentist brooklyn ny.” Yet, with natural language, all someone might say is “my son chipped his tooth.” With the API, it would immediately understand that a dentist is needed and, with local search, it would find one in the Brooklyn area.
“If you’re having a conversation with someone, you have to be adaptive. You have to pick up on cues,” Nordhaus said, talking about natural local search. “We have to make it more natural as you might communicate with another person.”
In the demo, Nordhaus showed how the system can take the natural search and then deliver the right type of business.
But as Nordhaus explained, when dealing with local search mixed with natural language, the number of possible queries become quite large. The API required Soleo to process 10^22 possible search query combinations, which is 10 billion trillion queries.
Nordhaus compared it to the size of our home galaxy.
“There are 100 billion stars in the Milky Way galaxy. Our local search system can handles 100 billion Milky Way galaxies of possible search strings.”
In a world where technologies, such as smartphones and virtual assistant software are becoming more prevalent, the ability to process and understand so many different search queries is going to become tremendously more valuable. If ComScore is right, ultimately, businesses that don’t understand the mix of local and natural will be left behind.
Jacob Donnelly is a Street Fight contributor. Photo by Shana Wittenwyler.