Up until now, one of the biggest roadblocks to the growth of voice search has been inaccuracy—both real and perceived. In its infancy, errors were rampant and query attempts were quickly abandoned for the reliability of a keyboard. Technology, however, has progressed leaps and bounds from those early stages of voice recognition with still more ground to cover. In 2017, Google’s word error rate dropped to 4.9 percent from 23 percent in 2013. That’s like going from missing one in every four words, to one in every twenty.