I was meeting someone at Grand Central Terminal last week (and for you non-NY‘ers, Grand Central Station is a subway stop; GCT is the commuter train station). We agreed to meet “at the clock” which everyone knows is the clock atop the information booth in the middle of the building.
I was a few minutes early and as I was waiting, I overheard the person in the information booth as she dealt with the line of folks needing help. What struck me wasn’t the speed with which she answered the questions (how the heck do you keep track of all those bus routes in your head?). Instead, I was impressed with something else, something that is the yet as unfound grail in digital, but one which Siri and other voice interfaces may be on the way to finding.
That’s the grail of “intent.” As I was listening to her, I was struck by how often the woman answered a question with another question that would clarify intent. For example, “where is the subway?” was met with “where are you going?” since the two main subway areas go to very different places. She wasn’t just answering the question she was asked; she was making sure she had been asked the right question.
I realize that all the major search engines try to guess at your intent as they build Search Engine Result Pages. Generally, they fail, in my opinion, because we don’t ask questions well (Where is the subway? vs. How do I get to West 86th Street by subway?). The search engines answer each question they’re asked without clarifying. That’s changing, however.
I don’t use Siri but I do have another voice-assistant (Speaktoit Assistant) on my Android phone. I’m interested in how often it asks a clarifying question before answering. I suspect it won’t be long before things such as tone of voice (“where is the hospital?” asked in a rushed, stressful way might prompt a call to 911) will be factored in.
Part of what I do with clients when working on SEO or SEM is to predict consumer intent. It’s hard, and I imagine when you’re trying to predict intent for something north of 34,000 searches a second (that’s just Google!), it’s pretty daunting. It cheers me a bit that we humans can still do that better than the machines, but as the machines become more human, that gap is closing.