One of the most striking things about CES – and what is happening in broader culture – isn’t necessarily what cool tech-forward devices people are using, but rather how people are able to (and are driven to) interact with them. Mulling over some of the high level trends that pervaded the show as I wandered through the Samsung area, I thought about some of the practical implications of these trends and how they impact actual human experience.
I joined an eager group of fellow event-goers crowded around a small stage upon which a young woman was beginning a demo – or sales pitch if you’re inclined to view things that way – of a number of Samsung’s connected devices. She stood in a well-crafted and strategically manicured model of a room in one’s house, one naturally tidier than the rooms any of us messy humans actually live in, which was of course stocked with what seemed to be smart everything: lights, appliances, thermostat, air purifier, window shades, speakers, among other things. She presented how she would go about her morning routine – and how much better smart devices made it. She spoke to Bixby, Samsung’s yet-to-catch-on digital assistant (think: Alexa), in some cute coded manner that I can’t exactly recall, like "Bixby, start my morning.” With that cue, she narrated that the thermostat would turn to the ideal temperature, the air purifier would go on, music would start playing, and some of the other smart things would come to life too in order to set the stage for her morning yoga routine. Whether or not her coffee automatically brewed and poured too, I can’t recall. But I can recall that seemingly wherever I went at CES, connected devices were touted as having the functionality to be controlled by one’s voice – either directly or via smart speakers.
An Easy Entry into Voice Tech
A connected daily routine sounds cool, to some, but what does it actually mean for research? I’m going to table a discussion about the data created by these devices and how it could potentially be used and instead focus on what she did. She spoke. She didn’t get out her phone. She didn’t even touch anything. And it was natural. Voice tech isn’t exactly new, of course. Most of you probably own or know someone who owns an Amazon or Google smart speaker. If you dictate to your phone, chances are pretty good it will come up with “Will you please pick up some bananas at the store” rather than the “Willow peas pick and sum aunt Santas and destroy” that it may have come up with a few years back.
The tech isn’t perfect, but it’s getting pretty darn close pretty darn quickly. So, let’s start with a really simple application of voice tech for standard market research. Do you have a survey that includes open ends? Of course you do. Do you know if people are using voice-to-text to complete these open ends? I can guarantee you that some are, but also that many people aren’t thinking about this. But they should. And you should remind them it’s an option. Don’t force them, but make sure it’s top of mind for those who take your surveys on a mobile device and are comfortable with the tech. I’ve conducted some research – along with my partners at MaritzCX and FocusVision – that has shown that voice-to-text responses are longer and of higher quality than traditional typed responses. To be frank, I see the upside far outweighing a very limited downside in prompting people to answer open ends using voice to text.
Can Voice Make “Surveys” Better?
Those of us thinking about the future of surveys and the future of research more broadly need to think beyond the immediate, practical applications and to how we connect with people to participate in research in an increasingly voice-driven world. Market research has fallen too short too often in meeting people on their terms and has expected, arrogantly, that people should take their time to take our surveys no matter how crappy and/or inconvenient they are.
As an industry, we’ve been through the wringer in trying to just make surveys palatable for those who take them on a phone (and still struggling significantly, IMHO). We need to be thinking about how we design research in a conversational manner – and actually testing this by scrapping much of the baggage that comes along with traditional quantitative online surveys. With requisite apologies to those of you designing engaging surveys, most research engagements are formal, boring, clunky, and everything but conversational.
We should want people to engage with us in natural ways, by speaking, texting, writing as if it were a conversation with a friend (chatbots, anyone?) and approach research with this in mind. If we want more honest and accurate data from more representative audiences, we should follow this path. I have seen/heard a few companies testing quant surveys through smart speakers – and kudos to them. And applications in the qual space may be even more immediately fruitful. I’m hopeful that voice tech spurs the development of more authentic and interesting approaches to research. With persistent talk about the “voice of the customer,” let’s think more about how we actually hear their voices.
For those of you still listening, how do you think we can effectively utilize voice tech to create better research?