A recent survey commissioned by Nuance Communications and conducted by Vocal Laboratories concludes that, “consumers have an appetite for voice biometrics”. I agree there is an appetite, but I wonder if that appetite is great enough to trigger a switch from text-based passwords to voice-based passwords.
The downsides to text-based passwords are: (A) You have to remember them. (B) You have to type them (which means occasional errors and backspacing, especially on mobile devices).
The downsides to voice-based passwords are: (A) You have to remember them (but it might be easier). (B) You have to say them out loud (which means occasional errors and repeating yourself, especially in noisy environments).
On the surface, it doesn’t seem like a slam-dunk, but I believe voice-based passwords will eventually be a part of the authentication mix.
Regarding the Nuance/VocalLabs survey, it’s worth pointing out that Nuance has an obvious motive in pushing anything voice-based because that’s their core technology. Not saying it’s wrong; just remember to consider the source.
Here are some thoughts:
1. To conduct the survey, Nuance showed the participants “a video that demonstrated a voice biometrics mobile application … Callers were prompted to speak a pass phrase … heard someone press a button and say ‘my voice is my password.’ They were then brought into the home screen of the application.” There’s a big difference between watching the ideal case and actually using it. By conducting the survey in this manner, participants avoided the number one annoyance with voice recognition: having to repeat the phrase because the machine didn’t catch it. That drives people nuts.
2. The stats they listed make sense but don’t exactly build the case they’re trying to make. For example, they report that, “96% said they make mistakes typing their passwords, 10% make mistakes every time.” Sure, but the real question is: Will a voice-based password be faster or more convenient, even if you have to correct a mistake while typing?
3. Then there’s this stat: “89% have more than 10 passwords to manage…” which is totally believable. But the “I have too many passwords” problem is separate from the “I don’t like typing passwords” problem.
4. When they got down to a more direct question: “[Would] interacting with a … mobile app through a natural spoken conversation be easier than typing?” 59% said yes – that’s a majority but definitely not overwhelming. So, people were shown a video of the ideal case of voice authentication and still, 40% said they’d rather type their password.
5. I was really happy to see this one: “85% of consumers would rather use a mobile app instead of calling a company for routine inquiries such as checking their balance, checking flight status, and other tasks.” This certainly helps their case even if it doesn’t speak directly to voice-based authentication. It does however speak to the growing role of smartphones for customer service and the ongoing dislike of dealing with call centers.
Don’t mean to dump on Nuance, they’re clearly market leaders in speech tech and I do think voice-based authentication will be part of the mobile world but I don’t think this survey is the conclusive proof we’re looking for.
What are your thoughts?