Voice Commands and Mobile

I’ve been lurking on a discussion about the use of voice commands in the mobile channel.  The proponents suggest that this is the next evolution for mobile computing, as one of the greatest limitations of the channel is in the difficulty of using the keyboard to input text.   It seems like a rational premise, but I rather side with the opponents of the proposition.

The problem is that the mobile device is not a verbal channel – which is self-evident in that it is called a “mobile device” rather than a phone.  The telephone is one of the least-used applications, and users prefer text messages and email.   In general, users seem to prefer the privacy of their mobile device, particularly in public spaces – and their interaction ceases to be private when they speak aloud. 

Further evidence is the lack of popularity of agent programs such as Apple’s Siri, which at first was not used because it was painfully unintelligent and tedious, but which has remained unpopular even now that the voice recognition and logic have been significantly improved.   People simply do not care to use voice commands with their mobile devices.

There are a few edge cases in which voice command makes sense: use of a mobile keyboard while driving is particularly dangerous and unwise, and using a voice interface seems a better choice.   I don’t have the sense there is much benefit it doing so – given the studies that show that drivers engaged in phone conversations, even hands-free, are just as impaired as drunk drivers because it is the distraction of their cognitive senses, not just the loss of a hand, that causes the greatest danger.

And of course, there is the indication that voice command might make the mobile channel more accessible for the vision-impaired user seems entirely plausible, though it seems this is a niche market and a visually impaired person might find it more convenient to simply place a call rather than attempting to use the voice command system of a mobile site or application.

The suggestion that the voice interface would be popular when using mobile device in a private environment seems a bit contrived.   It can neither be proven nor disproven that users do (or would) find voice commands to be preferable, as their behavior in private cannot be witnessed and self-reporting is likely to be inaccurate, either way.

At this point, the discussion has died down to the point where I felt it useful to take a few notes.  If it is resurrected and the proponents of voice commands can find a valid argument in its favor – but for the present I’m satisfied to rest on the conclusion that it is an interesting concept that is likely not going to be practical or commercially viable.