Audio Comments

The fun part about blogging is the new people you get to meet and converse with. This happens in a variety of ways. Some people post comments, some write emails, some tweet, but recently I got my first voice message comment.

It was from Chris, a Telecom Specialist in North Carolina, whom I have not had the pleasure of meeting in person. This blog has a gadget that allows visitors to call. Chris called, and left me a voice mail. The voice mail was 2:22 long and was delivered as an audio file and a text file. Here is an excerpt:

i’ve been upon of I P telephony for a long time bud skype in this is there google voice that’s pretty cool tooth the only one comments yeah i love doing telephony from the P C and the desk and you know from P C the best time to the only thing is is the PCs gotta be on and logged on to do that it…

This was using voice recognition software provided with the Google Voice service (voice mail transcription). As you can see, it isn’t perfect. Here is the revised version I created after listening to the message:

I’ve been a proponent of IP telephony for a long time (Skype and Google Voice stuff is pretty cool too). Only one comment, I love doing telephony from the PC from the PC desktop, Only problem is the PC has to be on and logged on to do it…

It is actually a little more clear in the Google Voice inbox because the more ify words are in gray and when you press the play button, the cursor highlights the text as it reads it. I don’t expect the voice recognition to be able to get the unspoken parentheses, but why can’t it capitalize the “i’ve” like a word processor can?

Anyway, the point is this blog is hosted by Blogger (yet another Google tentacle in my life). Why not offer a solution that allows the voice mail (with the audio, text, and play button) to be within the comments section?

It really doesn’t even have to be Google Voice. Companies like Twilio or Voxeo should be able to offer a widget like this (though unlikely for free).

This is the great part of UC, being able to apply the pieces into custom(izable) solutions offering new capabilities. The solutions can be from anywhere and be implemented anyplace where it makes sense.

There is no ROI in audio comments on my blog site (though clearly it would be nice), but the ability to take things like a voice mail recording and update an application has lots of potential. For example, a field tech that calls in to close a ticket – complete with reason codes and free form text. Lots of websites rely heavily on comments – such as Amazon (they even rate the comments) – why must contributors be forced to type?

There are speech recognition systems that work better than shown above. Of course, when you know what to expect, the systems get very accurate – like a “yes” or “no” choice. Free=from text is quite a bit trickier. One of the problems with Google Voice transcription is people don’t talk slowly and clearly to voice mail as they do when they know they are talking to a speech recognition system.