The business of moving bits has clearly become a commodity, much to the chagrin of Cisco. I can't help but imagine the sales messages they must be delivering as they fight what is obviously a losing battle.

We're in the midst of launching a small ASP business and we needed a handful of Layer 3 devices.  I sat down with my engineer to try and map out the budget for these devices. The conversation went something like this. We could buy brand-new Cisco Catalyst boxes for about $6,000 each. OK - I wrote that down. OR he quickly said, we can get them refurbished for about $3,000. I quickly scratched the $6k for a $3k unit.

But that wasn't enough, he said that Dell had some L3 switches with similar specs that we could get new for $3k. I quickly changed the DESCRIPTION from Cisco to Dell.

Obviously, my engineer was on a roll, he then said that Netgear had pretty much the same stuff for about $1,900. I realized I probably was going to need a pencil to keep up with this as I marked both the price and description again.

So in the course of a 10 min conversation, we went from $6,000 per unit to $1,900 saving effectively 66% of where we had started for what my engineers says will get the job we need done. Cisco no doubt is advocating superior support, better quality and and and, all of which reminds me of the same message I got from General Motors when Japanese cars first arrived. Do you see how this story might go?

Well the L3 switches we need are scheduled to arrive next week, my engineer let slip that he managed to find the Netgear boxes we wanted on the used market and the price, well, $400 each.

Whether you like it or not, most long distance calls today are carried on a VoIP network and practically all international calls are. Your international call may ride thru various wholesalers on it's way to your end destination. Your telephone company may send it to 'X' who decides to send it to 'Y' and may send it to 'Z' before it hits the telephone you're calling.

Along this path, XY and Z may often decide to compress your phone call using any number of known CODECS, most notable G.729a to save on bandwidth costs (cheap being a winning factor). Your phone call may actually be compressed and decompressed before it actually gets anywhere.

The result is terrible voice quality. It seems sort of logical, I'm calling the other side of the earth, it's a long way, anything can happen, it's not going to sound as good as calling next door. If we get a really terrible call, we hang up and call again. It's not until you use Skype to call half way around the world and get crystal clear "in the room with me" quality do you start to realize how awful the telephone network is.

God help you if you make an international call from your mobile, which starts it's journey from your handset being compressed and less than toll quality.

But who do you complain to? Most folks simply shrug, what can you do? Many have never experienced the pristine quality of a Skype call so they simply don't know that a better solution exists. The telephone companies tell me it's too hard to change and no one really wants 'better' voice quality - it's not a marketable 'feature'.

Or is it? AT&T is disconnecting about 33,000 residential telephone lines each day, people are downloading Skype about 95,000 times a day. Do you see a trend here?

I've long advocated that VoIP needs to be both cheaper and better, something Skype (which is non-standard) has clearly done and is winning doing it.

Update: The Wall Street Journal, quoting sources at British Telecom, reported that SpinVox is basically using humans to listen to your voicemail and transcribe this. If

Back in April 2007 (original posting), I was down at CTIA and met this arrogant prick executive from SpinVox. They were going to conquer the world transcribing voicemail messages to text. Spinvox had given me a free account and sure enough, it worked quite well.

But I dug a bit deeper and found out they were transcribing my voicemail by having a human listen to it and type it in. They sent my voicemail messages to whatever on the earth they could find cheap human beings. Hey - I was cool with that, didn't matter to me so long as it worked, a low cost receptionist I thought, though tougher to wring costs out of it.

The prick executive was explaining how they hoped this transcription service could also be transformed into a "record of communications" for large corporations. When I asked how an CEO from a public FORTUNE 500 company might feel about some Jamaican being the first person to hear a voicemail to his CFO about their quarterly results, well let's just say the Spinvox executive got a bit muffed up. When I confronted them with how the transcription actually worked and their bogus patent, well, let's just say it was much shorter meeting than I'd planned.

Gigaom is reporting effectively that today as Spinvox is in a spin after all of the things I knew over 2 years ago are now becoming more public. If things aren't what they seem they should be, investigate.

One financial fund manager gave me the best advice about CEOs. Three basic rules. Remember, they lie, they lie and the last and most important one, they lie.

The PBX and IP Centrex guys lost in the woods with nothing new to talk about  have grabbed hold of 'cloud computing'. Yes, cloud computing is going to save them. Why, they haven't fully explained but clearly they are fearful the herd will move there and leave them in the dust.

I'm an alligator, preferring to do nothing, looking at everything, moving only when I'm sure it's compelling and within reach. Call me lazy.

The axiom poles aren't likely to change to get something to take off, fundamentally you have to solve a problem I'm having (real or imagined) and do it at a compelling price. That simple. Ask yourself that daily when you imagine some new product idea. Hitting one BTW probably isn't enough to get you in business.

Cloud computing is the modern day version of 'time share'. I can get massive computing power and rent when I need it for however long I need it and turn it off quickly. So using my axiom, cloud computing is a great idea. I need massive computing power for some application but don't need it all the time and the Amazon EC2 pricing is fantastic, just pennies per hour. Great idea, let me acknowledge this.

But running a telephone system? As Gigaom pointed out in a recent article, bandwidth costs in the cloud are prohibitively expensive and sadly VoIP doesn't allow an easy way for 2 devices to talk to each other directly (blame the VoIP guys who have successfully took 1970 Class 5 switching technology and upgraded it).

Many geeks have successfully figured out how to run the open source Asterisk in the cloud, I played with it. Neat Saturday afternoon experiment, but the 'why' question is never answered. J Dodd @ Digium, who admittedly is about every customer having their own Asterisk server did the attached spreadsheet. Spreadsheet http://www.loligo.com/asterisk/misc/amazon-ec2.xls

His basic calculations figure it cost a small business ~ $36 a month to run an Asterisk server for typical 20 seat business, so about $1.80 a user. He figures it you use Amazon EC2 it's gonna cost you $72.00 a month to run this same server. I'm waiting for nut jobs to pop up and start talking about "total cost of ownership" as the cloud advantage. I used to throw vendors out of my office when they used those words which effectively mean, we don't have a business model.

Now J Dodd's calculations were based upon running the service 24 hours a day, now clearly I'm not in my office 24 hours a day, but easy math says that if I need the telephone system only 12 hours a day, my EC2 cost would be ~ $39 a month, better, but not compelling.

If the day comes that Amazon EC2 can run my telephony system for 1/2 of what I'm currently spending, call me, I'll be interested, until then ... well ... don't call me.