The telcos have told the FCC that they think the cost of providing free voicemail to subscribers displaced by a Katrina-like emergency would be a substantial obstacle. I calculated a capital cost of less than a penny per subscriber to provide every subscriber with ten megs of storage for voicemail. Reader DG Lewis points out, correctly, that there may be licensing costs involved. Here’s his comment in full:
“Another suggestion would be to reach out to one or more voicemail system providers and engage their support - and provide some more realistic cost figures than ‘a penny a mailbox’.
“Sure, disk storage is cheap - but the cost of voicemail systems isn't in the hardware, it's amortizing the software development and support. You will not be taken seriously telling the FCC that voicemail costs a penny a mailbox when carriers are able to take them purchase orders written to Comverse and Avaya with prices of a couple of bucks per mailbox. Unless you plan to go into the voicemail development business and sell the carriers two million lines of voicemail capacity for $20k.”
I am NOT planning to go into this business (important disclaimer since I am pushing for the VM requirement).
I liked Lewis’ suggestion that we get the opinion of a VM expert so I urged Craig Walker, founder of voicemail provider GrandCentral, to comment. Here’s what he wrote:
“Great entry. In response to some of the prior comments here, it is true that voicemail costs can be more than the raw storage costs. However, most of these costs are licensing costs by the voicemail vendor. If there is a decision to require the telcos to provide voicemail temporarily in an emergency, that order might also be able to provide relief for these licensing costs as well, particularly if the service is only being provided temporarily. As such the true costs would only be for incremental storage costs.
“Regardless, the costs to the telcos all depends on their specific voicemail systems and agreements with their vendors. Its also likely they might have a license for nearly unlimited accounts, in which case the incremental license costs would be zero, or they might have (God forbid) built their own systems...which again would have the incremental costs down to the storage costs.
“We (GrandCentral) built our own voicemail as we weren't interested in paying the high licensing costs described above and are now able to offer free local phone numbers and voicemail to the homeless and temporarily displaced in the San Francisco Bay Area through our Project CARE initiative (kudos to PacWest Telecomm for providing the DIDs for free). If a 12 person start-up that has been in business for 6 months can do it, I'm certain the RBOCs should be able to figure it out.”
Now I know from my time at AT&T that it can be very expensive to do very simple things in a huge bureaucracy and Craig may be too generous in assuming that an RBOC can do what his twelve person startup can. So let’s do a reductio ad aburdsum on the cost argument and put it behind us forever.
Telcos sell voice mail to their customers for between $4 and $7/month. We’ll assume that there is NO profit at $4, that this is the true cost to them. I know this assumption is absurd but I want to prove a point. Now let’s assume that, in any given year, one percent of a telco’s customers need to be given free voice mail for an average time of three months each. Again a very exaggerated assumption. If the telco is going to recover this cost by increasing prices to the whole customer base (because this is insurance for the whole customer base), the monthly price increase would be one penny. Here’s an example:
Megatel has 10,000,000 subscribers. By our assumptions, 100,000 (one percent) will need disaster voicemail each year. That’s 300,000 months of voicemail which we are assuming cost Megatel an astronomical $4/month each or $1,200,000 total. With 10,000,000 subscribers there are 120,000,000 monthly bills sent out each year. If each of them is one penny higher, Megatel recovers its $1,200,000. If there aren’t many disasters, they get to keep it.
Of course, Megatel could also hire Craig Walker or someone else competent and reduce this cost by 99%.