Telcos to deliver software as a service?
Usually Om Malik is right on the money with analysis, but his recent feature on telcos exploring software as a service missed the obvious problem with the concept: the telcos don’t do software.
Well, that’s not true. It is better to say that, by and large, the telcos are atrocious about using useful software to solve their own problems. They rely heavily, heavily, heavily on hardware (why use a few lines of code when a bigger switch will do the job), and they don’t do much that is considered smart and innovative with their software.
Case in point: I’d surmise that 90% of the churn to voice over IP is not about costs. Sure, the blog-o-sphere is full of guys who tout the latest VoIP over mobile data networks to shave a couple of dollars off your phone bill. That’s nice and all, but very few people in the SMB market–the core market for software as a service–actually care. No, VoIP wins when it wins because of features and the overall value that they can create. Many–if not most–of the features can be implemented on POTS/TDM infrastructure. The telco’s don’t do so, however, because as a general rule, they aren’t interested in solving problems to create new value. Maintenance and marketing are the mainstays of large telecom, not innovation. (Not that the so-called innovative start ups rehashing the same VoIP tricks twenty different days like five day old leftovers are any better–scale itself requires more innovation than is often assumed by the smug little “single rack” guys).
Still, a couple of examples can shed some light on the matter. A few years ago, when I was fronting an in-house developed Asterisk-based PBX that was a little ahead of its time and way underfunded to succeed, I spent some time talking to SMB owners to find out what they actually needed. I very quickly found out that they had no idea. Basic questions as to line utilization at peak usage were returned with estimates that smacked more of horse-betting than information-driven business decisions. Somewhat more recently, I did some Metro-Ethernet infrastructure troubleshooting/diagnosis on behalf of a cable company. One of their clients was school system that, according to their own IT staff, needed (and supposedly was using) gigabit speeds connecting all of their locations–justifying annual expense of hundreds of thousands of dollars on Metro-Ethernet infrastructure. In the course of diagnostics, statistics showed they should have been funding a dozen more teachers instead: peak usage over the course of a month was only about 1-2 Mb/s a second, nowhere near their supposed gigabit demand.
Here is a universal business problem: the average SMB customer has very little insight into their own telecommunication needs and usage. This presents a huge problem for the telcos (I’d argue that, believe it or not, many of the SMB customers going to VoIP platforms are doing so because of transparency of their usage rather than actual features per-se. The number one feature of VoIP for business such as law firms that bill by the minute will be the powerful usage reporting tools and the ability to integrate the usage statistics directly into billing systems). It also presents a huge opportunity–small little features like a portal showing usage, stats, and so forth in an easy to use manner accompanied by all sorts of pretty graphs would do wonders for retention. Even better, hand out all sorts of APIs like candy that eventually get integrated into various billing, CRM, and communication platforms–you want to maintain a customer, get your services integrated as tightly as possible into their business systems.
Cut the marketing budget in half and actually deliver something useful. Your world, delivered? International communication was a nice feature of 1960–let’s see some new features out here worthy of the 21st century.
A couple of friends of mine are working heavily on projects that are brilliant at making sense of large amounts of traffic associated with communications: Kelly Storm with Ajaxalytics and Alex Balashov with Evariste Systems. The former, Ajaxalytics, analyzes web server logs and uses a AJAX-ified interface to seamlessly pull the needed information out of them. Evariste is in development of a system that will seamlessly transparently pull all sorts of traffic, accounting, and usage information out of virtually any sort of voice communication infrastructure (be it TDM or VoIP based) and help make sense of it all for executive and management analysis. These are the type of systems that telco’s need, and need badly–the way to manage large amounts of information, make sense of it, and, in the end, also share on with their customers.
The problem is, however, that as I asserted at the beginning, telco’s don’t do software well. Software as a service from a telco? First, I’d like to see software providing a useful service at a telco…