Moving away from the commodity game: the evolution of the network is accelerating, making speed to market—not end-to-end control—a key ingredient for

Telecom Asia: During your keynote talk at Carriers World Asia last month, you said we're in year ten of a 25-year network build out.

Vab Goel: If you really think about it today, the Internet is still in a very early stage. IP is a protocol that doesn't really care about the underlying physical interface--it runs over cable, PSTN, TDM and over the wireless network. But right now multimedia and video are still not happening over IP networks. The day is going to come where you'll be able to watch any movie or any TV program from anywhere sitting right in your living room. TVs are going to come with an Ethernet interface, and you'll be able to see what channels are playing in the UK and select what you want. What has to happen to allow that? The network has to change. If you look at most of the broadband networks getting built today, they don't have enough capacity. DSL is the first phase of the deployment. In the US you can't get 300 kbps most the time.

The network is going to change through new equipment deployments. None of that is going to happen over night. The next step is how to really integrate video with the IP network, and then how does it integrate with Wi-Fi in the home. The challenge is making it as easy as turning on the TV. The biggest mistakes we can make are thinking this is the best we can get or that it will all happen in the next year. Is it going to be ten or 15 more years, I can't predict. But one thing for sure is that a lot more innovation is coming and a lot more work has to be done. So the view is that the industry has to be ready for change and to think out of the box. And it has to invest in new ideas and technology because the status quo is not going to get us there.

How does this impact incumbent carriers?

It puts them back against the wall. The more applications are available over the Internet, the more their top-line revenue goes down. I think there are still opportunities for carriers and a tot of possibilities for new services, but they need a service delivery platform that is auto-provisioning and low cost. They also need to have a partnership model vs thinking they need to own everything. They have to think beyond where do I own the fiber and look at the AOL model, the IBM hosting model and the Google and Yahoo models. They have to look at what they have and totally change the way they do business. It can't be a 5 or 10% improvement; they need to look at 200 or 300 or 400% improvement--both in the business plan and the network architecture.

Going to Nortel, or Siemens or Lucent and asking them to be your supplier may not be enough. Carriers may have to take a stronger approach and say they're going to partner with entrepreneur companies and build new services which give them a 200-300% shift. Those that don't do this, their days may be numbered.

Who would have thought that AT&T would be sold for half its revenue. This is a big alarm for everyone and people have to accept it or we can stay in denial. AT&T for the longest time was in denial. They didn't have a single IP network--they had multiple networks and were late coming to the party. Sprint was the first telecom company to adopt Internet--AT&T was two to three years behind. Time to market is going to be key, and if you don't take advantage of it, someone else will.

What do you see as the pathway to higher margins?

Like I mentioned before, carriers need to think through forming partnerships and figure out how they're going to have a global footprint without investing capital in each market. This is self serving, but they need to partner with companies like Virtela that don't own the network but have design capabilities, network capabilities and a service delivery platform, which allows them to deliver complete managed network services, because every enterprise needs connectivity within the country, within the region or globally. They need to take advantage of services like video conferencing, which is becoming commonplace now, but many enterprises are doing it themselves and carriers are getting no revenue out of it. They need to offer directory services and interconnection on the IP layer between the different service offerings.

Say you have a Tandberg as a product as well as a Sony product, can you as a service provide offer a directory, which is a simple software, so enterprises, even though they're not within the same company, can communicate with each other. They basically need to open the platform--you can't keep the network closed.

What are enterprises looking for in service providers?

I think they are most frustrated by customer service and support, which seem to be at an all-time low. Every time a customer calls now, the reaction is "oh, my God he's calling me one more time." The challenge is that enterprises have issues, because he just got attacked by a virus, and he's going to need service providers to help to figure out which IP address these attacks are coming from. So service providers really need to update their backend systems and network management and monitoring systems so they're more proactive. And enterprises are willing to pay for it, because 100% up time for a real-time enterprise has never been more critical than now. But you can't just throw more people on it to provide the required level of service and support. It requires a new provisioning system and base of network management and monitoring, and requires different skill sets. Telecom companies need to be ready to change some of the skill sets they have and empower people who come from non-telecom environments.