Making money on value-added IP video MCU services

We’ve been saying for a long time that voice is trending to $0 revenue. Everyone is giving away voice minutes now – the latest gimmick being Verizon with its ‘call anyone on our network for free’ friends and family plan.  The peer-to-peer voice offerings on the market – like www.skype.com – provide more 10 penny nails to seal the voice revenue coffin.

Unless we see an effort by the telcos to purposefully degrade the ‘best effort’ basis services that are available today, peer-to-peer voice calling will work quite fine for many people with broadband.  As we’ve said before, we have used VoIP for our everyday communications, and with some notable exceptions, the quality is more than quite fine – it’s free.

Peer-to-peer is going to contaminate all the telco business plans for added revenue whenever two parties are involved, because these are the easiest forms of communication to handle without some form of intermediary server. So not only will voice trend to $0, but so will any form of peer-to-peer applications sharing and video services.  AOL already provides free video, albeit somewhat unreliably at this point. Other peer-to-peer offerings are coming out everyday as plug-ins to instant messaging platforms and other presence-based services. It’s just a matter of time before these are perfected to the point where they simply work over the Internet and a broadband connection.

So what will people pay for as an added expense in the near term?  We think the money is in the masses – that is, the masses on the same communications connection. For example, customers have flocked to three-way calling, benefiting both the communications service provider (where along with Caller ID and Voice Mail, such features have driven decent revenue to the carriers), and the communications hardware providers, where two line phones offer conferencing buttons and are considered a ‘must have’ for anyone who wants to bridge two lines (when was the last time you saw a two-line phone without conferencing).

But as you step up to four parties, five parties, six and beyond, it gets tougher to find such easy solutions. You get into the large multistream conferencing boxes, and network-hosted services, right away.

Group voice, group video, group Webcasts of PowerPoint presentations to hundreds of people, archival storage/replay… these applications are where the revenue lie.  These multipoint services are already well established in the audioconferencing world, but video is still nascent, with a few start-ups positioning themselves predominantly in the value-added IP multipoint conferencing play.

We think that telcos should plan on offering two types of services:  A managed IP services conferencing unit as well as a network-hosted service. For enterprise clients, the tradeoff between premises- based gear and network based functionality will run into the same bandwidth issues that the ISDN-based MCUs ran into when they were rolling out:. Specifically, if you are going to do this yourself with an IP MCU, you need to have the bandwidth to control and bridge all these different streams. The more streams you want to manage, the more bandwidth needed at your hosting location. For consumers, such hardware options are less likely to work – the approach for consumers is purely a hosted application. Few consumer broadband solutions (cable modems in particular) can handle the upstream bandwidth requirements for video over IP large group conferences.

And let’s not kid ourselves – some of these conferences can get big.  Santa Cruz Networks recently had a client that held a 365-person videoconference, which booked 4,413 minutes across 23 international sites. That’s a lot of bandwidth, and a lot of revenue. Not many companies can handle that varied an audience.

To date, the focus on IP multipoint solutions has been treated as a natural progression and add-on to point-to-point solutions.  We caution service providers not to focus on the peer-to-peer functionality, because it’s going to be hard to get revenue there; the big bucks (and pesos and euros…) will come from the group-oriented services. These requirements should be driving the decision-making for next generation IP conferencing gear.

Here are some issues to consider when supporting a group real time collaboration service:

1. Seamless Installation – Video/voice conferencing solutions based on the old H.323 standard utilize UDP direct IP to IP communications and do not seamlessly transverse NAT, firewalls etc. – (anyone who has tried to use Microsoft’s NetMeeting, or AOL’s new AIM 5.5 IM from behind a NAT router have experienced this problem).  To successfully support and launch profitable IP based group conferencing services, the installations need to be simple, seamless and straightforward.  If service providers need to manually install or provision group services, suffice to say, the service will not be popular nor profitable.  Using a client/server model is an excellent approach to IP-based group services.

2. Security – Look for the contact list/IM to be permission based, so if a company wishes to keep a ‘walled garden’ approach to group communications, they can.  If you are using the Internet for transport, you will certainly need some type of encryption, and SSL appears well suited to the dual task of security yet allows for a straightforward install as it transverses NAT and firewalls.  Some customers may also wish to run the group online meetings via a corporate VPN and further encrypt traffic via IPSec.  Finally, if security is the main issue (e.g., for a board meeting communications session), the service provider should consider offering the services via a private network and not the Internet – like MCI’s VoIP via Private IP service.

3. Bandwidth management – Look for centralized and easily adjustable bandwidth management by IT without having to install any hardware or software. Web-based consoles are a must. You need to be able to evenly distribute bandwidth among multiple users on a shared connection. Peer-to-Peer services can often swamp a network as there’s no awareness of other users and centralized bandwidth control. The software needs to constantly adapt to network traffic. Look for options that have failsafes built into their architectures, like dual paths for voice channels.

4. Scale – Scale has several different aspects to it.  Some carriers will want to start as a peer-to-peer offering and then grow to a group service.  Others may start out with small groups and then grow to support up to “n” number of people in an on line meeting.  Nothing is more frustrating than finding a useful new service or product and quickly reaching the capacity limit.  The on line meetings need to scale to support dozens of concurrent users and, ideally, hundreds.  On line, group collaborations tend to be viral and somewhat addictive – the number of concurrent on line users in a single meeting will grow.  In addition to concurrent users, the service provider needs to ensure that the underlying architecture of the group conferencing solution can scale to support hundreds of thousands or even millions of customers.  Lots of hardware companies ‘talk that talk’ but few ‘walk the walk’.

5. Integration with new services – Anticipate that the underlying group voice/video/IM and collaboration services will only be the beginning and that the service provider will need and want to integrate a number of other offerings – for consumers, multi- player games and business customers, archiving of meetings, etc.  The service provider needs to insure the deployed solution can be easily enhanced to support new services. Look for open, standards-based interfaces.

6. Pricing/CPE – Similar to the early broadband days, and the recent cellular battles, there will be a ‘land grab’ for new service revenue in the online group meetings space.  We believe that a successful model will include ‘free’ CPE, i.e., camera, headsets, etc., for a commitment to some period of service, perhaps one year.  This type of pricing is good for both the service provider and their customers, as it gets the service provider the committed annual revenue and customers they need and it saves the customer from having to buy the necessary CPE. It also eliminates possible CPE incompatibilities.

Multipoint video is going to be a big winner for all. The fast pace of Webcams, cell phone multimedia, and other similar services and products show this clearly. The trick is getting ahead of that curve to get the Video over IP real estate before others do.