I have started this new trend of inviting experts and gurus to share their thoughts and publish their essays over here on GigaOM. Earlier this month, Dan sent us The VoIP Insurrection. This week pitching please welcome Aswath Rao, who has 20 years of experience in the telecommunications field, makes his second appearance on the blog, with our open letter to Steve Jobs and Apple, telling them how to really bring VoIP into the Apple fold. Rao had earlier contributed Why Skype is No Different? Rao is telecom industry veteran and for past 5 years he has been working on VoIP related issues. Long before intelligence at the end became acceptable, he advocated “functional terminals” in ISDN. His proposal for Inter Connect Function has been incorporated in the TIPHON architecture and currently it is known as Session Border Controller. He has developed ways to offer PSTN subscribers many of the features available to VoIP subscribers. In a classic engineer’s vernacular he suggests to Steve Jobs, “please build a device that facilitates users to communicate with their peers without the need for any other third parties and at the same time provides a rich man machine interface through which users can easily invoke features.”
Dear Mr. Jobs
Recently you have been publicly advised on why you should heed the call of VoIP: some like Salkever suggest that you should build on iChat and follow Skype’s model, while others like Wallingford say that iChat’s lack of interconnection to PSTN is a fatal flaw and so you should acquire a company like Packet 8 or BroadVoice. While agreeing on some of the points raised by these two authors, my recommendation is radically different and goes against the common view in the industry. The purpose of this letter is to elaborate on my line of thinking. As you read this please keep in mind the famous ad taken out by your company during the 1984 Super Bowl, which established the idea of “empowerment”.
A superficial analysis would indicate that it is better to enter the service provider business. Most of the current market activity centers on offering VoIP service to consumers. The proposition looks very tempting: it is easy to enter the business with very little capital expenditure and eager wholesalers facilitate interconnection to PSTN; in one swoop one can have national, nay, international presence; lenient regulatory environment, and the list goes on. Vonage has enrolled more than 150,000 subscribers, but boasts very little infrastructure. Most of their expenditures seem to be related to customer acquisition. Skype claims it has gone one step better: almost zero acquisition cost; no infrastructure whatsoever, save for a directory server; still can support millions of users. But a deeper analysis will indicate that this level of success can not be sustained.
All service providers offer two basic services: 1) provide a directory service whereby people can get the currently reachable IP address of a subscriber, 2) assist users in addressing NAT/Firewall traversal problem. Apparently these two are not value added services, since all the service providers offer them for free in the hope that interconnection to PSTN will provide sufficient revenue. Evidently the tariffs offered by VoIP service providers are very low compared to the traditional service providers. This difference is attributed to the efficiency afforded by IP technology. But PSTN service providers can and many do use the same technology in the backhaul and realize the same operational efficiency. So the explanation must lie somewhere. The current regulatory environment is the major contributor for this disparity. Currently, the VoIP service providers are free from many regulatory fees that are imposed on the PSTN service providers. But this will not continue for long; eventually both of them will have similar regulatory imposed cost structure. Given the users’ lethargy and limited price differential, the incumbents can and will aggressively fight the insurgents and prevail. (It is fashionable to suggest that PSTN is like a dinosaur. That it might be; but VoIP service providers are not the fabled meteorite. Indeed a thriving PSTN is a precondition for their survivability. This irony seems to have been lost on the industry.)
As if anticipating this, the industry is betting on additional revenues from new services/features. Many VoIP proponents have suggested that VoIP technology will facilitate new services that we have not yet imagined and that PSTN can not offer these services because the technology is limited. Even this point withers away in the light of some analysis. Some of the anticipated unique services are virtual numbers, voice mail via web interface/email and buddy lists. But PSTN technology doesn’t prevent offering virtual numbers; iobi, a new service offered by Verizon already offers the other two. Services are enabled because of the richness of the call control signaling protocol and not the transport technology. In this respect, PSTN’s Q.931 and VoIP’s H.323 and SIP are equivalent.
Since IP affords an end-point to signal any other end-point in the network, end-to-end signaling is possible in VoIP. This means that almost all the features can be realized by the end-points without the help of a service provider. Indeed if you look at the way features like call waiting, transfer and conferencing are accomplished you will conclude that these services/features are realized by the user’s equipment. Even if we could develop a revenue producing service, the players end up offering it for free because the industry is so competitive. When Free World Dialup was started, the anticipation was that subscribers will be charged for features like voice mail. Even Skype echoed such a plan. But now almost all the service providers offer voice mail for free.
In summary, given low barrier for entry, uncertain revenue potential, formidable incumbent competitors and limited feature differentiators, I submit to you that you should not enter this business. To top it, it is contrary to your motto of user empowerment. Due to the popularity of Skype, it has been suggested that P2P is a technological breakthrough. P2P is a useful technology for applications like file swapping. In that application, a user needs access to a file but does not know where in the network it is available. P2P technology facilitates in locating a potential source. But communication is different. A user knows precisely whom to contact but may not know how to reach that person. There are more efficient and non-public ways to get that information instead of using a labyrinth of supernodes. So, do not consider Skype like technology or other “serverless” architectures just because they are currently in vogue.
The industry is placing a premium on PSTN interconnection service while at the same time trying to supplant in it. This is strange. Email didn’t provide a mechanism to interconnect to postal mail or to fax. Why not we as a society have both the networks and use them when and where appropriate? Indeed, James Crowe of Level 3 suggested this in an FCC Forum. It is interesting that nobody picked up on his point or challenged him. So if we exclude interconnection to PSTN, then it is clear that the real need is for a well conceived VoIP client. Many others have observed this, but immediately have focused on Vonage or Skype without realizing the irony.
So what are the characteristics of a well conceived VoIP client? In accordance to its corporate credo, Apple can develop a product that truly empowers the users and does not enslave them to a service provider. For VoIP to be effective, it still remains that the users realize the two basic services – getting the reachability information and solving NAT/Firewall traversal problem. A product could be built that uses only the currently available technology and business models to realize these two services autonomously.
There is one other matter that is being ignored by many in the VoIP industry. In the final analysis, the user can invoke services and features only through a Man-machine interface and that will determine the usability of these services/features. Currently the popular arrangement is to provide a traditional phone with 12 buttons. This is so limiting even for the current list of features, let alone new and revolutionary features. The importance of MMI can be learnt from your yin-yang twin, Microsoft. You may remember that some six years back, they introduced a cordless phone that was tightly integrated to their telephony software. The phone never caught on and I theorize that had they solved the MMI issue it would have been successful, especially with those who access VoIP service through an ATA. So the development of a product that offers a rich user interface (Apple’s strength) will be a welcome contribution. This will be the “thumb wheel” to VoIP client.
To summarize, please build a device that facilitates users to communicate with their peers without the need for any other third parties and at the same time provides a rich man machine interface through which users can easily invoke features. I sincerely hope that you reach the same conclusion.