VoIP: Not yet ready for prime time

Having made major inroads in the wholesale telephony carriage business, VoIP is a service and a technology that is now on the verge of eclipsing the circuit-switched retail PSTN. Right‾
Wrong. There's more than an element of smoke and mirrors at work in that popular assumption. Actually, other than in the form of PC applications or enterprise solutions, the number of retail VoIP telephone terminations worldwide is insignificant compared to the estimated 1.2 billion regular TDM phones in use and the 2.5 billion circuit switched mobile phones on air.
'Principally it's still a computer market rather than a residential or enterprise market with large scale usage,' asserts Keith Miller, CEO of service delivery platform specialist Appium. 'I do think residential VoIP has been seriously overplayed in this decade. The current way of connecting to ADSL is done almost entirely via PCs and it requires a lot of knowledge to be comfortable setting it up, and is a niche market to date.'

Creating the market
This is not to dismiss the considerable achievements of the bigger VoIP pioneers such as Vonage and Skype. Vonage says that as of March 31 it had completed over three billion calls over its network while the Skype peer-to-peer VoIP operation registered 100 million registered user names on April 27.
'Skype's been fantastically successful, if for nothing else than highlighting to the customer that this technology works,' judges Jef Kent, chief of operations at discount international calling company and soon-to-be UK VoIP operator Liquid Telecom.
Nor is it to denigrate the serious impact that VoIP is having, and is forecast to continue to have, in the enterprise space. Both Vonage and Skype have enterprise as well as residential users - 30% of the total customer base in the latter case - and, driven by the ongoing replacement of legacy TDM PBXs with IP PBXs, the use of VoIP products and services by business enterprises of all sizes is seriously on the up. According to a recent report from WinterGreen Research, VoIP is anticipated to have 60% penetration in enterprise markets by 2008. WinterGreen also calculated that revenue from worldwide IP-telephony products was 55% higher in 2005 than 2004.
That's some of the good news about the end-user VoIP market. Somewhat less good is the fact that that some of the numbers aren't all that impressive. At the end of Q1, for example, Vonage possessed a modest 1.6 million active lines. And although Skype has said the numbers are negligible, it's been suggested that some of its users have downloaded the application to several different devices with different user names. More seriously it's not clear how many active and regular, money-spending users Skype has - at the time of writing this article there were just under 6.5 million on line.
Many industry observers opine that in the long term the standalone business case for these and other VoIP start-ups is hedged with uncertainty, with already low margins being squeezed by intensifying competition. Pressure on VoIP pricing is being exerted by the start-ups themselves - Skype's recent offer of free PC-to-phone calls in Canada and the US wasn't thought to have helped the cause of Vonage in North America - and also by established cable companies and incumbent telcos.

End of the 'free' ride‾
Added to this is a growing conviction - to date most loudly articulated in North America - on the part of the infrastructure owners that VoIP providers and bandwidth-intensive enterprises such as Google and Yahoo! have so far had something of a free ride on the network, and such users now need to pay a more commercial rate.


'There's an awful lot of network out there, and ultimately the guys that own the network have got to make money back on the investment they're making,' reasons Mark Ashdown, sales and marketing director at multi-service platform developer DigiTalk.    
Meanwhile, although the share of the enterprise telecoms market that is falling to VoIP is substantial and increasing, a peak - at least in terms of hardware sales - is thought by some to be in sight.
A report from Juniper Research forecasts that revenues from the sale of business VoIP hardware and software will reach $5.5 billion by 2007, driven by rapid new business growth in China, adoption of VoIP across existing western businesses and an echo of the year 2000 equipment purchase boom. The report then speculates that sales would stabilize and fall to around $3 billion toward the end of the decade as the rush to IP tails off, the Chinese market begins to mature and as competition from low-cost suppliers in China and India eroded prices.
Of course the retail VoIP story isn't just about the enterprise market and start-ups such as Vonage and Skype. Although there was an earlier and fairly widespread insinuation that the advent of end-user VoIP might turn out to be the 'slayer' of incumbent telcos, in practice most of the long-established service providers now have retail VoIP products and strategies. Part of this is defensive, part to do with the need for better house keeping.
'We are increasingly facing competitors that do not work on the basis of an extensive infrastructure - as we do - but on the basis of IP or software,' noted Kai-Uwe Ricke, chairman of Deutsche Telekom's management board speaking at the company's shareholder meeting in May. 'Take companies like Skype or Vonage, Google or Yahoo! These companies may well be using our infrastructure but have totally different cost structures from ours. They can produce voice and data services with far lower marginal costs. To survive in this competitive environment we have no choice but to shift our own production over to IP. To ensure we remain competitive, we need to significantly improve the efficiency of our production and slash our costs of production by implementing a radically simplified all-IP infrastructure.'
As well as improved productivity Ricke spoke of the new services that IP would enable. Like many other of the larger incumbent telcos, Deutsche Telekom has ambitious plans for services such as IPTV to sit alongside voice, Internet access and fixed-mobile convergence services. Some analysts argue that by bundling VoIP with an extended repertoire of IP services the established telcos will limit the opportunity for, and lessen the appeal of, service providers that are focused mainly on VoIP. So could the eventual big hitters in the VoIP league turn out to be the incumbents rather than the start-ups‾
'After 2010, PSTN will no longer be the main revenue generator in developed countries,' states Malik Saadi, author of 'IMS Opportunities and Challenges,' a recent report from Informa Telecoms & Media. 'There will be no justification for big operators to reserve a whole network for traditional PSTN voice traffic. This trend will increasingly push operators and network owners to gradually migrate their subscribers from traditional PSTN to VoIP.'
In much the same vein, a report from visiongain argues that VoIP products and services offer operators new revenue streams and means of growing their business.


'Operators should focus on deploying IP-based networks rather than worrying about possible threats from VoIP. They ought to look at VoIP as an opportunity to develop new revenue streams,' says visiongain analyst and the report's lead author Dr Jean-Pierre Aubertin.

How far, how fast‾
But how far along the all-IP trail the incumbents are likely to go, and how fast, is moot. 'If a telco opens up all of its customers to VoIP connectivity then anyone can offer service. This is far bigger than equal access as it's simple and cheap to attack that revenue,' contends Appium's Miller. 'Next generation networks will not seriously reach a lot of residential customers worldwide until the middle of the next decade, and most of the current ones only extend VoIP to the telephone exchange as well, which will slow easy-to-use residential VoIP even more.'
'There's nobody out there ripping up networks and starting again,' adds Ashdown from DigiTalk.
Appium, which is targeting the service delivery platform and IN replacement markets as the first real commercial step to NGN rationalization and an eventual SIP/IMS services infrastructure, finds it difficult to imagine residential VoIP overtaking the PSTN until the end of the next decade at the earliest. Somewhat contrary to the received wisdom that retail VoIP is a wireline play, Miller sees considerable potential for wireless retail VoIP in the form of hybrid GSM and Wi-Fi/WiMAX services, not least because of the current high costs of international roaming.
Roland Burri, executive VP for product management at specialist Swiss  provider of wireless roaming services Comfone, agrees with Miller about the prospects for cellular/wireless LAN VoIP. 'A crucial factor for a successful service if it is to reach mass market momentum is how user-friendly and convenient it is to use,' he reasons. 'With the increased availability and spread of new dual-mode devices - GSM/WLAN - which are the size of usual mobiles, have easy to use functionality and are simple to switch between GSM or WVoIP calls, wireless VoIP has potential to be extremely successful.'
But hybrid cellular/WLAN VoIP service provision has its own challenges. As Ashdown points out, many cellular operators are going to have to countenance losing revenues to hotspot owners. 'Some of the big name manufacturers have the capability to put WLAN and Wi-Fi access in a handset, but would an O2 or a Vodafone be happy to be doing deals with Nokia or BlackBerry or whoever when they're enabling alternative service providers to come into their networks and steal their revenues‾' he questions. 
According to Burri the other approach - where cellular operators build up and invest in their own hotspot infrastructure - is very time consuming and requires new in-house knowledge and skills. 'Or they outsource the whole WLAN/Wi-Fi component to a specialist service provider,' he suggests.
Either way, though, nobody really doubts that the future of voice communications, as with most other forms of telecommunication, belongs to IP. But if VoIP is coming to a terminal near you, it may not come all the way, it may not be the type of terminal you expect, and it may not arrive as soon as you expect.

VoIP Now


- Limited mostly to PC applications and enterprise solutions
- Incumbents have countered threat from VoIP start-ups with their own retail VoIP products and strategies
- Number of active, money-spending Skype users dwarfs stated 100 million registered users (only 30% residential)

On the Horizon
- Strong upside for wireless retail because of high costs of international roaming
- IP-based networks give operators opportunity to develop new revenue streams
- Business VoIP hardware / software sales to peak in 2007

The missing links

Although the initial attraction of VoIP to an end-user was low cost in the case of consumers, and low cost and the benefits of integration and flexibility in the case of enterprises, the technology offers a number of capabilities not provided by the PSTN. One is the nomadic utility of the VoIP terminal - to date generally a laptop.
'The roaming capability is very compelling,' argues Mark Ashdown, sales and marketing director at DigiTalk. Another is presence. 'It's the great feature of VoIP as compared with a PSTN,' ventures Jef Kent, chief of operations at Liquid Telecom.    
There are, however, some respects in which VoIP is not yet the equivalent of the PSTN, and there are still some gaps in the repertoire of VoIP.
On a rudimentary level, there are residual regulatory issues with the technology, and some governments have acted to ban or limit its use to safeguard the PSTN revenues of incumbent telcos. Elsewhere, regulators have had to give attention to attempts by network owners to 'tax' providers of VoIP services that compete with their own. Just how much telecoms regulation should apply to VoIP service companies is still the subject of debate in some markets.
Voice quality was once a major bugbear of public VoIP services, but this has been greatly improved in recent years. Is it now comparable with that of voice over the PSTN‾ Yes and no seems to be the answer.
'In places where you have broadband and IP bandwidth availability I don't think quality is an issue. Where the markets are not so well developed in terms of broadband or IP bandwidth being generally available, then QoS is an issue,' suggests Ashdown. 'But to be honest that's the case whether it's an IP call, or a GSM call or a PSTN call.' 
Some observers feel that the ease-of-use of VoIP could stand major improvement if the technology is to displace the PSTN. 'The key to doing it, and doing it successfully, is that it has to be as simple as plugging a telephone into the wall. Why should it be more complicated‾' asks Kent.
Interestingly, in a commentary on Skype's announcement of free Canadian and US PC-to-phone calls referenced above, Jan Dawson, VP of the Ovum consultancy's enterprise practice, remarked on '"&brkbar;the  hassle of using a PC-based service rather than a normal phone and the quality issues associated with Skype"&brkbar;' and thought these might contribute to some potential users passing on the offer.


There may still be gaps in the VoIP monetization chain, some doubtless arising from the possibility of creating bogus origination and termination gateways for calls. 'I think there are probably revenue assurance issues with VoIP, just in terms of the new opportunities for revenue leakage,' says Martin Creaner, CTO of the TeleManagement Forum. 'There's probably a lot more to be done on that.'
And although the PSTN is far from being impregnable, security is an area in which VoIP is viewed as having additional network vulnerabilities. These range from irritants, such as the less than delightfully named spam over Internet telephony (SPIT), to the more serious possibility, highlighted this year by the Cambridge-MIT Institute-funded Communications Research Network that the ability to dial in and out of VoIP overlays allows for control of an application via a voice network, making it almost impossible to trace the source of an denial of service attack.
- John Williamson

Skype: Mind your own business
Citing potential threats to security, one telecoms analyst in November last year urged extreme caution regarding the use of Skype in the enterprise. Ross Armstrong of Info-Tech Research Group outlined five possible weaknesses in the peer-to-peer VoIP application.
- Skype is not standards-compliant, allowing it and any vulnerability to pass through corporate firewalls.
- Skype's encryption is closed source and prone to man-in-the-middle attacks, and there are also some unanswered questions about how well the keys are managed.
- Enterprises using Skype risk a communication barrier with countries and institutions that have already banned the service.
- Skype is undetectable, untraceable and unauditable, putting organizations that are subject to compliance laws at risk.
- The question of whether VoIP calls constitute a business record is a legal quagmire, and throwing Skype into the communications mix further clouds the issue.
Have Armstrong's views changed in the interim‾ Now Armstrong says: 'All I would add is a caution - Skype has become very popular with home users and very small companies (i.e. those with less than 10 employees). My original point is that if any enterprise wishes to use Skype for business purposes, it had better be ready to secure and control it through centralized IT administration, just as you would for email, instant messaging, or any other form of corporate communication. Other than that, my position hasn't really changed.'