EuroTelcoblog No. 45: Wednesday 9th June, 2004 - an explosion of convergence

Yesterday was day one of the Voice on the Net (VON) Europe 2004 show at London Olympia, where I was honored to be given the closing slot in the morning's presentation line-up. Send me an email at james.enck@dir.co.uk, and I'll gladly send you the slides (Daiwa IT department would kill me if I tried sending it simultaneously to my distribution list...). Parts of it may be difficult to interpret without the accompanying commentary, so later in this post, I will provide some comments to help with interpretation for those who weren't there.

We saw and heard a lot of interesting things yesterday, which we also cover later in this post, but first we'd like to give attention to three announcements made by Northern European incumbents today, which look to us like shrewdly-timed return fire to the IP guerillas amassing outside the gates of telcoland (or in the Olympia conference center, to be more precise).

BT Group 21st Century Network
====================================================================

BT Group today announced more details on the timetable its 21CN program to deliver converged IP services throughout the UK. The principal components of today's announcement were:

1) A goal of mass migration to IP over a five-year period, starting in 2006, with the majority completed by 2008.
2) The company will initially conduct a PSTN bypass trial at two nodes in Cambridge and Woolwich (London) from October 2004, later expanding to the Faraday exchange in London.
3) 1,000 customers within the trial areas will be placed on an end-to-end IP telephony service (i.e., over its own IP network rather than the public internet) by January 2005, expanding to 3,000 customers in June 2005, as a precursor to wider rollout.
4) Initially Multi-Service Access Nodes (MSANs) will be installed in 18 exchanges around the trial areas, and by 2009 BT aims to make broadband dialtone available to around two-thirds of its customers, with large-scale migration kicking off in 2007.
5) Also starting in October 2004, BT will conduct fiber to the premises (FTTP) trials for 1,500 homes and businesses in five locations, running to September 2005.

The program will be funded out of BT's previously identified GBP3bn annual capex, of which 2/3 (increasing prospectively) of annual spend will be dedicated to the 21CN program. The transformation program targets delivery of GBP1bn in recurring cash cost savings on an annual basis by 2008/9.

With this announcement, we think BT is once again demonstrating its leading position among PTTs in awareness and adoption of new technology and services. Indeed, BT Wholesale CEO Paul Reynolds today referred to the need to "grasp the nettle" of transformation to all-IP, and we agree it is an absolutely inevitable development. We think this move will prefigure a rush of similar announcements from other incumbents in Europe in H2 and into 2005. The main question for us is, given the current competitive landscape in the UK, and rapid and accelerating pace of development in the IP application space (see our section on the VON show below), will a five-year migration be fast enough? Last week we saw Carphone Warehouse revise up its target for talktalk carrier pre-select customers to 900k by March 2005, implying a weekly net add rate of 9,900, down from the 11,200 seen in the March quarter. We think this target looks easily achievable, and there is more to come in the broadband space in the wake of the Bulldog acquisition last week by Cable & Wireless. Additionally, we also noted with interest the contract awarded last week by Tiscali to UTStarcom (supplier to Yahoo! BB) for IP DSLAMs covering eight countries in its European footprint. Given previous comments by Tiscali we think we can expect to see a serious push in the UK market via an unbundled DSL IP triple-play similar to that already on offer from Yahoo!BB in Japan and Free.fr in France, perhaps by the end of this year. Obviously, there is more involved in the BT plan than just the voice and broadband access markets, and there were references today to a range of new services, including video, which will be enabled in such an MPLS network environment. We think BT continues to make some audacious moves and clearly has its eye on the ball, and our ongoing neutral stance on BT Group is more related to what we see as an increasing lack of order and visibility in the industry (and the lack of confidence this translates into when it comes to forecasting and valuation issues), rather than any negative view of the company itself.

France Telecom VoIP and flat rate initiatives
====================================================================

As if we didn't have enough to focus on today, France Telecom today also unveiled a range of new services and tariff changes, including VoIP. They principally break down as follows:

1) A 2Mbps DSL product for EUR39.90, and a reduction in the 1Mbps product to EUR29.90, which brings it into line with the Free.fr product (minus the telephony service of course). The 2Mbps product will drop in price to EUR34.90 from 20th August.
2) Flat rate calling plans for PSTN users, under which a range of numbers can be specified and tariffs tailored to suit calling patterns. For example, if users want to identify five national fixed line numbers for unlimited calling after 6pm and at weekends, the monthly charge is EUR14, and EUR17 for 24/7 calls to the same numbers. For calls to all national fixed line numbers in France, the monthly cost is EUR20 for the 6pm and weekends option, and EUR30 for the anytime plan. These services will be available from 15th June.
3) From 20th August, Wanadoo DSL subs on packages of 512kbps or above will be offered VoIP via an ATA-based service for EUR20 per month, with an introductory promotional offer of EUR10 per month for the first year. Numbers are mapped to an 087 prefix, and national calls to fixed line numbers are free, as are calls to other Wanadoo VoIP customers.
4) A VoIP version of the EUR5 per month flat rate calling plan already in place on the PSTN, under which weekend and evening calls to fixed lines of up to two hours cost only EUR0.13 for the duration of the call. We think this is clearly aimed at potential VoIP customers who might not want an all-you-can eat plan, but who might otherwise move to an all-mobile/voice-over-broadband lifestyle and foresake the PSTN completely.
5) Unified account management for all FT services. FT states that
currently around 600k PSTN customers and 2m internet customers make some use of online account management facilities, but from July, French customers will be able to manage fixed, mobile and internet accounts from one location.

As with BT's announcements, we are impressed with what this says about FT's growing awareness of the need to "get with the program" in consumer IP and converged services, though it does not alter our underweight stance on the stock. Our concern with FT, as with all incumbents, is that we believe the IP voice world to be fundamentally a deflationary force, which will inevitably elicit questions about optimal headcount size (a very sensitive issue in France), and more fundamentally, what services will drive growth in future. We expect to have some clearer ideas on FT's thinking following the analyst day tomorrow, though we find it curious that a company so ready to embrace the internet has chosen not to webcast the event!

KPN's tentative foray into media
====================================================================

Though not as dramatic as the BT or FT announcements, we also had news today from KPN which serves as another datapoint on the road toward convergence. KPN today announced its intention to solicit tenders from leading cable companies to allow it to lease capacity on their networks for the provision of TV and radio programing, as is its prerogative under the Dutch Telecom Act. What exactly this says about strategy (for example, to increase KPN's visibility as a content provider in the eyes of cable customers) or the state of KPN's network (the press release states that the company has determined that its network is "still less suitable" for broadcasting services) are both open to speculation. The message for us is that we may see more of this as the PTTs try to define and refine their multimedia strategies.

A recap of our day at VON Europe 2004
====================================================================

Yesterday I was honored to close out the morning of presentations at the VON Europe conference in London with some thoughts on fragmentation in the industry. I will summarize this presentation at the end of this post. It was a good atmosphere, despite the apparent lack of air conditioning at the Olympia venue, and my rough estimate would be that something like 800 people attended over the course of the day (some photos can be viewed at organizer Jeff Pulver's blogsite http://192.246.69.231/jeff/personal/archives/000868.html).

Update from Norway
==================

I met a number of interesting people, and got some good leads to follow up for the future. I also met up with some more established acquaintances, such as Telio, which yesterday launched a pan-Scandinavian unlimited calling plan (to fixed lines) covering Norway, Denmark and Sweden. The new plan, which costs the same as their existing national plan at NOK159 (including VAT), is set to launch on 12th August, and CEO Espen Fjogstad said that an early check on the website showed that the new offering seemed to be stimulating a new customer addition every three minutes.

Meet Voipster
=============

In my presentation yesterday, I introduced, I believe for the first time in such a public forum, the existence of Voipster (www.voipster.com), a Dutch-based company of 20-or-so employees (with a team of software developers in Estonia), which has produced a P2P telephony service platform running on Overnet (please see EuroTelcorama No. 5 from 30th January, 2004 for more detail), which also allows break-out to the PSTN, SMS messaging, video, document sharing and a host of other features which differentiate it from other offerings in the market presently. A demonstration video can be viewed here (www.voipster.com/video.html). Download the file first and then view it, but be forewarned - it is in Dutch, and the scenario involves people collaborating on an autopsy, so it is somewhat visually graphic.

Speaking to the Voipster management face-to-face for the first time yesterday, it became apparent to me that they followed a largely parallel development strategy and timeline to that of Skype, but they are coming to market somewhat later, with quite a different set of features and marketing approach. Their intention is to remain a neutral software developer and provider, targeting the enterprise space, and more importantly for our focus on the consumer segment, as a white label solution for ISPs or telcos who want to position themselves in this space. Notably, they also want to establish peering relationships with other services. I am very curious to see what sort of newsflow follows from Voipster and will update as the business unfolds.

Takeaways from the Voice over Wi-Fi session
============================================

I had a lot of meetings yesterday and didn't make it to as many seminars as I wanted to, but I made it for part of the Voice over Wi-Fi session, which contained interesting presentations from The Cloud, Telio and Spectralink. The thing that interested me most about the session, however, was the moderator's questions to the audience between presentations, in which he asked for projections as to what the future might hold. One interesting question was about the timeline for full seamless multimodal interoperability in the wireless world (2G, 3G, Wi-Fi), and the majority of the audience seemed to be roughly evenly split between 2005 and 2006(my hand went up for 2005). Another question was to identify the key inhibitors to the take-up of VoWi-Fi, and predictably the largest number of hands up was for battery life issues and security, but I was surprised at how few people cited QoS issues. This seems to confirm our view that, as with cellular, adoption of VoIP in the wireless arena is likely to be more about ease of use, cost savings and features than about ubiquity of coverage and iron-clad reliability. One other interesting question to the floor was about the form of future devices - would they look like the devices of today or more like the Apple iPod? My money was on the iPod, along with a few others, but we were overwhelmingly outvoted. One other interesting development in this session was a statement from Telio's CEO about some of the promising opportunities that applications written in Symbian OS can give to access independent voice providers in the 3G world, and we are very eager to follow up on this angle.

The rise of Citynets
=====================

I also attended an interesting, but surprisingly poorly populated session on the issue of Citynets, or networks funded and created by local governments to achieve socio-economic goals (once again, see our EuroTelcorama No. 5 for a detailed discussion of this issue). Unfortuntely, panel member Esme Vos, author of the Muniwireless site focusing on municipal Wi-Fi projects, was a no-show, but we were treated to presentations from two people with hands-on experience in building such networks. Bertrand Chauvet of netcentrex tracked the evolution of Fastweb in Milan from municipal dark fiber initiative to full-blown triple play service provider, and highlighted the inflection point in subscriber growth which accompanied the introduction of bundled IP telephony which could break out to the PSTN. He also offered comments on video service which echo our own thoughts and also what we are hearing from our friends in the cable industry - video content is a high cost item to the service provider, and should really be viewed as the teaser to get customers in. What keeps them and drives the cashflow is the availability of voice, and increasingly in the case of Fastweb, video telephony. This, in our view, further highlights the difficult balancing act required of the incumbents in the new world. Mr. Chauvet also highlighted some recent projects in France, such as the FTTH project in the Pau region, in which service providers such as www.ipvset.com are offering a 10Mbps internet/television service plus IP telephony over the municipal infrastructure for EUR30 per month. The availability of the infrastructure was cited as one of the key ingredients in attracting a new university campus to the area.

The other Citynet presentation of interest was from Byron Sherafatmend of E@glecom (that's not a typo), who has constructed municipal Wi-Fi networks in Ville de Vevey and Lauzanne in Switzerland. The model here is similar to those we tracked many months ago in our PTT Pulse publication, and centers on indirect revenue generation for the area under coverage, though the original motivation in the Vevey project was to stimulate e-voting in the area. However, following a promotion of the availability of Wi-Fi "dans la rue", the sale of coffee in local cafes went up by 30%. The Vevey project currently covers 20% of the population via 10 hotspots and can handle 250 concurrent sessions, and coverage should be ubiquitous next year, partially funded by a hike in local service charges by the city government. Vevey is apparently looking at offering VoIP services over its infrastructure in future. Eaglecom has set up and run 60 free hotspots to date, and targets 400 by year end. A list of free spots in Switzerland can be viewed at www.freespots.ch and includes a large number of spots constructed and managed by E@glecom in partnership with service station chain AGIP.

Niklas Zennstrom of Skype
==========================

The last event of the day was the predictably standing-room-only keynote by Skype co-founder/CEO Niklas Zennstrom, who gave the audience some insight into Skype's thinking about the service. Mr. Zennstrom showed a map of the world with an approximate breakdown of Skype's 6m registrations to date. Interestingly, despite the global scope of the application and the frequent erroneous identification in the press of Skype as a US company (it is registered in Luxembourg), North America only accounts from around 16% of registrations. Europe dominates at 64%, with Asia at 13%, Latin America at 5% (which seems huge relative to broadband penetration in the region) and Australia/Africa accounting for the remainder. Given all the attacks from the SIP brigade and traditional telcos about Skype's current standalone nature, it was interesting to hear Mr. Zennstrom confirm that the company had taken the instant messaging model as the example of best practice for user interface (which is an issue I covered in my presentation), and also that PSTN connectivity was not viewed as a core activity for Skype, but as an add-on feature to be used opportunisitically, which he likened to the J-Fax service. Nevertheless, Skype Out, the PSTN interconnection feature is on track for release "pretty soon," to be followed by "Skype In" to allow users to receive incoming PSTN calls. On the issue of interoperability/inconnectivity with other VoIP protocols, Mr. Zennstrom was clear that Skype considers this an important issue, but also stressed that innovation was drawn from user feedback, where such demand is not currently strong. He said that demands for features such as video and the availability of Mac or Linux versions of Skype were currently much stronger. He also stated that interoperability with other VoIP protocols would not initially be part of the Skype In feature either (this may account for Voipster's stated interest in establishing peering relationships). On the issue of the revenue model, I got the feeling that a lot was in the pipeline beyond the pre-paid solution for PSTN breakout, but there was little detail forthcoming. A few examples of what might be on offer included conference calling features for more than the current five participants offered for free, and we are interested in seeing what sort of services might evolve out of the company's decision to allow a Skype API for third party developers. However, Mr. Zennstrom stressed again the issue that we have repeated since Skype emerged - its cost structure is such that it does not need a huge number of paying customers to be profitable. He likened the service to Yahoo!, where a limited number of users may spend around $9 per year on paid content or services. A question from the floor raised the issue of whether ISPs would begin blocking Skype packets to prevent dilution of their own nascent VoIP offerings, and Mr. Zennstrom again repeated the analogy with Hotmail, which has never been blocked despite providing a service in direct competition with the connectivity service provider. To do so would promote churn, and this echos comments from the MSOs presenting at VON in Boston last September.

Daiwa presentation
====================================================================

Going back chronologically to the morning session, here is a rough guide to what I said there for those who weren't able to make it (email me and I'll send you the slides). Much of the content derives from previous editions of this blog, as well as much longer works such as EuroTelcorama No. 5 and our Global Telecom Monthly features on VoIP and broadband, and so may be familiar to regular readers.

Slide 4 - setting the stage - broadband household penetration in Europe >17% in Q1, Q1 annualized growth rate implies 50% growth this year

Slides 7 - 11 highlight the generational change issues facing incumbent telcos - instant messaging use (the three strongest platforms average 250 minutes a month - equal to typical outgoing minutes of use on the PSTN in Europe), less reliance on voice, greater use of rich media applications and independent self-provision of applications. The implication for telcos is that if voice is really just another application on the web (as BT acknowledged today), then getting and maintaining the user's attention/buy-in may be tough.

Slides 12 - 15 deal with new user interfaces and hybrid services. I opened discussing the popularity of social networking sites such as Orkut, which offer communities of common interest, simple messaging (email), some calendar reminder functions, etc. all driven by the interconnectedness of users. What could be achieved by embedding a richer set of communications tools (IM, voice, video)? Related to this is the statistics from late last year showing that Friendster users spend nearly two hours a month on the site - this is a sticky application. Now we see other services taking this social networking paradigm on board, and the examples I cited were ICQ Universe, Dodgeball (a fusion of location based info service and social networking) and xfire (presence management for locating friends on interactive gaming servers). From there, the argument was that we might expect to see voice embedded anywhere, and I showed the glophone from Voiceglo and Myphonebooth.com (both interesting examples of browser-based softphones which we think ISPs or others might seek to white label), as well as the new version of Yahoo! IM, which has a huge range of features including voice and video. Lastly, all of this use of voice as a service differentiator makes me wonder how long Google will stay out of the game - a point I have raised before.

Slide 16 - A response to a recent diatribe against Skype by the founder of SIPphone, who lambasted the service as something only "hardcore geeks" would do. I made the point that every time I log onto Skype I see at least 300k other users on line, and that the assumption that the interface was something users would find alien seemed to fly in the face of the statistics we have seen regarding the rampant use of IM among younger user segments. We should not underestimate the power of this interface.

Slide 17 - Here I was playing Devil's advocate, posing the question of whether Metcalf's Law is really all it's cracked up to be. In other words, does it matter to me if I am connected to 300,000 people on Orkut or if 6m people have registered their copies of Skype? In the case of both of these services, the people I know within Skype or Orkut were known to me previously, so why do I need a huge network? What if there emerged a service which could enable a small group of users to establish a mini-Skype among themselves? I then highlighted three encrypted file sharing/IM applications (WINW, W.A.S.T.E., and BadBlue - all of which I have written about previously) as examples of services which might evolve to embrace voice at some point. Not sure if anyone in the audience liked this idea, but I thought the issue needed to be raised.

Slide 18 - An example of something I found the other day entirely by accident after wrongly entering a search query on Google. TeamTalk and Tiny TeamTalk are VoIP softphones with large scale conferencing functions and adjustable audio quality to accomodate narrowband users. Tiny TeamTalk is a Windows CE version which could sit on a wireless PDA, as with the lite version of Skype. This is the sole work of a graduate student in Denmark, and is available as freeware (Tiny TeamTalk is shareware). My point here is that if one person can do this and see 2,000 downloads of it each month with absolutely no promotion, what else might be out there?

Slide 19 - In this dynamic of high profile internet brands and independent guerilla programming both doing interesting things with voice, the incumbents may find themselves in a lethal crossfire.

Slides 20 - 25 - I tried to touch on a few themes we're seeing developing in the VoIP world. The first was IP as incumbent outreach product (examples included Wanadoo and Deltathree in the UK market, and BT Broadband Voice seeking to claw back revenues from cablecos. This was followed by VoIP as incumbent pre-emptive strike (BTYahoo Communicator, Telia Sweden's Bredbandstelefoni products as examples). Then followed VoIP as ISP differentiator (Free.fr IP triple play for EUR29.99, and Freenet.de's iPhone product as a smart tactic in a concentrated DSL market), and VoIP as an affinity product (Digilinea - an ATA-based service in the US targeting the Hispanic market, my point being that VoIP might offer some interesting variations on the ethnic calling card model, and the global ATA rollout from Morpheus). I then discussed the Cablecom hispeed 100 product as a good example of a common sense way to employ bundled VoIP in a product to attack an incumbent's most lucrative product set, followed by an examination of the foothold which naked DSL availability has given to Telio in Norway, effectively removing the duplication of access payment issue which some hold up as an inhibitor to the adoption of broadband telephony.

Slides 26 - 28 - Here I highlighted three innovative companies with unique products: firstly Paradial, a Norwegian start-up which claims leadership in NAT/firewall traversal for SIP/MSN video services. This has been one area where Skype has gained a lot of attention with its solution, but a company like Paradial may deliver something equally elegant for the SIP world. A competitor company, Ridgeway of the UK, was bought last month for $16m by Tandberg. Next came Voipster, covered above, and lastly I presented a screen shot of Gush from 2entwine.com, which fuses IM, RSS, and in this experimental example voice and video, all in an attractive Flash-based format with some unique features.

Slide 29 - The point I was trying to make in summary was that some of the key innovations coming out of the sector seem to be coming from small companies, and in some cases, non-companies, utilizing skill sets found mainly outside the telco world. In this environment, it is increasingly difficult to determine where the next wave of change will come from and how telcos will position themselves quickly enough to benefit from, or at least absorb, the impact.