Staying in the game

Richard Stastny has picked up on a release out of the US regarding CC1 ENUM, an organization comprised of AT&T, domain registrar GoDaddy.com, MCI, SBC Laboratories, Sprint, and Verizon, which will promote ENUM within the NANP area (US, Canada, Caribbean). I very much like ENUM as a concept/ideal, and I am sympathetic to, and respectful of, the needs for standards in VoIP. I know it has been in the pipeline for some time now, but then again, in light of who is involved, it is interesting to see this happening just now, one week after the Popular Telephony GNUP and Dundi announcements (regarding alternative numbering/identity schemes) at VON. Perhaps it's coincidence, but it seems to me that the carriers are growing significantly more concerned about issues of control in the emerging IP space. Any guesses as to how long we will have to wait before Euro CC ENUM is announced?

Resource update - OFCOM market review update

OFCOM is on a tear, following up their comprehensive market review of August with an update today, covering the period up to the end of September (for some data). Carrier pre-select lines are up by 13.5% sequentially in the most recent quarter, now accounting for 15% of all BT lines. The report contains a current estimate from Analysys that VoIP is in use in 0.35% of residential access lines, and o.14% of business lines. The UK has 24m residential lines, and 10.6m business lines, so this seems to equate to 84k residential users, and 14k business lines. Viewed as a percentage of the addressable market of broadband lines, this is more like 2% penetration. However, the 98k number implied clearly excludes Skype. Based on information on geographical distribution of Skype users, I believe there are currently over 200k Skype users in the UK market. This implies that the true VoIP penetration of total broadband lines in the UK is more like 6%, before whatever uplift we miss from voice IM.

(http://www.ofcom.org.uk/research/industry_market_research/m_i_index/cm/qu_10_2004/cm_qu_10_2004.pdf)

Belgacom giveth, and Belgacom taketh away

Belgacom launched VDSL today in a number of cities - 9Mbps down/400kbps up, for EUR59.95. This looks way steep, considering that next door in France customers of Free can get 15Mbps for EUR29.99 (including TV and voice), so poor Belgacom subscribers pay 3.3x as much per plain vanilla megabit. The Belgacom package is also capped at 15GB per month. Anyone doing anything halfway serious in terms of bandwidth consumption is going to hit this pretty quickly. Strange to see such retrograde pricing in a market which is already more highly penetrated than France, with serious cable modem competition (something France still lacks at present), but this is probably a stop-gap measure in a transition technology while the FTTH project proceeds. It certainly doesn't look priced for success on a long-term basis. Coincidentally, EUR59.95 is a good rough benchmark for Europe generally in terms of current monthly spend for PSTN plus mid-level DSL/cable broadband package. Maybe it is actually setting a reference price for future fiber offerings ? (http://www.belgacom.be/company/en/jsp/static/291004_vdsl.jsp)

BT port-blocking: fact or fantasy?

In an interesting twist, Jeff Pulver reports that so far no FWD users on BT DSL have reported problems with port 5060. Curiouser and curiouser... (http://192.246.69.231/jeff/personal/archives/001276.html)

Morgan, Merrill and J.P. - meet Andy, Om and Martin

Yesterday afternoon I had my first conference call with a certain institutional investor who just happened to stumble across the blog, the third such contact I have made in this manner. Very early in the conversation, he made the following remark:

"I find I increasingly get a lot more out of reading blogs than sell-side
research. I am doing more and more primary research myself, and blogs form
a very important part of that process."

Readers of this blog will know that I strongly believe this to be yet another very important angle to the disruptive technology theme, and one I have explored previously (http://eurotelcoblog.blogspot.com/2004/07/special-excerpt-from-upcoming-daiwa.html - the hardcopy version contains an annex with more pretty damning stuff in it). What follows is not a charm offensive or mutual admiration society - I read something like 45 blogs with regularity and all of them are good in their way.

I select these three [Andy (http://andyabramson.blogs.com/voipwatch/), Om (www.gigaom.com), and Martin (www.telepocalypse.net)] because they, to my mind, demonstrate the highly individual qualities of blogs which collectively deliver what brokers' research typically lacks. All three have very sensitive BS meters, and are not afraid to court controversy. All three possess wide expertise and that rare quality of 360-degree, joined-up thinking, which allows them to consider the broader implications of what Company A is saying/doing, rather than the all-too-typical broker treatment: "Company A announced X. This is line with our expectations. STRONG BUY." The former quality is what good fund managers increasingly seek out, the latter is something they find oppressive (because their in-boxes are full of it) and irrelevant.

Individually, their styles, and the issues they focus on, are very different. Andy is very tightly focused on VoIP and its impacts, and never misses a story related to the theme, but is also adept at making connections and seeing implications which others might miss. Om writes about virtually everything, seemingly, and if I'm not mistaken, recently seems to be including more financial results coverage and investment strategy angles (e.g., if you want to play the VoIP revolution, forget the carriers, buy chipmakers - we like it!). Martin is a deep thinker, with some fascinating and challenging views from a high level strategy perspective. Taken as a cocktail (or tapas, if you prefer food metaphors), investors could certainly do worse than to make these three a daily supplement to their research diets.

Couple these three with all the other insightful blogs out there (I get a lot out of reading Get Real, PaidContent.org, DigitalMusicNews, Slyck and a whole lot of others) and the message is pretty clear to me: eventually, and probably sooner than later, someone is going to pull together all these diverse angles on telecom/internet/media/hardware/applications/chips, incorporate some hard financial and technical analysis, and build a cross-sector investment research platform incorporating realtime tools (I mean blogging, IM, video conferencing and collaboration) rather than .pdfs and spam.

There is a business model here, and whether it's the financial media who seize upon it (Reuters and Bloomberg have the infrastructure and a lot of data, but are trapped in a walled garden mentality and put their journalists in the same sector-coverage silos that the brokers do), or the brokers (I'm skeptical, because I think they tend to be dismissive of alternative points of view, risk-averse, organized in sector and region silos, and anyway are focused on trying to kill one another), or a newcomer (CNET or something that doesn't currently exist), I feel certain that it is going to happen.

Investment banks: you have competition, whether you know/believe it or not. It behaves differently from you, uses different tools, and takes no prisoners. More importantly, its mindshare is growing, and you ought to be scared. What happens when one day your client base wakes up and feels confident enough to say they don't want your research?

Honey, my ISP ate the company

Interesting observation in the Financial Times today, which makes the case that the France Telecom buyout of Wanadoo minorities actually constitutes a reverse takeover of the parent company by the IP "young Turks." It certainly looks that way from where we sit. Deutsche Telekom announced plans recently to follow suit with T-Online, and today Swisscom has announced the reintegration of its ISP, Bluewin, into the Fixnet division (http://www.swisscom.com/GHQ/content/Media/Medienmitteilungen/2004/20041029_01_Bluewin_Fixnet_Fusion.htm).

We have always felt (see EuroTelcorama No. 5) that many of the telcos would try to adapt in the best way they could (clumsily), but would face many challenges and ugly choices, and would inevitably end up as companies with significantly smaller top lines and dramatically smaller workforces. It's interesting to see some of the European telcos circling the IP wagons relatively early, admitting that their ISP teams actually hold the key to the future, not just the key to cashing in on IPO mania back in 1999/2000 (as FT and DT did fairly unsuccessfully).

The transformational pains are going to be considerable, and the pressure to act is mounting, in our view. Yesterday France Telecom owned up to the fact that VoIP is taking a visible bite out of its French voice traffic, but as DSL unbundlers grow more aggressive in Spain, Italy, the Netherlands and the UK, we will see similar messages from others within a few quarters. One only need look at Om Malik's assessment of the recent AT&T and MCI asset write-offs to see where the asset sides of Eurotelco balance sheets are headed in the next couple of years (http://www.gigaom.com/2004/10/mci_att_write.php).

Paradial roadtest

I've written about Paradial's Real Tunnel firewall traversal program for MSN Messenger a number of times, and I saw a demonstration at CeBIT earlier this year, but I didn't download and install it until this week. I must say that I'm very impressed. It punched right through the firewall first time with no trouble, I registered my details and I was away. I just got off a test call with Kevin at Paradial up in Oslo (www.paradial.com), and the sound quality was very impressive, particularly when we spoke over each other on occasion - something I have noticed that Skype doesn't always handle terribly well. The other thing that's good about this is that, as it involves the MSN Messenger client, a user can move seamlessly into a Hotmail account or launch NetMeeting, as well as all the more frivolous things that MSN allows (background sharing, game playing). So, using MSN and Paradial together, you've got the Skype-like firewall-hopping feature, plus a lot more that Skype doesn't do at the moment (forgot to mention video), and it can also interconnect with the likes of Free World Dialup and iptel.

Skype probably shouldn't worry - Paradial has a few 10's of thousands of users on its public SIP service (servers in the US, Singapore and Europe), of which the European users are the most active. Moreover, the company's real target is the corporate desktop conferencing market. However, it gives an idea of what some of the IM players might choose to add to the mix later, or indeed what Google might come up with if the reports about IM development are correct. I've speculated for months about Google in the voice space (http://eurotelcoblog.blogspot.com/2004/04/daiwa-eurotelcoblog-no_02.html). Here's a quote from Eric Schmidt in the conference call last week, in response to a question on whether or not Google has a portal strategy:

"So we look for opportunities, Gmail being the most recent large-scale example,
where we saw an opportunity to solve a problem, that had largely been viewed as
a mature industry in a new and different way... We are going to look on a per
case basis and you will see us evolve this model as we get better on
understanding worldwide user needs."

So, I used the 'p' word in my earlier post, but seriously, how about something that stitches together all the features of the services Google has acquired over the past year (Blogger, Picasa for image storing/sharing) with the new features it has developed (desktop search, Gmail), in an IM client interface, with voice/video as an added bonus? They've got the servers. I picture something like Gush (which I wrote about ages ago http://2entwine.com), but with all the other Google stuff intermingled. Vague, I know, but I'd be amazed if something like this doesn't happen.

I started out on Paradial, and ended up on Google, sorry. Download Real Tunnel and try it.

BT port blocking

West Coast insomniac and all-round nice guy Andy Abramson has already covered this, but for anyone who missed it, Mr. Blog claims that BT is blocking port 5060. Oh dear. If France Telecom says that VoIP effects are visible in their huge volume declines, and now BT apparently feels bothered enough to block SIP, then maybe the carriers are a tad bit more concerned than they have let on in the recent past. When I met up with the guys from Gossiptel a few weeks back they said they were at 5,000 subs or so, France Telecom said today that they've sold 4,000 Liveboxes in the UK, and no doubt there are a fair number of Free World Dialup users and ex-pat Vonage users out there. Excluding Skype and BT's own VoIP product, I wouldn't think we're talking about more than 50 - 75,000 users at the moment in the UK market, which is hardly worth this sort of reaction, if it indeed is confirmed.(http://www.toyz.org/mrblog/archives/00000173.html)

More Q3 data points from France

It's that time of year again, when we dial in to the France Telecom conference call and listen to Stan Getz on hold once again while hurrying to finish lunch. Ah, the glamor...

Today's Q3 results contained more information of interest to those tracking the issues we look at in this blog:

Call volumes in the consumer segment have deteriorated over the first three quarters, and registered a 10.3% YoY decline in Q3. What makes this so interesting is that FT lost fewer fixed channels (54k) this quarter than in the previous one (and 80k fewer than in Q3 2003), local market share stabilized, and long distance market share increased this quarter, yet volume declines accelerated in almost every category. We think the Orange data shows that this is being driven primarily by mobile substitution, but we also suspected (and management explicitly confirmed today) that some degree of VoIP impact is coming into view at the margin.

FT has so far sold 426k unlimited calling packages, which it claims increases ARPU for those users by EUR2 – 3 per month pre-VAT. While this is a positive defense strategy, we calculate that the decline in revenue per fixed channel accelerated to 5%YoY in the quarter, and penetration of such packages must ramp up significantly to gain gearing against the underlying traffic decline, in our view. Management today revealed that FT had 30,000 VoIP subs in France, and had sold or rented 40,000 Liveboxes in the market. (Across its four core European markets over 100,000 Liveboxes have been sold to date since the product launched in mid-July). This is FT taking a leaf from enfant terrible Free, which today reported another 140k subscribers gained in Q3, taking it to 908k, or nearly 20% of the entire French DSL market.

Wanadoo succeeded in getting its share of DSL net additions back above the 40% mark this quarter, and at 47% overall market share, is within management’s target range of 45 – 50%. However, as we recently saw, unbundled lines grew by 43.6% sequentially in Q3, and fully unbundled lines by 290%, which has to call the sustainability of the pricing and market share situation into question longer term. Stripping out the unbundled lines from France Telecom's stats for the whole market seems to show wholesale lines flat-to-down slightly, as market growth is largely absorbed by the unbundlers. Management stated today that they believe that unbundled lines will grow by nearly 50% in Q4, to 1.5m lines (we have been looking for 1.35m previously), and account for 25% of the total market by year end. That means that unbundled DSL lines will have gone from basically 0% market share to a quarter of the market in five quarters. Is France really "Old Europe"?

A different kind of oil shock

It's not often that I get to write about the oil industry as a telco disruptor, but today Shell, which is a gargantuan entity broken into two separate listed companies (one UK, representing 40% of the equity, one Dutch, representing 60%) has decided to merge the two under a single board and list everything in London. Why should we care? Well, all things being equal, this will take Shell's weighting in the FTSE 100 from 3.8% to nearer 8%, which means that the weightings of other companies should fall by about 4.4% from where they are now.

Biggest telco victim - Vodafone, which we worked out should move from 8.21% of the index, to 7.86%. BT Group should go from 1.4% to 1.34%. These don't look like big drops on the face of it, but when the fund managers run this through their sector weighting models, coupled with whatever their own sector preferences might be, we could see some technical selling around this. The company is going to maintain a Dutch listing as far as we are aware, though whether it will still account for 12% of the market's capitalization in future is unclear. So perversely, KPN, one of the most embattled telcos in Europe, may go up on technical grounds.

Resource update - Scathing review of Orange 3G data card

We're far from being 3G bulls here at Daiwa, and much of the technology's prospects probably lie in simple things like ease of use and the intuitiveness of the user interface. This review brings up some interesting examples of where Orange seems to have gotten it very wrong, in contrast to Vodafone. As a long-standing and increasingly dismayed Orange customer, I must say I'm not totally surprised. (http://www.the3gportal.com/cgi-bin/framer/framer.cgi?http://www.trustedreviews.com/article.aspx?art=803)

Some Q3 datapoints

Had to pay the mortgage today by hunkering down and writing some maintenance research. My first two sets of company results doubled up on me this morning - Telenor and TeliaSonera. As usual, lots of detail to wade through, but a couple of points came out which are of interest given the orientation of this blog.

Firstly, Telenor revealed that 7% of its DSL line base is "naked DSL." In Q1 the figure was stated as 6%, and no clear figure was given in Q2, but management today said it was 5%. Personally, I have trouble imagining this figure going down in Q2, so I think someone was improvising, but that's something I can take up with the company later. The really significant thing is that, at 7% of its retail DSL lines, this translates to roughly 19,000 lines, on top of which there are, on my estimates, 60,000 fully unbundled lines in the market. In other words, I think it's possible that something like 17% of Norwegian DSL subs may have no PSTN subscription. Mobile substitution is the main culprit so far, but we also know that Telio is thriving in Norway, and this sort of scenario in the DSL market is something that the rest of Europe should probably watch with trepidation. All that said, it looked to me like Telenor is doing a pretty good job of defending itself (line loss was down again on Q2 and half the level of Q1, wholesale line rental growth has stalled, and Telenor's retail DSL lines are adding more users than wholesale/ULL, and picking up speed).

Secondly, TeliaSonera blew the market away at the EBITDA level (I was well above consensus but still 6% too low), and the share price rallied - which is a shame because I have an UNDERPERFORM rating. Try as I may, I can't get to the sort of valuation the stock is trading at now, but I won't begrudge them a bit of good news. Management almost sounded surprised to see a 1% YoY increase in fixed line traffic revenues in the Swedish market. TeliaSonera is still losing about 3% of its access lines on an annualized basis, so how did they swing this? Dig a bit deeper, and we see that total fixed line traffic declined by 1.5% YoY, which is very mild (Q2 was down 3.8%, Q1 a whopping 6.1%). Peel back below that level, and we see that national fixed-to-fixed traffic was down 2.6% YoY (again much milder than the 4.9% in Q2 and the 7.4% in Q1), international was flat, but fixed-to-mobile traffic is up 11.2%, and accelerating. In fact since Q2, outgoing minutes to mobile have increased dramatically and now account for 11.2% of outgoing fixed line minutes on TeliaSonera's network. It looks as though, while the fixed line user base is dwindling, voice usage among the remaining users appears to be stabilizing somewhat, ironically driven by the ubiquity of mobile usage. This is the short-term positive flipside of mobile substitution - higher revenue-yielding fixed-to-mobile minutes growing and increasing their share of the pie, allowing the telcos to eke out some top-line growth in traditional traffic revenues. The interesting thing to see will be, at what point does this apparently more stable legacy user base become more susceptible to cutting the cord completely, and will the arrival of wholesale line rental in the market accelerate the process?

France DSL market - 1m unbundled and counting

France may be part of "Old Europe" but the regulator was in the office on Sunday and issued a DSL market update for Q3, showing over 1m lines unbundled (998k partially, 51k fully). This represents sequential growth (that's from June 30 to September 30) of 43.65%. Jaw-dropping. (http://www.art-telecom.fr/eng/index.htm)

When the going gets tough, the tough move the goalposts

Tele2, who suffered some brutal treatment at the hands of the market after last week's Q2 results, have announced a reshuffle of reporting lines, moving the UK and Irish operations from the Southern Europe division into a newly defined UK & Benelux region (http://www.waymaker.se/bitonline/2004/10/26/20041026BIT20650/10262065.htm). My six year-old's world map (and the miserable climate here) have always indicated that the British Isles are not in Southern Europe, but Tele2 has previously argued that the classification was more to do with market dynamics than actual physical location (on this basis perhaps markets should be delineated by average customer height or national costume). Interestingly, Southern Europe was the most disappointing element of last week's results. However, in light of the fact that Tele2 has never issued any meaningful information on the UK market, or any other specific country market, and also just made a radical reshuffle of reporting lines in Q1 of this year, unravelling the underlying performance of the business is guesswork.

IPWireless/Atmel tie-up - it's a small world

Last week, while all eyes were on the Geek Woodstock (or Altamont, depending on your point of view) known as VON, IPWireless announced a partnership with US chipmaker Atmel (http://www.ipwireless.com/news/press_101904a.html) to produce a UMTS-TDD VoIP (how's that for acronyms?) handset. We knew handsets were definitely in the pipeline for IPWireless (our understanding of the roadmap was Q1 2005), and it's good to see this coming through apparently right on time. What makes this release particularly interesting is that there is an unnamed OEM mentioned in the press release, and one of Atmel's OEMs is none other than Atonics of Taiwan (http://www.atonics.com.tw/company.htm), with whom Peerio signed a development agreement last week. Given Peerio's recent successes in striking agreements with a variety of players in the telecom food chain, is it a stretch to think that we might see the "C'est Peerio" logo on some very robust mobile hardware in the not-to-distant future?

This is just the sort of development that should have mobile bulls reviewing their revenue assumptions. To date, much of the market seems to have dismissed VoIP as a primarily fixed-line phenomenon, or at best, one that was trapped in the wireless ghetto of the Wi-Fi hotspot. As my mobile/Southern Europe colleague Jacqueline Millan noted in her recent initiation note on mmO2, despite apparently rampant competition in the UK and German markets, there is really very, very little differentiation in pricing between the mobile players in the consumer space at present. And make no mistake about it, the players with 3G licenses need to carefully manage voice pricing as subscribers ramp up on the new platforms.

How much market share for $5m a month?

It's interesting to see the likes of Telio saying the company will be profitable at the net income level this year, and what seems to distinguish it is the complete lack of marketing spend to date. This is probably partially related to VoIP market dynamics in Europe, but also probably to management aspirations. Contrast with Vonage, which yesterday awarded French advertising group Havas its marketing campaign reportedly worth between $50 and 75m (http://www.adage.com/news.cms?newsId=41824, probably requires registration, but the site is free). At the mid-point of the range, this suggests a monthly advertising spend of $5.2m.

Admittedly, the markets are different, but just to put this in perspective, BSkyB, which is in a turnaround phase in its evolution after a couple of quarters of poor net additions, has earmarked an additional GBP50m ($91m) in above-the-line annual marketing spend this year. Remember, this is a $17bn company with 7.5m customers, and an entrenched market position to defend. The longer I follow this sector, the more I'm convinced that for investors, the way to play it is actually through ancillary sectors which might benefit from the competitive mayhem underway. As an example, take global advertiser WPP (which is reporting Q3 results today). In the first half, telecoms was the second-highest growth revenue category for WPP, at 15 - 20% annual growth. How many telcos are reporting those sorts of numbers?

SkypeOut price reductions ahead

One of the criticisms of SkypeOut voiced in some quarters has been that pricing hasn't been all that cheap (I would disagree, as would probably many people living in draconian telecom regimes around the world). This Thursday, all that is set to change, with some dramatically lower pricing to some destinations(http://www.skype.com/company/news/2004/sa_skypeoutratesoctober.html). I assume this reflects Skype's ability to renegotiate some termination agreements based on higher call volumes coming through. Calls to Japan for under two eurocents per minutes looks pretty competitive to me. Calls to French mobiles for 16.4 eurocents also look pretty good, when you consider that BT (remember it is possible to swim from the UK to France) customers who pay GBP1 per month extra receive a special "discounted" rate of 30 pence (that's 43 eurocents) for the same call. This is what makes disruptive technologies so interesting - they expose just how much further prices need to collapse before consumers start getting real value for money.

Paradial jumps the firewall into 3G

I have been remiss in my blogging duties. A tiny little Norwegian company called Paradial (www.paradial.com), which I have written about a couple of times previously in relation to possible consolidation candidates in the race for the ultimate desktop conferencing solution, demonstrated a video call (last Wednesday at VON) from a SIP client behind a firewall to a Hutchison 3G Italia terminal. Sadly, they haven't posted the press release to the website yet, so I can't link to it. I'm probably stating the obvious here, but this is just the sort of commonsense application which could really broaden the proposition for video-over-3G as a practical technology - i.e., it's not just an expensive video walkie-talkie that is useless unless one of your friends with a 3G terminal calls you to say, "Hey, isn't this cool." This is a genuinely useful application of both SIP and 3G which can enrich both today.

Daiwa EuroTelcoblog No. 83: Monday, 25th October, 2004 - Moshi-moshi Japan, Skype calling

Skype was obviously holding back some bullets until after the VON event last week. Yesterday, we had the TOM.com announcement in China, and now today, Skype has just announced a partnership with Livedoor (TSE 4753) to deliver a co-branded Skype to the Japanese market. As we pointed out earlier this morning, Japan appears underpenetrated in terms of Skype users versus some other markets. If this partnership delivers similar results to the Taiwan partnership, Skype user numbers should accelerate significantly. Of particular interest tomorrow will be what sort of share price reaction Livedoor sees to the news. The stock has underperformed TOPIX by 22% year-to-date, so the "Skype magic," if any, should be evident on the public markets for the first time. Press release is here: (http://www.skype.com/company/news/2004/livedoorlaunch.html).

Resource update - DTT in Italy

Those who have observed the runaway success/deflationary bloodbath (depending on your perspective) which is the Freeview platform in the UK market, may be interested in this overview of the DTT situation in Italy, from Mediaset's perspective. The slides show that the past six months have seen 720,000 set-top boxes shifted, and that the Italian market is just behind the UK in terms of DTT multiplexes. (http://www.gruppomediaset.it/corporatefiles/DTT_updateX.pdf)

Broker sanity check - would you buy a used car from this man?

A couple of years back, I used to publish a weekly update on the PTTs, called "Weekly PTT Pulse," and the one thing I really liked about it was that every week I tracked the mismatch between analyst sentiment and sector performance, in something I affectionately named the "analyst whipsaw index." Sadly I have let the data points lapse, because I have more important things to do, but I thought it might be interesting, particularly for those without access to Bloomberg data, to examine this issue.

Firstly, a few macro points:

• Of the 18 DJ STOXX industry groups in the STOXX 600 index, the benchmark we follow, Telecom has performed best over one month (+6.8% absolute, +3.7% relative) and second-best over three months (+2.6% absolute, +5.3% relative). It is important to point out that over that period we have seen some pretty savage underperformance from the Food, Media, Health, and Auto sectors.
• Telecom is the third worst performing sector (after Tech and Insurance) over five years, down 27.6% relative. A five-year long position established in October '99 in Basic Materials, Construction, Banks, Energy, Food or even Chemicals (!) would have generated excess returns over an equivalent investment in Telecom by a margin of between 39% and 72%.

Virtually everything I have written over the past two years has focused on an impending shift in the nature and economics of telecom service provision, and the immense challenges telcos will face in keeping an even keel through the transformation. Almost every client meeting I have had over the past two years has revealed that investors largely share my concerns. I would assume any short-term outperformance in the sector is being driven to some small extent by company de-leveraging, increased dividends and share buybacks, but mainly by technical considerations and the poor fundamentals of other sectors. In other words, beware of sustained long positions in the sector into the new year, because I expect January could be the start of another trough.

Then again, I am only one voice, and a relatively minor one at that. I've never appeared on Kudlow & Cramer, and I don't even like Ferraris. What do the other brokers say? I took some time this morning and tabulated Bloomberg broker ratings on 18 stocks in Europe, comprising 14 integrated operators and four mobile pure-plays. All-in-all there are 510 current broker ratings on these 18 names, equating to an average of 30 analysts per stock (that's a lot of maintenance research, and a lot of deforestation). Of the 510 ratings, on average 82% were either positive or neutral. The least-loved stock, TeliaSonera (48% negative ratings) has performed second-best over the past three months (I have an UNDERPERFORM rating too, because I don't buy the buyback story), following Telenor (94.7% positive rating and my preferred name in the sector).

The data is here, but Blogger doesn't handle tabular data well AT ALL, so apologies for the layout:

Belgacom Buys – 6,Holds – 9,Sells – 2,Total – 17
BT Group Buys – 5,Holds –14,Sells – 11,Total – 30
Deutsche Telekom Buys – 33,Holds –9,Sells – 2,Total – 44
France Telecom Buys – 29,Holds –9,Sells – 4,Total – 42
KPN Buys – 19,Holds –14,Sells – 11,Total – 44
mmO2 Buys – 9,Holds –11,Sells – 7,Total – 27
OTE Buys – 9,Holds – 8,Sells – 3,Total – 20
Portugal Telecom Buys – 15, Holds – 9, Sells – 4,Total – 28
Swisscom Buys – 5,Holds – 15,Sells – 5,Total – 25
TDC Buys – 8,Holds - 7, Sells – 6,Total – 21
Telecom Italia Buys – 21,Holds –12,Sells – 2,Total – 35
Telecom Italia Mobile Buys – 20,Holds –8,Sells – 7,Total – 35
Telefonica Buys – 28,Holds –13,Sells – 3,Total – 44
Telefonica Moviles Buys – 10,Holds –17,Sells – 11,Total – 38
Telekom Austria Buys – 6,Holds –9,Sells – 3,Total – 18
Telenor Buys – 13,Holds – 5,Sells – 1,Total – 19
TeliaSonera Buys – 7,Holds – 5,Sells – 11,Total – 23
Vodafone Buys – 34,Holds – 7,Sells – 2,Total – 43

Total Buys – 243,Holds –174,Sells- 93, Total – 510, Neutral/Positive - 81.8%

Beyond the apparent reluctance to be negative, it is also interesting to see some of the price targets out there. Take a company like BT Group (Daiwa rating UNDERPERFORM, price target 179p), where sentiment is relatively bearish: the average of disclosed price targets according to Bloomberg is 192p, or still 5.4% above today's open. At the other end of the spectrum, the three most bullishly-rated large caps (Vodafone, Deutsche Telekom, France Telecom) have average price targets giving upside of 12.9%, 17.8% and 19.4% upside, respectively. After such a strong run over three months, and with a wide range of secular challenges and uncertainties facing the sector and the individual operators, what in the name of Juan Villalonga do the analysts think is going to happen to support this sort of rise?

Resource update - Research and Development league table

The UK's Department of Trade and Industry (DTI) has produced a global league table of R&D expenditure for 2004 (http://www.innovation.gov.uk/projects/rd_scoreboard/introfr.html). In our dear old telecom sector, it's interesting to see companies like BT Group and TeliaSonera outgunning the likes of Vodafone (!) and AT&T. Tiny Belgacom spends as much per employee as AT&T, and TeliaSonera three times as much. Not a single RBOC appears on the list. It's also interesting, though hardly surprising, to see a company like Yahoo! outspending the average telco on a per-employee basis by a 7-to-1 margin. Even UK online bank Egg spends nearly three times as much as the average telco on this basis. A couple of very sharp bloggers have recently observed that "the IT industry has declared war on telecom," a view with which I wholly concur. If the development pipeline is genuinely what is going to differentiate the two camps, then data like this paints the telcos in an evermore dismal light.

Liberty Media rolling up more Euro cable

Given the aggression with which UnitedGlobalCom (owned by Liberty) European unit UPC is pushing VoIP, it is interesting to see Liberty take control of Callahan's 14% equity stake (21% voting rights) in Telenet, Belgium's leading cable player and one of Europe's most technically advanced MSOs (http://www.telenet.be/nl/bezoekers.php). Coming on the heels of the Noos acquisition in France, Liberty is further consolidating its position as the major pan-European cable player, and it will be interesting to see what the next technical/product developments are for Telenet, as well as what strategy/product cross-pollenation occur inside UPC as a result.

It is my understanding that the Belgian cablecos have thus far been leaders among a small group of European MSOs to indicate an interest in the Ultrawideband-over-coax technology on offer from Pulse_LINK (offering 1.2Gbps capacity boost per node on 750MHz systems, and 600Mbps on 450MHz systems, with enhanced upstream speeds which make the E1/E3 markets addressable for cablecos, at an upgrade cost of less than $100 per sub http://www.pulse-link.net/pr-aug10-2004.html). So far most interest has come from Japanese and Korean MSOs, where the threat from fiber is much more pronounced and immediate. So it is with Belgium as well (incumbent Belgacom begins IP TV trials next month, is rolling VDSL, and check out this list of blown fiber contracts from Emtelle http://www.emtelle.com/?id=308).

Next door in the Netherlands, KPN CEO Ad Scheepbouwer has recently been making noises about FTTH to Bloomberg journalists, and UPC Netherlands has announced trials of 30Mbps and 50Mbps access technologies with unnamed vendors (my understanding is that there are two involved, but I have been unable to find out who they are). As commentators in the US decry the state of competition following the FCC's RBOC fiber ruling, my guess is that cable in Europe does not roll over and die in the face of fiber competition from the PTTs, and in some markets may well beat the telcos to the home with HDTV offerings and higher access speeds plus VoIP.

Daiwa EuroTelcoblog No. 82: Monday, 25th October, 2004 - Skype and TOM.com partner in China

Chinese portal TOM.com yesterday announced that it will be launching a co-branded version of Skype imminently (Chinese - http://tech.tom.com/1121/1794/20041024-133551.html, English http://ir.tom.com/en/news_20041024.html). As we noted recently, Skype's only other co-branded venture, with PCHome in Taiwan, seems to have been instrumental in taking Taiwan to the No. 2 spot globally in terms of Skype users (9.2% of total). China's broadband user base is over five times greater than that of Taiwan (c.4m), so a similar level of uptake there would move China decisively to the top of the Skype user league table.

It is interesting to now see a confirmed Asian focus in Skype's choice of ISP/portal partners, and even more interesting to note that after the US and China, the two largest broadband users bases are in Japan (12.8m subs), which ranks only 12th in terms of users, and South Korea (11.7m subs), which doesn't even appear in the top 20 (reproduced again below). We now fully expect those markets to be included in this co-branding push, and wonder what the scope for an Indian partnership might be over time. We may be about to witness a step-change in Skype adoption and usage.

Top 20 Skype user countries

[1] United States - 10.33%
[2] Taiwan - 9.24%
[3] Poland - 8.78%
[4] Brazil - 7.24%
[5] Germany - 6.18%
[6] China - 5.89%
[7] France - 5.53%
[8] Netherlands - 3.50%
[9] Denmark - 3.07%
[10] United Kingdom - 2.94%
[11] Israel - 2.94%
[12] Japan - 2.61%
[13] Canada - 2.46%
[14] Belgium - 2.10%
[15] Spain - 1.82%
[16] Sweden - 1.76%
[17] Australia - 1.46%
[18] Italy - 1.44%
[19] Switzerland - 1.22%
[20] Mexico - 1.07%
Source: Skype

Cracks in the wall of indifference

Yesterday afternoon I got a call from a salesman in one of our European branches to ask me an urgent question:

"James, have you ever heard of a company called Skype?"

I took a deep breath, counted to five, and tried to diplomatically explain that, not only had I heard of Skype, but in fact 13 months ago I was the first among the broking analysts anywhere to write about it, one week after it launched. What's more, I've written several thousand more words since then on Skype and a lot of other related issues and companies, and have given countless presentations to institutional investors which left many of them profoundly unsettled. "So much for competitive advantage in research," I thought. I also thought about the late (but immortal) Rodney Daingerfield's stock phrase: "I don't get no respect."

Later I paused for reflection and realized that this incident, though galling to me personally, is probably significant, and here's why:

The person in question is a Luddite, overworked equity salesman trying to cover the entire Japanese equity market, and dealing with information flow in two languages which are not his native tongue. The fact that he has remained ignorant of Skype until now, despite my best efforts, is not all that surprising, given the nature of his workday: endless order executions, client calls, lunches, company roadshows, IPOs, etc. I understand that. What is important is that, even a year later, this relatively insulated Luddite is now aware of Skype, and moreover, is scared about what it means for a whole raft of companies. I took the opportunity to bring him up to speed on Popular Telephony yesterday, and now he is really scared.

Obviously, I won't get too excited about it until I have a few more examples under my belt. However, if people, who have up to now somehow managed to miss the millions of column inches of coverage Skype has received in the press over the past year, are now independently waking up to the implications, and to a sneaking suspicion that even more disruptive things may be just around the corner, we might be about to see one of those hockey stick adoption curves that brokers love to put in research.

I, robotelco

Alright, this is a bit off-topic, but in light of all the hoopla in recent weeks about the digital living room, the networked home, the converged kitchen, etc., I thought these stats on the growth of robotics from the United Nations Economic Commission for Europe were interesting (http://www.unece.org/press/pr2004/04stat_p01e.pdf). The idea of there being 610,000 domestic vacuum cleaning and lawn-mowing robots worldwide is absolutely staggering to me, as is the forecast of another 4m units in operation by 2007. (My only partially tongue-in-cheek recommendation to the investment banking world is that robots may be good candidates for writing maintenance research in future, as my conversations with investors lead me to strongly believe that no one reads it in any event. See http://eurotelcoblog.blogspot.com/2004/07/special-excerpt-from-upcoming-daiwa.html for background).

Seriously though, it's interesting to think about how companies involved in both telecoms and domestic services (like Centrica in the UK) might try to build on this phenomenon. BT Group already has a home security monitoring service in operation which involves both wireless and the internet (http://www.bt.com/homemonitoring/index.jsp) - is it so much of a stretch to believe that telcos might move to offering wirelessly-networked domestic assistants as an accessory to home networking product offers as a differentiator? A couple of years out, when residential broadband access levels start to peak, voice traffic revenue evaporates (as we know it will), and the telcos have struck all the low-margin content partnerships they can, we may see all sorts of bolt-on products/services that we never viewed as telecom before.

If it's Thursday, there must be another Peerio partnership

Popular Telephony wins this week's newsflow award hands down, and it is quite daunting to think that the company has achieved so much on the basis of angel money to date, while other, better-known companies have been landing additional VC money in ever-larger quantities. However, it seems that the market is waking up, and the company is at last getting coverage in some higher profile venues (http://www.business2.com/b2/web/articles/0,17863,725862,00.html). It's interesting to step back and remember that Skype was more or less a whisper this time last year, and now returns 913,000 hits as a search query on Google. Peerio returns 906, but that should change.

Today Popular Telephony has announced a deal with Atonics Inc. of Taiwan, to license Peerio for use in Atonics IP phone products, including Wi-Fi handsets (http://www.atonics.com.tw/product.files/WiFi_preliminary_Sep_2004.pdf). This is interesting in light of last week's announcement with Global IP Sound and my speculation that this might relate to a push in the wireless arena. It is also interesting to see the company moving into the Asian market, where the stakes are probably higher in the long run than in Europe (remember Skype's only co-branded ISP deal to date has also been in Taiwan, and this seems to have contributed to making the Taiwanese the second largest group of Skype users in the world).

To date, the market, and particularly the brokers, have tended to downplay the potential collateral damage to cellular operators from VoWi-Fi, and this is another data point along the way to suggest that the issue is not to be ignored. (Anecdotally, I have been interested to see the number of people wandering into my site initially looking for information on Wi-Fi handsets from producers like Senao, also a Taiwanese company. At this point I would have to say that search queries related to Wi-Fi handsets are probably the third most common ones I see coming in.)

Sweden - more wireless woes on the way

In a recent post on the Flarion/Siemens tie-up to take advantage of 450MHz frequency refarming (http://eurotelcoblog.blogspot.com/2004/10/flarion-and-siemens-hotwire-450mhz.html), we noted that Sweden was expected to reach a decision on licensing by the end of the year. The consultation has been published (http://www.pts.se/Nyheter/nyhet.asp?ItemId=3802), and regulator PTS has gone for an auction of 2x1.8 MHz in the frequency range 463.0-464.8/453.0- 454.8 MHz. This range is to be vacated by existing users by January of next year, and the license process envisages that 80% of the land area of each county in Sweden will be covered by 1st July 2007. This is actually a much less draconian build-out requirement than that faced by the UMTS players in the country. Given how tough the Swedish mobile market is already with only three 2G players and a handful of MVNOs, the arrival of a new player, perhaps with a superior data product, should do nothing to help sentiment towards operators with big financial stakes in the market (we have TeliaSonera as an UNDERPERFORM, and no rating on Tele2).

Fixed line, schmixed line...

Nokia and MORI have just released the results of research into mobile/fixed substitution (http://press.nokia.com/PR/200410/965166_5.html), which claims to have found that in the markets of the UK, USA, Germany and South Korea, something like 45m consumers exclusively use their mobile phone for voice. I work that out to be about 9.3% of their combined populations, overall, which is pretty similar to the kinds of numbers we have anecdotally heard in other European countries. A few interesting insights are included:

• In South Korea, a significant number of people report that they are less likely to take out a new fixed line subscription after moving house/flat, and 65% of the respondents there report they are already almost entirely dependent on mobile for voice. This house-moving phenomenon is something that France Telecom mentioned in Q3 last year as a trigger for non-renewal of PSTN subscriptions in its domestic market.
• For respondents with broadband access already, there is keen interest in switching to a wireless broadband service, and an expectation that this may be a viable alternative within one to two years.

Telio goes white label

Telio has announced a white label partnership for the Netherlands with InterNLnet (www.internlnet.nl), to go live in Q1 2005. The press release (http://www.telio.no/about/?436)contains no mention of pricing, but I assume it will be highly attractive. In Telio's home market of Norway, unlimited fixed line calling to all of Western Europe and North America costs EUR19.50 per month. This deal appears to initially cover only calling within the Dutch market, but it will be interesting to see if the Europe/North America offer is extended to InterNLnet's subs over time. I am also curious to see if this is followed by similar deals in other markets.

I have until recently been fairly unconvinced about the value of white label/co-branded offers of this type, but Skype's experience in Taiwan has given me a new perspective. Recall that when Skype revealed its geographic user breakdown recently (http://eurotelcoblog.blogspot.com/2004/10/daiwa-eurotelcoblog-no.html), Taiwan, where it has its only co-branded deal with PChome, came in at number two, despite being a smaller market overall than many others which appear further down the list.

Remember, AOL has two heads

Much speculation and rumor surrounds AOL's mooted VoIP foray, and while I would generally agree with some of the downbeat assessments coming out of the US market (http://blog.tmcnet.com/blog/tom-keating/voip/voip-blog/aol-and-dialup-voip-update.asp), I would also say that AOL Europe is a very different beast, and I see Europe as a market where an AOL VoIP offering might have much sharper teeth.

• AOL Europe accounts for about 20% (c. 6.3m) of AOL's total customer base, and unlike the US business, has seen more stable customer numbers (and even some growth) over the past three years (see page 6 of http://ir.timewarner.com/downloads/trending_07_28_04.pdf).
• The last official data I can find for the UK, for example (August 2003), shows AOL's overall internet market share stable at 18%, which made it at that time the country's third largest, just behind Wanadoo. Sadly, new super-mega regulator OFCOM doesn't seem to track this stat any longer (not that I can find), but generally speaking AOL's broadband pricing is much more in line with competitors in Europe (http://www.aol.co.uk/ and check out France where 1Mbps AOL DSL costs the same as 60-hours of dial-up http://sabonner.aol.fr/accueil/accueil.html?promocode=524187), and I have every reason to believe the company is holding its own on this side of the pond on the access side.
• Arguably, in markets like France, where DSL/VoIP bundling is gaining significant traction, AOL will find that having a similar product is a defensive, rather than market share expansion, tool. In other words, they probably have to do it just to stay in the game.
• Lastly, Nielsen//Netratings data for Europe from April shows over 10m AIM users, and 4.7m ICQ users (ICQ is still hugely popular in Germany), so it's reasonable to make a case that AOL Europe has some pretty attractive cross-marketing opportunities through these channels.

Telepocalypse

Not only is this my favorite blogname of all, but I think this piece may be one of the better I have ever read on the subject, and has the added value of coming direct from the heart of the Geek Woodstock which is the VON show in Boston. Read it, and weep, as they say. I am meeting up with author Martin Geddes on Friday and looking forward to it immensely. (http://www.telepocalypse.net/archives/000485.html)

Skype chips in to the numbers game

Amazing to see so many stats coming in a couple of days. Skype has just reported that concurrent users on the network broke the 1m mark today. I had a feeling we might hit it today, because late yesterday afternoon I could see over 980k users online. This is up from around 800k observed just a couple of weeks back.

The numbers are starting to add up

Xten Networks has just reported that it has 500,000 softphones deployed (www.xten.com/news/20041020.pdf) globally. Add that to the 1.8m number reported by Voiceglo today (which as far as I know involves no double-counting with the Xten number), plus the 5 - 6m that we think represents the total Skype base, and we're talking about pretty significant softphone penetration worldwide. What would be even more helpful would be if any of the four competing IM platforms offering voice (AIM, MSN, Yahoo!, iChat) would deign to tell the world what sort of voice usage they are seeing. Then we could add that data to the list to get some sort of rough estimate of access-independent, device independent VoIP users. My guess is that it's a pretty respectable number.

Q3 reporting season off to a lousy start

As Yogi Berra said, "It's like deja vu all over again." The Q3 European telecoms reporting season kicked off today just like in Q2, with some disconcerting numbers from Tele2 and a hefty share price fall (11.2% as I write). This time around, Tele2 has delivered net subscriber intake right at the bottom of the consensus range (1.03m), and seen negative fixed line/internet subscriber additions in Southern Europe and Benelux. Southern Europe (-10k net adds) curiously includes the UK, where the company has yet to disclose any subscriber numbers, but part of the weakness could stem from the CPS market there. As a testament to how ugly the competitive pressures are in this Southern region, net subscriber decline of 10k translates to a 3.7% compression in EBITDA margin versus Q3 2003. The release also makes clear that the source of weakness in the Benelux was the Netherlands (no surprises there).

The group revenue line was weak too, which is disturbing for a company which has traditionally delivered high growth rates relative to the incumbents. In its current reporting structure, Tele2 has delivered sequential revenue growth of at least 2% every quarter for the past five quarters - before today. Q3 is typically a seasonally strong one, but this time the top line is unchanged from Q2's result, and I calculate a 4% sequential ARPU decline this quarter. Last year's Q3 sequential ARPU decline was 5%, but the company also added 60% more customers in that quarter than this year.

Sweden, the critical home market, also proved problematic once again. Mobile EBITDA margin in Sweden contracted by 6.5 percentage points compared to Q3 last year, the single largest YoY compression in its history. Even so, the poor performance in many other operating regions means that Tele2's reliance upon its domestic mobile business is actually increasing again (46.6% of total EBITDA this quarter, up from 43.6% in Q2) - this is a significant reversal of trend, and one which the market would not like to see sustained.

As the first company to report in Q2, Tele2 proved to be something of a bellwether for the turgid Northern European PTT results that followed (with the exception of Telenor). I have a bad feeling that we may see a repeat of this scenario in Q3. Given its scope and scale, Tele2 could be considered the best proxy for the sector overall in Europe, and the evidence we see today - of rising churn, poor customer growth, revenue stagnation, and margin compression, paints a pretty ugly picture for what may be in store.

IP troika?

Sistema, the Russian telco behemoth intending to seek a London stock market listing (http://www.sistema.ru/_en/press/news/200410062114-2020.htm), today followed through on its September announcement of a DSL triple play contract with Alcatel, by adding Tandberg TV (www.tandbergtv.com) to the mix as provider of encoding, decoding and streaming solutions. Tandberg TV is one of a growing number of smaller European companies which are carving out a nice little niche in the TV-over-DSL arena. Another one which just hit the public market is Orca Interactive of the UK (www.orcainteractive.com).

Free rolls with ADSL2+

Iliad SA, parent company of French enfant terrible ISP Free.fr, today announced the launch of ADSL2+ services (http://www.iliad.fr/actualites/CP20041020Eng1.pdf). Like Telio, this is another company which is moving forward at a blinding pace. Just a couple of months back they announced that customers choosing the fully unbundled DSL product would get an additional 4Mbps of bandwidth for the same price as always, EUR29.99. Today this rises to 15Mbps (11Mbps when the video service is active) at the same price point, plus free national fixed line calling and 110 TV channels (42 pay, 68 free). I think that this time next year, the UK market may look very much the same, and not many will shed a tear.

Voiceglo weighs in

We should see Voiceglo enter the VON press release melee later today and announce some customer numbers that may genuinely surprise people (pick a number between one and two million), and a geographic spread that may make it the global number two after Skype. It would be really interesting to see a breakdown by country or region, as we got from Skype recently, and also to know what the conversion rate is from free to paying customers on the glophone. The latter will be a key data point for the industry (whether it comes from Skype or Voiceglo probably doesn't matter all that much).

Peerio - another day, another partner

W.C. Fields once famously said that, in future, wars should be fought personally by national leaders using socks filled with manure. Being VON week in Boston, the IP communication industry's weapon of choice is the press release, and the salvos are flying, but the contents are considerably more substantial: two interesting releases from Nimcat in as many days, a blockbuster release from Peerio yesterday on global numbering (message - "you will be assimilated"), and now another piece in the Peerio strategy revealed.

Today Popular Telephony announced that it has signed up Stealth Communications (www.stealth.net) as a partner. Stealth, a major global peering agent, launched Voice Peering Fabric back in April of this year, and at that time had on board Free World Dialup, Net2Phone, Packet8, Addaline, Acropolis, MIT and Yale University as part of the VPF peering agreement. The list has since grown significantly, and now includes minnows like Musimi (now part of Telio) and a giant in the form of China Telecom (http://www.thevpf.com/index.php?action=display_participants).

While the VPF infrastructure acts as agent for IP voice traffic peering, the business also has what it claims is the world's first commercially available ENUM database. Today's agreement will in theory allow Peerio/VPF to cross-map ENUM and Peerio's GNUP numbering plan, opening the door to the kind of seamless interoperability between platforms which underpins the GNUP plan. Yesterday's GNUP announcement was the vision, and this appears to be the first stage towards making it reality. Given the widespread range of ENUM initiatives underway, things could get really interesting (http://www.centr.org/kim/enum/index.html#43), and it's only Tuesday!

Skype - billions and billions served

A few weeks back I posted on Skype's website redesign and noted that they seem to be making a tongue-in-cheek reference to fast food culture with the "minutes served" counter (http://eurotelcoblog.blogspot.com/2004/09/daiwa-eurotelcoblog-no_109519552619085606.html). That was 14th September, 35 days ago, and the ticker stood at 1.5bn minutes. Today it has just rolled over 2bn minutes a matter of minutes ago. That suggests something like 14m minutes per day (this doesn't include SkypeOut minutes).

To put this in perspective, in the most recent quarter for which we have regulator data, the UK fixed telephony market generated 1.65bn minutes of outgoing international traffic, or around 18m minutes per day. However, that traffic was spread over 34m active PSTN lines, rather than the more limited number of Skype users (I presently see 969,000 users online). Skype usage is undoubtedly much more intensive, rather than mere voice minute replacement, and appears to be accelerating. It took a year to reach 1.5bn minutes, and only five weeks to add another 500m minutes. How can any self-respecting telco look at these stats and say that the VoIP story is blown out of proportion?

Resource update - Naming and shaming

UK premium telecom services regulator ICTSIS has published its quarterly update, which makes grim reading in light of the growing concerns over the new flavors of fraud and abuse which may be enabled in an IP world. The graphic on page 9 suggests we're looking at a near-300% YoY increase in consumer complaints in 2004, and ICTSIS highlights the growth in automated marketing voice spam. Page 12 contains an extensive list of fines and access bars imposed on a raft of unusually-named operators. (http://www.icstis.org.uk/icstis2002/pdf/icstis_quarterly_summer_2004.pdf)

Battle of the P2P voice press releases

Yesterday saw two pretty hefty announcements from the P2P VoIP camp, and the tit-for-tat looks set to continue today. This morning Nimcat Networks announced that Canadian WISP Storm (http://www.storm.ca/Coverage.html) is trialling the nimX P2P platform in-house as a replacement for the legacy PBX. The press release on the deal contains a statement from Storm's CEO to the effect that the company is contemplating selling nimX to its own customers, based on their positive experience to date. Press release is here (http://www.nimcatnetworks.com/Reading.aspx?fid=32&ftype=1). I assume Peerio has something else up its sleeve for later today, and Skype, never one to miss a PR opportunity, remains eerily silent so far...

Peerio plays an ace

A couple of months back, I had an hour or so of face time with Dmitry Goroshevsky of Popular Telephony, in which we covered a lot of ground on the Peerio development roadmap. Along the way, he alluded to the significance of numbering, and today we see the outcome - the Peerio global numbering plan (or GNUP). This system allows interconnection - inbound and outbound - between any VoIP applications and also the PSTN, via a single unified numbering scheme created by Popular Telephony. This opens up some interesting scenarios for interconnection of currently incompatible platforms (for example SIPphone-to-Skype, Skype-to-MSN Messenger) with no gateway (except in the IP-to-PSTN transaction - interesting given Popular Telephony's recent efforts to get Peerio embedded on gateways http://www.peerio.com/news9.html). Popular Telephony will be enlisting service providers which will support the plan (I assume this will initially mainly include ENUM specialists, but who knows? A list is apparently forthcoming at www.gnup.org/connectivity), and will also build a global user directory (www.gnup.org/yellowpages).

This announcement adds another layer to the Popular Telephony strategy. To take stock of what has been announced so far and where all this is heading (as far as I understand it):

1. Peerio takes steps towards becoming the preferred embedded VoIP solution for the hardware space and on the enterprise desktop, effectively making a play for disintermediating the PBX and Centrex players in the process (the previously announced deals with Adtech, Atemis, Logicom and VONTEL). In other words it gains a position at the end-points and also can serve as a replacement for part of the legacy local network;
2. Peerio gains a place on media gateways, i.e., further up the network, to facilitate transactions between itself and legacy wide area networks (the Quescom deal).
3. Peerio teams up with an acknowledged leader in audio codecs, whose products are proven to perform well in an environment of high packet loss, for which I read "wireless" (the Global IP Sound deal from last week).
4. Now, rather than merely hold itself up as "a better alternative to Skype," or something along those lines, and pit Peerio head-to-head with everyone else, Popular Telephony sets up Peerio GNUP to forge compatibility between incompatible platforms (but at the behest of the end user, not the application provider). It effectively oversees interconnection arrangements, numbering and the directory function - globally.

Whew. At this point I'm left with the inescapable conclusion that Popular Telephony is trying to rewrite the rules for pretty much the entire industry with the exception of access. Perhaps now some more mainstream media attention will follow...

Telio flattens the playing field

Vonage's impending (and long-delayed) entry into the European market is getting a fair amount of coverage in the States and in the UK (Telecom Markets this week speculates that a deal with NTL may be on the cards in the UK). However, when Vonage arrives, they may find that the market has moved on a bit since they first talked about European expansion a year ago. For example, in the UK market, Gossiptel (www.gossiptel.com) has about 5,000 users and is currently adding about 125 - 150 subs per week to its consumer service with zero marketing spend - but more importantly, the various business units comprising Gossiptel also contain all the components of a successful one-stop white label offering (hosting/managed services, back office, call center, interconnect). I suspect it was a similar asset portfolio which led UK ISP Business Serve just last week to acquire Pipemedia (http://www.pipemedia.com/pressreleases/BusinessServeAcquisition.pdf, which operates the Pipecall service blogged by Andy Abramson on Saturday).

Another area which may prove challenging is pricing. Today Telio, "the little Norwegian telco that could," is to announce a flat-rate calling plan for its subscribers covering calls to all fixed lines in Western Europe and North America, for the same fixed price of NOK159 per month. In other words, for EUR19.50, the monthly price that Norwegians pay just to have an access line in the traditional fixed line world, unlimited fixed line calling is now a reality to something like 400m lines on two continents. When Telio went live officially in February, EUR19.50 covered flat rate calling in Norway alone. This was expanded to all of Scandinavia only in June. So, while Telio pricing has remained constant, the flat rate calling "catchment area" has expanded by a factor of around 114x in eight months. For heavy international callers this must equate to effective annual price erosion of something like 70%, if not more. No doubt the Nordic competition regulators' working group, who today issued a report (http://www.kkv.se/bestall/pdf/rap_telecompetition.pdf) stating that telecoms pricing in the region still has significant room to become more competitive, would cheer such rapid price compression benefitting consumers.

Telio's press release from today again highlights its intention to expand this offer to other consumers in Europe, and we expect this will come both by means of footprint expansion, as well as in the white label arena, where we think a deal may be imminent in Northern Europe. (Incidentally, it is my understanding that a number of private market transactions in Telio equity recently indicate a public market value of EUR25 - 30m if it were a publicly traded company, and that it will be profitable at the net income level this year. Just goes to show what's possible when companies don't burn cash in marketing spend.)

Did I give you my card?

The always interesting Get Real blog from Stowe Boyd over at Corante (www.corante.com/getreal) has already posted on this, but it is interesting enough for me to add my two cents' worth. Back at VON Europe in June I included in my presentation some data from Nielsen//NetRatings (http://www.nielsen-netratings.com/pr/pr_040416_uk.pdf) which showed that European IM users rack up roughly 180 minutes per month across the seven major platforms, and if we take only the three most popular (MSN, Yahoo! and IRC), the figure is around 240 minutes. This, I noted, was a pretty good approximation for the average outgoing minutes of voice on a typical European PSTN line in a given month. The conclusion, and one that Skype seems to have grasped more than any other VoIP player to date, is that among its devotees, IM is a very serious source of voice traffic displacement. AOL's new AIM revamp takes this acknowledgement a step further, by providing a template for AIM users to create an AIM Buddy Card (http://aimtoday.aol.com/features/main_redesign.adp?fid=aimbuddycards). Conventional business cards still often contain largely useless information like FAX numbers and other relics of the pre-IP age. Why shouldn't we be including our IM/VoIP user IDs instead as more relevant bits of our identities?

Resource update - the German take on VoIP

While Boston hosts the VON Fall 2004 shindig, on this side of the pond, German regulator RegTP today opened its Telecommunications Forum in Bonn with a speech from RegTP President Matthias Kurth. The speech (http://www.regtp.de/imperia/md/content/en/aktuelles/Rede_Kurth_TK-Forum_2004_en.pdf) makes fairly familiar reading for those conversant with regulatory issues surrounding VoIP, but it is interesting and encouraging to see the regulator publicly commit to giving careful consideration to naked DSL.

Nimcat gets embedded

Nimcat Networks of Canada has today announced a deal with Canadian hardware maker Aastra Technologies (www.aastra.com) and Broadcom to incorporate Nimcat's nimX P2P solution into Aastra's 480i IP phone, embedded on Broadcom chips. This interesting deal highlights the interplay between the chip maker and application developer (Broadcom is an investor in Nimcat), and echoes the previously announced deals between Popular Telephony and hardware partners. In keeping with Nimcat's stated market focus, the Aastra phone is targeted at the SME space specifically. Popular Telephony, in contrast, seems to be widening its net to encompass a much larger market segment (more to come on this issue).

Resource update - Legal vs. P2P

Interesting graphic in this press release from NPD Group (http://www.npd.com/dynamic/releases/press_041013.html), which shows that P2P and legal music download user numbers have moved pretty much in tandem so far, at least until May/June, since when P2P has seen a growth spurt and legal services appear to have stagnated. NPD says this relates to the end of promotional offers from legal sites during that time frame. This seems to seriously call into question the assumption that legal services constitute some sort of magic bullet to overcome P2P.

On-Instant on the move

On-Instant, which announced a PSTN breakout service in partnership with Level(3) a couple of weeks back, launched a new site overnight, which includes a user guide that gives more insight into the service (http://on-instant.com/userguide01.html#first). I spoke with founder and CEO Nick Ogden this morning, and he says the company currently has users in 161 countries, and looks to be on track for 100,000 users by year-end. Not bad for a company of 16 people which only launched service in mid-May. Look for a significant announcement next week.

High fiber anxiety

While the FCC yesterday seemed to hand the RBOCs back a significant degree of control over their destinies with the decision on FTTH (Om Malik's stinging assessment is here: http://www.gigaom.com/2004/10/shrinking_consum.php), Europe is shaping up to be a horse of a different color, at least in countries like the Netherlands. Housing corporations are big in the Netherlands (and in other parts of Europe, an issue we covered in some depth in EuroTelcorama No. 5) - they own about 3m of the 7m homes in the Dutch market, and increasingly are looking to differentiate themselves by offering additional features and services, and that includes FTTH.

There are a number of significant projects in the pipeline (hopefully I will be in a position to update in greater detail in a few weeks' time), but already the cable companies are getting their feathers ruffled. Last week, they were so upset by the 1,500 home pilot project being launched in Tilburg en Breda, that cable association Vecai wrote a formal letter to the Dutch Housing Minister to complain

(http://www.vecai.nl/downloads/docs/Brief%20Min%20Dekker.pdf). Their case is that the housing associations are acting outside their remit, effectively with a state subsidy, and that tenants will pay in the long run in the form of higher rents.

The housing associations would probably argue that they are providing a superior service at a superior price to what the market has delivered so far, and that in fact, this issue goes further than a competitive telecoms platform. Their next step is probably to integrate security monitoring, telemedicine, and e-government services to meet the needs of an aging tenant base. This arguably makes it more of a social welfare issue than a telecom market spat, and the government would probably have a tough time arguing with that.

This tension between public policy goals and the shorter term concerns of the capital markets is an issue we may see highlighted elsewhere in Europe in the not-too-distant future (check out Bryggenet in Denmark, where the tenants of a housing development got fed up with waiting for broadband and just did it themselves http://bryggenet.dk/).

Woops! Flarion/NetGear jumps the gun

Flarion accidentally issued a press release today which was meant to be delayed until Monday, and have subsequently asked for it to be treated as under embargo until then. However, it involves a publicly traded company, NetGear (NTGR), and this should be price-sensitive information, so it's going to be interesting to see what pans out later today when Wall Street wakes up. No prizes for guessing that this phantom announcement relates to the integration between FLASH-OFDM and Wi-Fi, which sounds like exciting stuff. More news on Monday, or was that later today?

Re-purposed Skype

VoIP blogger extraordinaire Andy Abramson has a fascinating post on his site about a third party work-around for connecting Skype to the PSTN (http://andyabramson.blogs.com/voipwatch/2004/10/personal_skype_.html). This reminds me of some of the IM-to-PSTN solutions we have seen previously (like the BAFO Messenger Call Box), but with the added attraction of Skype's bullet-proof voice codec. There is also another interesting multimedia post from the field, which is probably a pretty good representation of how the world is changing for the telcos(http://andyabramson.blogs.com/voipwatch/2004/10/always_connecte.html). BT Openzone might be making GBP6 from Andy for an hour, but the cellular and fixed line carriers are not seeing one red cent from his prodigious voice traffic. Given the level of roaming charges, the GBP6 for an hour probably seems like money well-spent.

Telco triple play, easier said than done

This article from Light Reading highlights some teething problems in Swisscom's IP TV deployment in partnership with Microsoft (http://www.lightreading.com/document.asp?site=lightreading&doc_id=60886), though the article strongly suggests this may have more to do with the vendor than the concept.

Peerio's roadmap - another signpost

The Popular Telephony announcements are accelerating at a nice clip, and later today, the company is to unveil a partnership with Global IP Sound (www.globalipsound.com), to make GIPS' "VoiceEngine" the voice processing technology of choice for Peerio444 (the free consumer client) and PeerioBiz (the enterprise desktop edition). It will also be the recommended voice processor for the middleware to be embedded on hardware bearing the "C'est Peerio" designation (again, think "Intel inside"), though it is my understanding that embedded product manufacturing partners will have discretion if they prefer/are already using another voice processing solution.

This announcement marks pretty rapid momentum by GIPS as well, as a follow-through of the announcement they made only late last month regarding Embedded VoiceEngine(http://www.globalipsound.com/newsroom/newsroom.php?newsID=79). Beyond their expertise in the field, it is noteworthy that GIPS has client relationships with Nortel and WebEx, as well as being a member of the Intel, Motorola and Texas Instrument design/developer networks. This is further evidence of the traction that Peerio is gaining in getting buy-in from credible, established partners who themselves have good client/partner relationships to build on.

GIPS claims that its products can withstand up to 30% packet loss and still maintain telco-grade quality, and indeed the company wrote the codec which makes Skype the robust client it is. Those who use Skype are undoubtedly aware of the voice quality, and may have used it under extreme conditions with no noticeable degradation. I have used it (inadvertently, because of a 3G network connection failure) over GPRS, and was unaware at the time what was happening. The resiliency of the GIPS codecs, coupled with the connections to Intel, Motorola and TI, raises some interesting questions about the timescale for seeing Peerio on wireless devices.

Resource update: Measuring the value of the BBC

This is interesting in its own right, but within the scope of this blog, it may offer some useful background as the PTTs gird their loins to enter the brutal world of television. The study shows the "consumer value" of the BBC to be GBP18.35 - 18.70 per month, and when asked how much they would be willing to pay to keep the BBC in existence, 81% of respondents said they would pay at least GBP10 per month (in other words, the same as under the current license regime). The result in the previous survey in 1990 showed that 90% of respondents would be willing to pay the same amount as the current license fee (then GBP6 per month). Nearly half (42%) of the respondents in the 2004 survey said they would stump up GBP20 per month, and 19% would be willing to pay GBP30.

However, when the emotional appeal of "saving the BBC" is taken away, a different picture emerges. Respondents were asked about price levels for a subscription-based service, and 4% said they would pay GBP7 per month for BBC One (!), 17% said they would pay GBP9 for BBC One and Two, and 22% said they would pay GBP11 for a package of BBC One, Two and BBC Digital channels. A full 58% said that, if these were the only packages available, they would opt out altogether. This 58% figure is very, very close to the household penetration rate for multichannel TV in the UK (59% in Q2, if we include analogue cable, 56% if we include only digital). In other words, we might go out on a limb and say that the only people who appear willing to pay (if it were a voluntary system, rather than compulsory, as it is now) at this point are people who would otherwise go without. Those who have a choice and are able/willing to take it appear to have voted already.

There are probably a lot of complex factors involved in this development (the growth of pay TV, public dissatisfaction with BBC programming, the advent of the internet as a leisure activity/information source), but the message seems to be fairly clear for anyone looking to enter a mature TV market. If you expect someone to pay for it, it has to offer something different, it has to be good value, and it ought to be a good candidate for displacing something else in the consumer's already crowded media diet.

The BBC survey can be downloaded here:(http://www.bbc.co.uk/thefuture/pdfs/value_bbc.pdf)

Sky's the limit

I have a meeting with Sky scheduled for this Thursday, with a view to taking up formal coverage. My impression is that most of the people covering the company currently come from the media sector (which is a real mess in terms of the rather ham-fisted industry classifications under the 18 sectors in DJ STOXX 600), and I'm looking forward to throwing some telecoms/internet questions at them and seeing what comes back. I'm particularly interested to see if they have any market share assumptions for TV-over-DSL and overall how much of a credible threat they (claim to) believe it to be. Among the many questions I have for them, the most interesting to me are: since BT is apparently about to declare war on Sky with the mooted Freeview tie-up, a) is there a telephony strategy afoot (clarification - I should have added "beyond the legacy SkyTalk carrier pre-selection product, which has c.350k users"); and b) what are the chances that we will see Sky enter the broadband market in the not-too-distant-future?

I VERY STRONGLY believe the answer to the telephony question is "yes," though I will be interested to hear the company line on this, and I believe the answer to broadband must also be that it is a natural defensive move. I eventually envisage some sort of Sky+ "turbocharged" edition PVR, with ADSL modem embedded. Clearly there may be some sensitivities around cannibalization of the DTH product from a broadband platform, but it is also interesting to contemplate how the business model might evolve, particularly in light of moves by Murdoch stablemate NDS in the area of DRM for IP TV(http://www.nds.com/newspdfs/VGS_100904.pdf) and for content distribution more generally (http://www.nds.com/newspdfs/SVP_Alliance_100904.pdf).

We should not underestimate what a telephony/broadband challenge from Sky might mean for BT Group. A concerted push by Sky into telephony would mean that 7m digital TV homes (which are contractually obliged to have their set-top boxes connected to a PSTN line) would be candidates for a bundled offer from Sky - that's nearly three times the size of the total digital cable footprint in the UK. And given the success that the two UK cable companies have had in bundling services (90% uptake of telephony services, 22% triple play penetration, on average in Q2), this is even more interesting, or conversely worrying, if you're a telco.

Resource update - DAB in the UK

Last Wednesday, while I was flat on my back with a mouth full of painkillers, the Digital Radio Development Bureau of the UK issued a digital audio broadcasting (DAB) market forecast update to 2008 (press release http://www.drdb.org/article.php?id=299&from=lat - there is a more detailed .pdf version which I received direct from GWR Group, but I haven't been able to locate a version online). The industry is looking for a significant ramp-up in receiver shipments, from a projected 1m cumulative by year-end 2004, to 4.5m by year-end 2006, and on to 13m by 2008. On an annual basis, that equates to sales of 750k units in 2004, 2m in 2006, and over 5m in 2008, for an annual market value of GBP500m by that year.

Clearly, the $500m deal last week between Howard Stern and Sirius in the US has rung some alarm bells about the disruptive potential of alternative digital broadcasting platforms, and the UK may prove to be another case in point, if from a different angle. And, we're not just talking "radio" here. The DAB infrastructure holds latent capacity for other, quasi-telecom services, as in this case (http://www.gwrgroup.com/documents%2FLivetimepresentation%20final%2Epdf). It's going to be intriguing to see how this platform develops, and what sort of challenges/opportunities it throws up to more mainstream service offerings.

Sunday Papers

UK weekend paper The Business today carries an article which highlights just the sort of risk we have been expecting in the UK market for some time now. It claims that France Telecom's broadband ISP unit Wanadoo will launch a marketing offensive in January 2005 which will see its agressively-priced DSL offering (GBP17.99 per month for 512kbps) add unlimited national calling to fixed lines via VoIP for no additional cost. This is hardly a surprise, but it is nevertheless a perfect confirmation of the nightmare scenario we have long envisaged - wherein PTTs adopt the strategies which attackers have used so effectively against them in their home markets (France Telecom should understand this better than anyone) to create market share gain opportunities in other European markets where they already have a reasonable foothold.

We already expect this of Tiscali in a number of markets, where it has contracts out with UT Starcom for an IP triple play solution. France Telecom may prove to be another proactive spoiler in the Netherlands and Spain, as may T-Online in France and Spain, should it wish to pursue such a path. For those markets where this dynamic is at work, we expect the end result to be some significant volatility in price points for both consumer broadband and pay TV. It is difficult to see what sort of comeback players like BT or KPN may have in this scenario, given their respective lack of consumer ISP businesses in Europe with which to strike back.

The article contains one tasty bit of market forecasting from the chief executive at DeTeWe UK: "...telecoms incumbents worldwide generate around $450bn in free cash flow. By 2007, 30 - 40% of that will be sucked into the black hole of voice over internet."

Flarion and Siemens hotwire 450MHz

Flarion Technologies announced late yesterday that it had entered into an agreement with Siemens to collaborate on mobile broadband technology operating in the 450MHz frequency, which is expected to result in an end-to-end solution from Siemens available from Q2 2005. The 450MHz range was the old analogue frequency for mobile in much of Europe, but has now largely been supplanted by GSM, and only remains in use in some remote areas of Scandinavian countries, where GSM coverage may be spotty or non-existent. The 450MHz range has been re-farmed in parts of Eastern Europe (Romania, Czech Republic, Russia), where CDMA450 has come into operation (you can view a presentation from last summer detailing the current state of 450MHz deployments here http://www.cdg.org/news/events/CDMASeminar/040615_CDMA450/Session1-IA%20450%20presentation%20Warsaw.pdf).

This is interesting from a number of perspectives. Firstly, it is a nice counterpoint by Siemens to the efforts that Ericsson and Lucent have made in the CDMA450 arena so far, and also to the push in some quarters towards GSM400. It is also interesting because Flarion has previously announced discrete, country-based trials of its technology with individual operators (DT in the Netherlands, Vodafone in Japan, Nextel in the US, various Korean cellcos, Telstra in Australia). By tying up with Siemens, the company gets significant backing from a major equipment player, to deliver a one-stop package to operators across Europe, which may give the technology a faster adoption rate than it has had so far. Lastly, the potential for further European mobile market fragmentation obviously grows larger if this gets off the ground in a significant way, especially if backed by proactive regulation such as in Sweden, where new 450MHz licenses are expected to be issued by the end of this year, as far as we know.

What's in a number?

(This is my first post in a few days as I've been laid low by some medical matters since Monday. Hope to resume regular service soon.)

There's an old joke about people's willingness to take statistics too literally: "Statistics say that one in five people in the world is Chinese, but I come from a family of five and none of us is Chinese." Today, in conjunction with its legal salvo against file sharers (in tandem with the IFPI's suits against 459 heavy users across Europe), the British Phonographic Institute (BPI) has stated that 15% of file sharers are responsible for 75% of content on P2P networks (http://news.bbc.co.uk/1/hi/entertainment/music/3722428.stm). If recent P2P user statistics are accurate, this suggests that the labels should ultimately pursue action against 1.5m people globally, which won't do much for industry margins.

Anecdotally, it's pretty widely accepted that P2P sharing activity is not symmetrical, and the "free rider" phenomenon is a well-documented source of network inefficiencies in P2P networks. However, we also have to acknowledge that the labels are shooting at a rapidly moving target. The P2P user base is almost certainly expanding in line with broadband growth, and the most explosive growth is in some places where the mechanisms for enforcement along these lines are far from clear - China is now the world's largest installed base of DSL users, with 13m lines. Moreover, it's far from clear that the underlying motivations of file sharers are as straightforwardly sinister as the industry would have us believe, or that the net effect of sharing is actually negative (both issues which we touched on here http://eurotelcoblog.blogspot.com/2004/09/daiwa-eurotelcoblog-no_22.html).

Lastly, the knock-on effects from actions such as this have historically been unpredictable. An industry analyst quoted in the BBC article linked-to above rightly states that a certain proportion of the youngest generation of file sharers may prove hard for the labels to convert to willing consumers of CDs, due to conditioning. However, it is inconceivable that goodwill towards the labels among such young consumers will be enhanced by litigation, while the lure of file sharing as "something naughty" may conversely be reinforced. The industry is targeting KaZaA, Gnutella, and eDonkey users in Europe this time, though as with the RIAA lawsuits in the US previously, this may merely lead to user migration to other platforms with lower profiles, or worse from the perspective of enforcement, closed systems, such as W.A.S.T.E.

Daiwa EuroTelcoblog No. 81: Friday, 1st October, 2004 - Skype's top 20

Skype has kindly given me a snapshot of their user breakdown by country, in percentage terms. It should be noted that this is not definitive, because the Skype network is a function of the people who download and use it at any given time - and the composition of this group has changed dramatically since the service launched a year ago. At the moment, I can see 750,000 users online, but my experience is that this number fluctuates from under 500k to well over 800k in any given 24-hour period. Given the kind of registered user numbers Skype has publicly claimed (c.10m users out of 25m downloads to date), it is difficult to get a clear reading from this breakdown as to precise total user numbers by country on any kind of consistent basis.

[Anecdotally, I use Skype on two different machines, so at any given time, I may appear as two users on the network, and other Skype users I know have said the same. If this were true of most of the 10m users, then a conservative estimate of Skype users would be more in the 5m+ range. However, a proportion of users may only use it when travelling or at work, so I would tend to think that the "true" base lies somewhere between 5m and 10m. If that is true, then the number of active users visible at any given time probably represents 10 - 15% of the total base.]
In any event, the breakdown is interesting and prompts many questions to which there may be no apparent answers:

• There seem to be few discernable correlations between population or broadband penetration rates. Tiny Israel and Denmark come out ahead of Japan, Poland punches well above its weight and broadband penetration, France looks to be almost double the user base size of the UK (despite identical population size and a broadband user base only 15% larger than that of the UK) and to my surprise South Korea is not even in the top 20, despite a rabid broadband market.
• Maybe the lack of consistency is price-driven, related to the large ex-pat, student and guest-worker diaspora. A caller in Brazil to the US on the Embratel network might expect to pay an undiscounted price of EUR0.21 per minute at current exchange rates. AT&T's discounted call plans show charges (for calls to fixed lines) of EUR0.14 per minute for Brazil (EUR0.11 for Rio and Sao Paolo). Both of these are fairly steep, and in this case there may be a good case to explain the Brazilian prominence in terms of toll-bypass. However, for AT&T users (for example), Germany is in the 10-cent band with the rest of Europe, and calls to Taiwan from the US on AT&T cost only 9 cents per minute (7 cents to Taipei). Denmark has some of the lowest calling rates in the world, and is an affluent society with a small ex-pat population. Its prominence in the rankings thus throws up more questions.
• Taiwan at No. 2 is also interesting, as it is the one market where Skype has a co-branding agreement with an ISP.
  

All of this is very intriguing, but it is probably fruitless at this point to look for common triggers across these cultures for Skype uptake. It's probably part curiosity, part convenience, part social/viral, part generational and part economic, to varying degrees in each country. This in itself is a worrying message for the telco community, as it suggests that perhaps something more complicated than simple price/feature comparison is at work in the mind of the consumer (interesting in light of the savage price war which has just broken out between Vonage and AT&T in the US).

[1] USA - 10.33%
[2] Taiwan - 9.24%
[3] Poland - 8.78%
[4] Brazil - 7.24%
[5] Germany - 6.18%
[6] China - 5.89%
[7] France - 5.53%
[8] Netherlands - 3.50%
[9] Denmark - 3.07%
[10] UK - 2.94%
[11] Israel - 2.94%
[12] Japan - 2.61%
[13] Canada - 2.46%
[14] Belgium - 2.10%
[15] Spain - 1.82%
[16] Sweden - 1.76%
[17] Australia - 1.46%
[18] Italy - 1.44%
[19] Switzerland - 1.22%
[20] Mexico - 1.07%