Thursday, August 31, 2006

Third fiddle

comScore Networks has launched a monthly ranking of European online properties, and the inaugural edition contains some interesting observations. Most dramatic is that Google's market share in Europe is significantly higher than in the US, at 75% vs. 60% across the pond. There is also an intriguing country-by-country breakdown of the top three which shows Google and Microsoft slugging it out for top position in every country except for Germany and Norway. And for the EuroTelcos, there's a rare bit of good news: they are surprisingly well-represented. Vodafone manages to claw its way to number 10 on a pan-European basis, and in Belgium, Denmark, France, Italy, Portugal and Switzerland, the local incumbent ranks third. Most staggering to me is that Telenor beats Google into second place in the Norwegian market.

And now for ComBOTS

Well, after months of buildup, tomorrow is the first view the general public will have of ComBOTS. So far ComBOTS seems to have been totally ignored by the English blogosphere, though I have come across some earlier coverage in German, more than one of which suggests it is a memory and CPU hog (the first article claims 130MB of working storage and 20 - 70% CPU utilization), and more recent coverage suggests an inauspicious start. Still, it would appear that there are some impressive graphics and features involved. Tomorrow will tell.

Without a trace

I was just having a conversation with a friend earlier today about the latent demand for anonymizers which would likely arise from growing concerns about internet datamining and the unforeseen consequences of otherwise noble intentions of law enforcement. Little did I know that in my inbox sat an announcement heralding the launch of Browzar, a self-anonymizing web browser from the founder of Freeserve, Ajaz Ahmed. Not only does it claim to be self-cleaning, but can also be run without a download if the user so desires. Harvest that!

Makeover for an ugly duckling

In some of the presentations I've done for clients and others I've sometimes touched (in the context of what Google, eBay, Yahoo!, et al are up to in voice) on the somewhat puzzling case of Amazon, which I guess some people view as something of an also-ran in web services. Indeed the stock has underperformed Google by 26% and Yahoo! by 8.5% year-to-date (neither of those have been stellar, incidentally), and I guess the company must be feeling pretty unloved right about now. So what's the way forward towards reinvention? According to my man Thomas Anglero up in Norway, it's about enabling applications developers and businesses with massive storage and virtual computing. And just in case you find his views far-fetched, check out the comment following Thomas' post, which apparently was made within three minutes of the post itself, by the co-founder of indy video newcomer Cruxy, which is entirely run on the S3 service (and seems pretty damned fast).

Battle of the press releases

At 9:39 I got an email from Telefonica/O2 announcing a new set of bundled mobile/DSL offerings in the German market (the English version shows the top-of-the-range 16Mbps offering coming in at EUR50 per month for O2 contract subs, including flat rate calls to domestic landlines and O2 mobiles). By 10:15 DT had fired its IFA press cannon, revealing the first steps towards quad-play offerings in Germany with a triple play flat-rate offering starting at EUR81 per month.

Blogospheric re-entry

Well, mega-uber value readers, I'm back for real now, and hunkered down for a busy autumn, during which I promise to endeavor to post more frequently and interestingly to both this bloglet and the somewhat neglected Chaotica. As I warned before I went away, I knew something major would happen while I was unwired, and indeed there was no shortage of interesting news, most of it predictably coming from outside EuroTelcoLand:

Google/eBay: I don't know how surprising this deal really was, given the click-to-call obsession of the past year and the efforts Google has been making towards achieving universal interoperability with other IM clients. I have for some time been of the suspicion that Skype would be forced at some point to pursue such an arrangement, though the official line has been that interoperability would be driven by user demand, which was reportedly not strong. Then again, what we have here is an agreement to "explore interoperability between Skype and Google Talk via open standards to enable text chat and online presence." Not the whole hog, but a step in an interesting direction. Andy makes an interesting observation about Skype becoming an alternative, communications-centric browser, which I thinks jibes well with some comments made by Niklas Zennstrom at VON Stockholm, to the effect that Skype has value as a conduit/enabler for other applications. This is going to run and run.

I reckon more than one telco music download product team hastily updated their CVs upon the revelation of Universal Music's deal with SpiralFrog. I am skeptical that the formula will work (DRM was made to be cracked, and there is no shortage of WMA conversion software available), but then again if we see another major label jumping on board maybe it will grow teeth.

I could go on and on about the fascinating developments around YouTube, Goople, etc., but the point is that telco newsflow continues to be pretty turgid by comparison:

  • While the global VoIM players engage in an ever-changing alliance exercise resembling a game of Twister, on the ropes, in the corner swinging wildly, BT counters with a me-too VoIM client notably not co-branded with Yahoo! (the fine print alone is enough to make the Skype or Gizmo user titter with contempt).
  • If you can imagine what the word "Duh" sounds like heard through a stifled yawn, then you can imagine my reaction to the Vodafone Belgian retreat. This is step one on my long-established roadmap to a slimmer, trimmer Vodafone, and I also expect a Swiss-shaped hole in the global domination map by Christmas. And, once again, despite all the "asset-lite" rhetoric, I fully expect some broadband acquisitions will follow.
  • VDSL T-Day came and went in Germany, with the European Commission using some strong language and also urging a fast-track to naked DSL. Read T-Regs' excellent overview.
  • Dick Notebaert of Qwest wants to protect the hapless consumer from the evils of net neutrality. If anyone can translate his comments into a language I can understand please send your version along.

So, compare and contrast: expansion, risk-taking, innovation versus regulation, retreat and paranoia. God am I glad to be back in EuroTelcoLand!!!!

Friday, August 18, 2006

Blue screen

Yes, I know I only just returned from a short break, but I'm sitting on a lot of unused holiday time and so am off yet again for a short break ahead of what I expect to be a very busy September. I feel EuroTelcorama No. 7 may be coming on (number 6 nearly killed me)... Be forewarned, something major will undoubtedly happen while I am in a converted barn in the backwaters of Dorset.

There's no justice!

Milton Keynes??????!!!!

UPDATE: Okay, I really meant this as a humorous jab at one of the UK's obvious town planning whipping boys, but this post really got the mega-uber value readers going. Long-time reader and Palladium Club uber-charter member Keef wrote in almost immediately to point out that Pipex is snug with local government via their commercial relationship, and fellow Palladium Club member Dean also chimed in with an interesting observation.

Something for the weekend

Plenty for EuroTelcos to ponder over the coming weekend. While the Dutch market will no doubt continue to astound, there is much for others to be paranoid about. Reuters is quoting a spokesman from Viviane Reding's office as saying that Monday is D-Day (or should that be T-Day?) for German broadband. In the UK, not content merely to offer free TV for customers which take its fixed telephony product, NTL/Telewest/Virgin has opened up another front in the free broadband wars: offering 8Mbps DSL free for a year for new customers on Virgin Mobile's SIM-only packages.

Second gear

Yet another jaw-dropping development in the Dutch access market, which is doing much more than making up for the late summer doldrums elsewhere in the sector.

Today we learn (again from Vincent Dekker at Trouw - can this man smell a story or what?) that one of Rotterdam's social housing corporations, Stadswonen, is next week to launch an open FTTH network to 5,500 homes in the inner city, serving primarily the student and first-time buyer markets. As with the proposed Amsterdam muni project, this initiative is also managed by Telecom Italia subsidiary BBNed. The article quotes Stadswonen's CEO as saying that the inspiration for the initiative came from requests from American and Korean ex-pat students who missed having high-speed connectivity in their accomodation (presumably being accustomed to fiber on campus back home), and the company decided it couldn't wait for whatever fiber might be coming from the municipality.

Now for the really exciting part - the pricing.

For residents who want a fiber connection, they pay an incremental rental fee of EUR5.40 per month if no fiber tail is present in the building already and only EUR1.80 per month if fiber is already present (which is the case in some buildings). The fact that it's bundled into the monthly rent presumably makes this almost an invisible amenity, in the same way that refuse collection or landscaping is just part of the total package - and it's difficult to see anyone arguing with such a nominal fee, though there is an opt-out for those who don't want it.

Better still is the pricing for connectivity. Students will receive service for free from SURFnet, and for non-students, the first ISP offering to launch on the network (Pilmo, a BBNed ISP, but there will be others, no doubt) delivers 10Mbps and 30Mbps symmetrical services for EUR4.50 and EUR7.00 per month, respectively. THIS IS NOT A TYPO! Not only is this top-tier service equivalent only to the price of a couple of beers (a benchmark for students, I guess), it also represents an 88% discount to UPC's top-tier offering of 20Mbps down/2Mbps up for EUR59.95. Time to get out the strategic pricing manual.

Thursday, August 17, 2006

How to pad your data ARPU

A Palladium Club mega-uber value reader recently returned from a working vacation in Southern Europe writes in with the following tale of woe. Please forward to Ms. Reding.

"Mobile 3G connectivity has dropped to acceptable price levels… or so I thought. EUR30/month for 1 GB of traffic is acceptable for professional usage (as a last resort for email checks). Apparently the thickheaded telco marketeers finally got the point that there is something like price-elasticity, and that high utilization of your extremely high investment in licenses and infrastructure gets you more business and recurring use/customers than creaming off the top with exorbitant pricing and alienating your customers.

Until I got my recent invoice. I had to use the connectivity during my vacation in Spain, connecting through a mobile operator who advertises with the same type of price level: EUR15/mnth for 1 GB.

Well let's make a quick and dirty calculation. Both operators combined (home and Spain) want to get in total EUR45/mnth for 1 GB for connectivity. Add some transport costs for 1 GB (EUR 2 max) and extra billing/handshaking for lets assume EUR10/month. Total EUR57. A hefty profit margin added, let's assume a round number EUR100 for 1 GB.

No. The bill amounted to a mindboggling Euro 8000 per 1 GB traffic."

Haarlem reshuffle

Damn, this Dutch broadband market is getting so interesting, it may justify a spin-off blog. First, it would appear that FTTH pioneer Lijbrandt is in negotiations with the city of Haarlem, presumably over the prohibitive EUR22 per meter repaving charge which Lijbrandt has previously cited as a disincentive to investment. One city alderman describes himself as "overjoyed" with the prospect of a Lijbrandt network, and one respondent in the comments section states, "It's the least the Municipality can do to enable competition to UPC."

So, with fiber (muni, entrepreneurial and incumbent) nipping at its heels and KPN pushing DTT, it would seem the right time for UPC to go on a charm offensive, er, no, instead it's shopping two of its file sharing subscribers' identities to Dutch artists' rights group Brein. The background here is complex - UPC initially denied that it had passed information to Brein, then claimed that the information was only included in court papers to demonstrate UPC's internal controls, though Brein claims that earlier it was explicitly promised such information. Whether this reflects malice or merely ineptness at UPC is probably academic now, as the public is outraged and feeling that trust has been violated. Watch those churn numbers fly!

Monday, August 14, 2006

Early stage diagnosis

At the end of my last post on Dutch fiber developments, which certainly must be one of the most interesting stories in the industry at present, I mused about the possible regulatory complications of a market where access becomes fragmented down to a highly localized level. As if regulators didn't have enough to grapple with. Now, once again from Trouw comes an update on the situation, in which industry consultant Hendrik Rood of Stratix Consulting urges the Dutch Parliament to address the situation explicitly.

I can't work out precisely what measures he envisages with respect to a privately funded network like Hillegom - it seems to me that if Lijbrandt, the company behind the Hillegom network, has gained the appropriate planning permission and is operating within the law, then restricting its (or others') investments on the grounds of preserving the telco/cable duopoly would be difficult to defend. Not that I'm trying to put words into Mr. Rood's mouth, because I'm still not clear on what exactly he proposes.

He does have a good point, however, regarding a likely duplication of infrastructure investment, stating that the Hillegom situation suggests we are about to witness a proliferation of projects. He says, "In Hillegom a commercial company is laying an FTTH network in the ground. Up to now most people thought that this would be too expensive, that it could only be done with subsidies. If this turns out to be possible and many become aware of that fact within the next few moths, the floodgates will open. Then you will see similar initiatives in a lot of cities and towns. The big players will also have to join. For he who doesn't build an FTTH network, will lose 70 percent of his market." At that level of market share, maintaining one's presence in a local market dominated by the new FTTH player looks highly unpalatable. Mr. Rood points to Nuenen, a project I have written about in the past, which has also attained roughly 70% market share among the 7500 households in its footprint, and claims that neither UPC nor KPN could be covering their costs of service in the area with the remaining 30% split between them. So, I guess the question for cable and telco alike in such situations is whether to build, or as KPN appears to be doing in Haarlem, to pursue a wholesale agreement with the new "local incumbent."

I assume the regulators are buying in a lot of aspirin in anticipation of chronic migraines as the traditional relationships and market definitions break down. I guess they will also have to hire more people to deal with an increasingly complex and fragmented market. Perhaps all of this could somehow be viewed as a vindication of the single neutral open network concept...

Thursday, August 10, 2006

It's how you tell 'em

Oh dear. I arrived at work this morning at 6:45 to find that Deutsche Telekom sent out a release at 11:49 PM last night, basically as a pre-release of results and profits warning. Foul! Then again, how does one deliver such a raft of bad news? I could go into detail here, but given that I've been doing that most of the day already anyway, you'll understand if I refrain for now. Suffice it to say that, as with France Telecom before it, this seems to underline just how fast the markets are transforming (along the way, CFO Dr. Eick gave a few datapoints - German DSL pricing at the bottom of the range is now where the company previously forecast it would be at the end of 2007, legacy data revenues are down twice as far as was expected) and how slowly large, complex companies can respond sometimes. In Q1 we're comfortable with our guidance, but come Q2, we need to take a EUR1bn haircut. Woops!

Wednesday, August 09, 2006

Haarlem shuffle

I may have spent the last few days in a tent with two small children, my only connection to the outside world an AM radio, but over in the Netherlands Vincent Dekker at Trouw has been continuing to dig into the FTTH story. As usual, he has come up with some very interesting information.

Firstly, a commercial (as opposed to muni) FTTH project in the small town of Hillegom, halfway between Amsterdam and The Hague, has reported some pretty astonishing progress in its rollout. Of the 7,431 homes in the town, 5,400 are passed by the network, and of that number, 4,600 homes are connected and taking at least one service - a penetration rate of 85%. Of this base, 90% take telephony and/or TV. In other words, within its footprint, the Hillegom fiber network has a 73% share of the markets which have traditionally belonged to cable (Casema) and KPN, leaving these two companies to fight for the remaining 27%. What's more, the total cost of construction and lighting the network has worked out to EUR1200 per home, which should give one pause for thought when we consider that the enterprise value (EV) per sub in the recent Casema transaction was north of EUR1700.

So encouraged is the company behind the project, Lijbrandt Telecom, that it now plans to expand to the nearby town of Lisse (10,000 homes) in October, and to 120,000 homes within the Dutch Bollenstreek heartland within three years. Perhaps most interesting is the revelation that the company behind the Hillegom network, Lijbrant, is owned by entrepreneur Dik Wessels, who also owns construction conglomerate VolkerWessels, which is already involved in a number of muni fiber projects, and whose Reggeborgh company recently acquired a very high quality national backbone. It looks to me as though Mr. Wessels is gradually amassing assets which could pose a real challenge to both KPN and cable.

Next, Trouw trains its sites on Haarlem (yes, American readers, this is the original), a city of 145,000 bordering the Bollenstreek region, where it transpires that apparently KPN wants to do FTTH and share the cost of digging with Lijbrandt, the very same company behind the Hillegom project. Lijbrandt reportedly isn't hot on the idea, principally because the city charges EUR22 per meter for restoring the streets to their original state. The article goes on to report that KPN is in negotiations with Lijbrandt about offering services on the Hillegom network, which presumably means IP TV.

All this points to a new way of working for the incumbent: partnering with those who have the expertise in infrastructure projects to share costs in new build (even if doing so enables another strong competitor in the process, it's better than missing the boat completely), and being a wholesale customer on networks where it has already been trumped (as in Hillegom). It also points to a future market with much wider regional/local variations in terms of level of dominance and market share, and presumably, once again, a new set of challenges for regulators. Does 13% market share (my estimate) in a town like Hillegom really equate to significant market power? Should the incumbent which finds itself in such a situation be held to the same price controls in this particular town, or should regulation be restructured to suit a more varied regional or local situation?

Wednesday, August 02, 2006

More summer reading

Check out this detailed survey of internet usage in Macedonia, where a USAID-funded program has delivered free wireless connectivity with Motorola's Canopy kit.

Take the Amino challenge

I got a message from IPTV set-top box makers Amino asking me to take part in this survey, so I thought the more the merrier.


EuroTelcoblog and Chaotica are heading off for a short camping trip near the White Cliffs of Dover. One of the interesting bits of beach reading packed in our bags will be this study of IPv4 address exhaustion from the Japan Network Information Center. If you prefer the original Japanese version, it was published in April and can be found here.

Tuesday, August 01, 2006

Sleeping with the enemy

At first I thought it was just my imagination during the KPN conference call this afternoon, but it sounded to me as though CEO Ad Scheepbouwer more or less totally ignored a question about muni fiber. (I love how often this comes up in analyst meetings now as a source of concern, well after, ahem, certain people in the blogosphere began talking about it.) Now, thanks to Vincent Dekker at Trouw (my personal vote for telecom journalist of the year), we discover that apparently KPN plans to opportunistically take advantage of the open ground exposed by the Amsterdam Citynet project to lay its own fiber alongside that of Citynet (presumably in its own ducts), and share the cost of restoring pavements, etc., as well. Vincent's opening gambit is a classic: Amsterdam is on the verge of claiming a world first - two competing fiber connections to the home.

UPDATE: A couple of Palladium Class mega-uber value readers have taken me to task over this post and the previous one on KPN, claiming that I have oversimplified the situation. Sometimes I worry that I am writing in too compact a fashion and assuming too much in terms of my readership, to the detriment of what I'm trying to convey. In fact, some friends have occasionally chided me about the need to be clearer, to spell things out, and this looks like a case in point. Mea culpa. It wasn't intentional though - I was merely trying to convey a sense of what the casual listener might have walked away with from the Q&A on KPN's conference call, namely a distinct impression that wholesale was the only option. So here's my followup, and I hope it's clearer:

  • Both readers say that surely KPN must be envisaging some collocation provisioning, and yes, I would be surprised if they weren't. One goes on to say that this may be forced upon them by the regulator in any event, so why not play nice and make provisions ahead of time? Agreed, though it seems that the regulator is being pretty passive at present.
  • One states that while the ULL players would undoubtedly find building out to all 28,000 nodes unpalatable, collocation probably does make economic sense in certain dense areas. I agree, but I still have questions about just how willing some local authorities will be to grant permission for new street cabinets to companies other than those with established rights of way, i.e., KPN and the local incumbent cablecos. For those areas where it is uneconomical to collocate, there will still be the KPN wholesale offering. The choice for the competitors is then whether to take this as a compromise and move forward with a better product, or stick with DSL and wither.

What I like, and dread, about this whole situation, is that it calls so many fundamental assumptions into question. Yes, KPN's plans may necessitate/stimulate further investment from competitors, which after all is what the EU is after. So KPN is actually helping to advance the EU's vision and the free market. In the case of Amsterdam, surely having two competing fiber offerings, plus coax and DSL, is an improvement on the status quo?

But let's get real for a minute. When we talk about stimulating further investment from competitors, exactly who are we talking about, and how crucial is the Dutch market to any of them? Is the pain actually worth it?

Start with Versatel, on which I and many others have expended far too much print in the past. Now a part of Tele2 (whose share price got hammered once again yesterday on some wobbly results), it is, in my opinion, just one small piece in a large and fairly incoherent puzzle - and probably not worth the effort.

Tiscali - this company has withdrawn from so many other markets that I've lost count, and I'd be surprised if the Netherlands ranked highly among the three which remain.

bbned - there have been countless stories about Telecom Italia looking to dispose of this business, and I can't honestly believe that anyone at the mothership regards this as a core business, especially if there are major new investment hurdles to be jumped. Perhaps someone in the Netherlands does see it as a strategic asset and will buy it - only time will tell.

Orange/Wanadoo - France Telecom has some major fish to fry in the UK, Spain and its home market. Q2 saw negative net adds in Dutch broadband, and the mobile business in the Netherlands is one which, in my experience, many analysts/investors think the company could do without.

I know there are others, but this is a pretty representative list of the majors. While I understand the pro-KPN case in theory, I'm not sure it matches up well with the reality of the situation, and perhaps a direct way of summing it up would be that KPN is moving the goalposts and testing the appetite for those remaining to chase the ball. Part of this relates to technology change, obviously, and the need to get more bandwidth closer to the end user, but it is also a convenient strategic coincidence in that it necessitates a change to the rules of business as usual.

This is, after all, exactly what telco executives are paid to do, thwart and exhaust the competition, and KPN shareholders should be very happy with recent developments. The beast is awake, thinking, calculating its own capacity for pain against the predators which threaten it, searching for a way to convince the attackers that pursuit is not worth the exhaustion. I would never fault anyone for fighting their corner, playing the hand they're dealt, etc., and KPN ranks top of the list in this regard.

I guess my concern is that the whole model is founded on very slippery ground. KPN (and every other EuroTelco) sprang to life in a carve-out from a post-and-telecom state monopoly, funded over decades by excessive call charges and taxes on citizens, who were then given the chance at IPO to buy back a piece of something they already owned. The stated purpose of this was to promote efficiency and competition, to the benefit of the consumer, and arguably it has worked to a certain extent. Fast forward to the present, and the same former state entity is looking to decommission its local exchanges and sell the associated real estate (i.e., assets ceded to it by the state in the first place) to recoup its expenditure in FTTC, a move which we must accept many of its competitors will not be able/willing to follow. Is this what the founding fathers/mothers of telecom privatization had in mind? Could they have even envisaged a world in which this was possible/necessary?

I don't have the answers - I am struggling with the issues, which is why I return to them so often. If I felt entirely comfortable with my position, I would move on to another issue which interested me. However, for the past few months I've been haunted by the possibility that the entire access model is flawed and that telco privatization will prove to have been a cataclysmic miscalculation. I don't have any beef with KPN - they are smart, they are working the system against competitors with their legacy cash flows, i.e., they are doing what any public company should be expected to do. The question for me is (and not only in relation to KPN, but more generally), is access to information something we are comfortable with as a pawn in a game of chess?

One week later and it's still open

Back from the trenches

I couldn't contain my mirth today on the KPN conference call when it became apparent that some analysts at big banks have clocked what FTTN really means for Dutch unbundlers.

Question: What's the regulator's view on the rights of the unbundlers?
Answer: We're having negotiations and discussions with LLU customers at present and OPTA is staying out of it for now.

Question: Is it possible that you might have to compensate them for (effectively) the loss of their plant?
Answer: That's something we'll discuss with them as part of the negotiations over our wholesale offer.

At no point was collocation ever mentioned. It seemed to be a foregone conclusion that in the all-IP world of FTTC in the Netherlands, KPN wholesale is the only game in town.

Baffling, but at least they're Orange

I've been getting spam from Orange for the past few weeks telling me about a special project and asking if I was interested in taking part in a secret location. Judging from the image attached to the latest spam update, it's around Lake Baikal or somewhere in Mongolia, and the 50 special people they have recruited to help them have erected gigantic orange plastic charicatures of the four "totem" figures they have come up with for their tariff plans: the cat, the racoon, the owl and the dolphin. I think I'm more of a bear. This is apparently in support of a new TV campaign which launches tonight in the UK. Hopefully this will also be accompanied by some sort of "the making of" reality progam, featuring 50 vapid, self-obsessed "yoofs" desperate to land showbiz careers.

From the Q2 trenches

Datapoints for anyone who keeps track of these things. KPN lost 253k consumer voice lines in Q2, that's annualized line loss in its consumer segment of 21.1%, versus 15.5% in Q1. The company also added 83k bundled VoDSL packages in the quarter, which interestingly is almost exactly the number of conventional call package customers it lost in the quarter (86k). But check out the consumer segment's EBITDA margin, down 4.3 percentage points over Q2 05, and down two points from Q1.