Amsterdam: FTTH penetration at 35%

When Amsterdam, KPN and Reggefiber recently announced that they were close to entering Phase 2 of the munifiber build-out, they were somewhat secretive about the number of homes activated. A local journalist pulled it out of one of the managers: 3k, on a total of 43k, which makes a penetration rate of 7%. Since work started 3 years ago, that was a less than impressive number.

However, things need a little clarification, as an insider informs me. First, hooking up was made available just 2 years ago. Second, yes there have been all sorts of everyday problems in rolling out. Third, Amsterdam not only counts homes passed (43k) and homes activated (3k), but homes connected too. The latter number stands at 9k, for a penetration rate of 35%.

What a little PR can do.

Still, the number of homes passed would still be of interest to me. I guess we need some 'performance measurement' standards when it comes to measuring penetration rates.

Lastly, the FTTH Council numbers last week also build upon the number of homes passed. See this IDATE information to find out some detail on the number of subscribers (= homes activated). The Netherlands stand at 350k homes passed and 29% (101.5k) of that as homes activated.

The case for FTTH - once more

In defending FTTH, one runs up against two problems: price (it's expensive) and speed (do we need all that?). Of course these are interconnected (if demand is high enough, pricing can go down and FTTH will be more successful and the business model works).

Skeptics abound, especially among the cable industry, so it takes tirelessly summing up advantages of FTTH (viz-a-viz cable). Here's another (concise) effort.

• Growth. At current growth rates, access networks will soon run out of capacity.
• Video. The internet was originally designed with data in mind. Soon graphics came along, and now video is all over the place. All the e- and tele- applications of the world probably have a video component, often for upstream too (symmetrical, that is). Of course, there is IPTV too. Plus: video evolves toward HD, 3-D and holography.
  
• Apps. New applications will come as FTTH spreads, pretty much the way electrical instruments were developed when the electric grid became universally available. Check out this Alcatel-Lucent release (on the ng Connect Program) and this one from Aepona, to see how application development and how to bring them to end users are hot today. What it all means? Telcos need to partner if they want reap new revenue steams.
• File sharing. This goes beyond illegal P2P. Some examples that make files go up, down, left and right through the network: file sharing (legal versions), place-shifting (Sling), cloud computing, storage and online backing-up.
• Femtocells. These in-home miniature base stations backhaul mobile traffic over the user's BB connection (freeing up both outdoor base station capacity and mobile operator backhaul capacity). They will become an essential part of 4G networks operating in higher frequency bands (2.6 or 3.5 GHz) and put a strain on the subscriber's BB connection.
  
• Copper doesn't cut it. Copper networks were designed with just voice in mind. DSL did wonders, but is running out of steam now.
• Timing. Even if we don't need it today, we certainly need it in 10 years time. Build-out takes long, so we better start today.
• GDP grab. If you don't build it, somebody else will. Cities and entire nations are already competing to build out FTTH networks in order to grab some extra GDP growth. Hence, broadband stims are finding their way to the market.
  
• Opex. FTTH is cheaper to service than copper.
• Green. Redundant FTTH networks save on travel expenses, fuel, carbon emissions, etc.
  
• Redundancy. We need excess capacity because we hate waiting. We also need low latency for fancy apps such as gaming.
  
• Moving. Perhaps not everybody needs FTTH, but people move - on average, once every 7 years (could be a bit slower today). So, not every body but certainly every home needs FTTH. Also, even if only families (roughly one third of all homes) are the addressable market, one has to realise that singles (one third) and couples (one third) tend to morph into families.
  
• End-game. Not just copper owners switch to fiber, alternative carriers and utility companies (with no legacy to worry about) also choose FTTH - not coax, mind you. Even cable execs acknowledge that FTTH is the end game; and those that do not, call their Docsis 3.0 service 'Fiber Power'.

(Who can switch the order in such a way that it makes a nice acrostic and the red letters connect to a relevant term?)

Skype sale limited by Silicon Valley clusters

Skype was probably rumoured to be for sale the day after eBay bought it. Recently, at the Q4 analyst meeting, new speculation found furtile soil. I added a little myself.

There is an interesting case for incumbents to take a look at Skype (which neither is a very new idea). Not only would they buy back some lost revenues, but they would also expand internationally (including emerging markets, where they all want to focus their investments at this point - talk about following the crowd). I particularly like the option to buy an international brand name that could be extended to include other services (such as operating the services and the active layer in FTTH networks throughout Europe).

A very valued reader pointed me to a limitation regarding Skype's sale. It has to do with how Silicon Valley seems to work. Apparently, development activity in Silicon Valley is grouped together in a number of clusters, and the sale of proprietary technology that Skype sits on may be limited to companies within the cluster. It could possibly even be a little more pronounced than that: the technology (check out this 1998 company and see what has become of it) serves as the basis for VC groups to set up a range companies. In Skype's case: the Joltid Global Index sofware was first used for Kazaa, then for Skype and next for Joost.

I'm not quite sure how these clusters are formed, but supposedly there is one around Redmond (Microsoft - don't mind the exact geo location), one around Palo Alto (Sun) and one around Stanford (Google, Yahoo!, eBay). This seems to preclude a sale to anything like a European incumbent. But how about KPN - doesn't its iBasis subsidiary have a deal for SkypeOut?

Vodafone and 3 Australia: why no network sharing?

Vodafone Australia and 3 Australia are merging into a 50/50 joint venture VHA, apparently to better compete with Telstra and Optus.

Three comments:

• Mobile markets worldwide seem to disagree on the number of players required for just the right amount of competition. In Germany, calls have been heard for years for a merger of the #3 (E-Plus) and #4 (O2) operators to merge. In the Netherlands, the market went from 5 to 3 players, but now the competitive level seems to have dropped below the ideal level; a fourth player could enter once more, once 4G spectrum is auctioned off. France is a comparable case.
  
• Why not do a full network sharing deal, for both passive and active (RAN) elements? Vodafone and 3 would vastly improve their cost base and could still market their own products and services. Vodafone has a solid position in the prepaid market, 3 has more data-centric postpaid subs. The venture is going with the Vodafone brand, so 3 is effectively getting out of the market, in exchange for a 50% share. Migrating 3's subs to Vodafone may cause an incredible headache.
• Are more deals in the making? Rumours have been going around about 3 giving up its position in Italy and the UK. Vodafone comfortably has a position in those countries as well. Australia could be a testing ground for 3 to quietly get out of the mobile business altogether.

There is more to FTTH than triple play

This summary is not available. Please click here to view the post.

New Research Brief: Skype for sale?

Today my employer publishes a Research Brief on Skype. It elegantly counts 7 pages, 7 paragraphs and 7 tables.

Paragraphs:

1. Skype may be up for sale
2. Skype is an established VoIP player, with an extensive product range and ecosystem, but is not a primary line service
3. Usage growth outstrips user growth, but revenue growth lags behind
4. Revenue growth disappointed from 2006, but the margin has reached an estimated 17%
5. The network effect delivers high growth, but revenues from SkypeOut could be in danger
6. European incumbents could expand internationally on the wings of Skype
7. Google-like growth would imply a USD 600-700 million valuation
   

Tables:

1. Skype products
2. Selected embedded technology partners
3. Selected distribution partners
4. Users, usage and year-on-year growth rates
5. Revenue, year-on-year growth rate and usage
6. Short P&L
7. 2008 Results, market capitalization, total enterprise value and multiples

Open access: If you can't beat them, join them

Paul Budde has the 'official' news: Sol Trujillo is out at Telstra. Another big win for the open access movement - hopefully (it remains to be seen who will replace him).
Paul rightfully refers to KPN's open access strategy, but let's not forget that it was not just NGN thinking that forced it upon them. KPN must have felt pressured by the loss of fixed lines (down 50% to 2.5m in just 3 years), heavy competition from cablecos (who have near 100% coverage), and last but not least: Reggefiber's stealth advance in the markt. In FTTH towns such as Nuenen, cable and KPN are almost extinct.

Some thoughts on KPN's vertical integration

Amsterdam finally published the Phase 2 plans, with the KPN/Reggefiber joint-venture, for its munifiber Citynet network. Phase 1 included 43k homes. Some notes:

• No timescale for the 100k homes, let alone the final 250k ones.
• I asked Ad Scheepbouwer about his views on KPN's stake in the passive layer. It has an option to acquire a majority in the Reggefiber joint-venture, but it could also allow it to dilute by raising third-party funding and ultimately get out of the infrastructure business altogether. Alas, Ad keeps his options open at this time. Reggefiber, for that matter, has no ambitions whatsoever to become a service provider (they do have a stake in the XMS joint venture with BBned).
• I liked mayor Job Cohen's referral to earlier PPP projects: the Amsterdam banking industry (400 years ago), the North Sea - Amsterdam channel, Schiphol airport and the wildly interesting AMS-IX.

KPN will also be operating the active layer (BBned is the active operator on the Phase 1 section), and be service provider as well. In other words, it will be a vertically integrated operator/provider once more. No exclusivity in the active layer, though (which BBned has). Some more notes:

• It remains to be seen if it works. The regulator should monitor KPN's actions (pricing) very closely.
  
• Will only SPs compete, or will there be alternative active operators as well? In the Phase 1 areas, competition is limited to SPs, because of BBned's exclusivity. However, the number has gone down drastically, and two (Alice and InterNLnet) are owned by BBned itself. Is there no interest in competing in the services layer only?
• Competing in the active layer is a lot more expensive, but I believe this is equivalent to LLU investments in the DSL world (cutting out the wholesale payments to KPN to improve the business case). This implies two things: 1. You need scale. 2. Since FTTH is the end game, there is no risk of becoming obsolote (which has happened to LLU operators, who were subsequently hoovered up by KPN). In other words: a new entrant, operating the active layer (and bringing its own SP) could alter the Dutch landscape dramatically. I see a big opportunity for other incumbents to first buy BBned (it's for sale!) and then expand.
  
• Obviously, some things stand in the way of operators expanding into the Netherlands: 1. The global financial crisis, 2. focus on expansion into emerging markets.
• Still, there is also a case for European incumbents to enter the Netherlands as service providers only: find some extra growth from an infrastructure-light approach, get experienced in an open access FTTH environment, hurt your competitor (KPN), etc. Now, of course you would ideally need a good brand name to make such a move. KPN is expanding Simyo (its MVNO) across Europe, which could serve as a mobile equivalent. How about Skype (it seems to be for sale ...)? It could be used as a vehicle for any incumbent and subsequently be enhanced with new products, turn it into a pan-European SP (and next get into the active layer).
  

Mysteries around KPN's FTTH trials disolved

Here's a short follow-up to some questions around KPN's success in an early FTTH deployment, mentioned in my previous post.

• ARPU = EUR 58 seemed low relative to the 65, 80 and 110 triple play packages KPN offers. The answer to the mystery is embarrassingly simple: ARPU numbers exclude VAT.
• 2x5 towns refers not only to 5 FTTH projects, but to 5 FTTC (+ VDSL) trial projects as well, as mentioned in the Q3 report. I'm quite confident that KPN won't go with FTTC.
• 16% penetration refers to the number of activated homes as a percentage of the homes connected (which is a number not much below the homes passed number). Here it becomes apparent that KPN has a much easier up-sell than Reggefiber. KPN can simply offer a migration with very little risk of service interruption. Reggefiber, as a stand-alone unit, has to compete with existing offerings from KPN and cable. Hence, Reggefiber's strategy to assure itself of some 40% sign-ups in advance.
  

Rotterdam Fiber Lab: KPN, Daily Media

Today I attended Rotterdam Fiber Lab 2009, hosted by OBR (the Rotterdam city department in charge of the local munifiber network, and other assets) and R-IX (the Rotterdam Internet Exchange). Speakers included Ed Vermeulen (director of KPN Glas, the service provider on KPN’s FTTH networks), Hans Appel (CTO at Sun) and Gert Suur (CEO at Daily Media, the media center company). Here are some highlights, with comments. Make no mistake, Ed Vemeulen’s statements need a lot of clarification. Read all about it at my employer’s site.

Ed Vermeulen, KPN:

• Almere has 230 homes connected. COMMENTS: We hear about delays all the time, in Almere, Amsterdam and Rotterdam. These reports remain unconfirmed, however. Earlier this week, OBR said (in a local paper) that it plans to pass 60k more homes in Rotterdam in 2009. They should do better marketing, because for as far as I know, this is double the amount they mentioned last summer.
• In a different town, probably smaller but not mentioned by name, KPN started the roll-out of FTTH 8 weeks ago. It has reached a penetration rate of 16% and an ARPU of EUR 58. Line loss was reverted to a gain of 5% (instead of the national KPN average of -2%). Almost 100% of the target group now has single access share. COMMENTS: Is the penetration a % of homes passed or of the entire town? People need to take the full triple play, costing 65, 80 or 110 EUR/mo, so how they arrive at 58 is a mystery for now (promotions?). Also, no wonder the target group has a ‘single access share’ of nearly 100% (meaning: they cancelled other subscriptions, such as ADSL and cable?).
• As usual, FTTH (dedicated, max in the multiple Gb/s) was promoted relative to cable (shared, max at 400-800 Mb/s). COMMENT: In the end, every network is shared, because traffic is assembled at a certain point anyway. FTTH is the end-game to most, except for NLkabel (they believe coax is comparable for the local loop) and possibly John Cioffi (the DSL godfather). Everybody else, including many cablecos, acknowledge FTTH is the end-game.
• The KPN strategy has 3 steps: 1. Get ready (2008), 2. Trials in 2 x 5 towns (2009), 3. Decision on nationwide roll-out (June/July 2009). COMMENTS: So far, KPN said they were building in 5 towns, now the number miraculously seems to have doubled. Ed Vermeulen appears to be a true FTTH aficionado. He showed a very cool promo, that I suppose is used at demo sessions in the towns where KPN is rolling out. Only one thing may delay KPN’s nationwide FTTH roll-out: the sale of its MDF locations (as part of the All-IP project); the proceeds are meant to finance the FTTH roll-out.
• KPN strategy per area: 1. Awareness (promotion), 2. Physical presence, 3. Frontline activation (roll-out). They distinguish residentials (families; elderly; singles & couples; youth) and city stakeholders (sports clubs; schools; banks; libraries; councilors). COMMENT: I suppose the relevance of the latter is to attract anchor tenants, which helps a lot when trying to make the business model work.
• Q (from the audience): What happened to KPN; they promised so much over the past few years but delivered so little? Are they re-monopolizing? A: I can only speak for myself. The network is completely open, we invite everybody to provide services. What counts, is the utility rate. COMMENTS: KPN needs to be more clear about the 3-layer model. Yes, they will be a monopolist on the passive layer (apart from the omnipresent cable network), but as long as regulation results in decent wholesale tariffs: who cares? Why don’t cable companies start providing services over KPN FTTH, instead of complain all the time over KPN’s abuse of market power? Generally, cable suffers a very bad image in the consumer market. KPN however has an equally bad image in the professional market: slow; indecisive; unreliable, etc. As to the utility rate: this is the first time I heard a fiberguy explicitly mention the term. I suppose he reads this blog.

Hans Appel, CTO:

• Young people have 4 screens to the world (not the usual 3): mobile device, desktop, TV and car.
• Unfortunately, the Woodstock generation (baby boomers) is running the world. They should take a look at how Generation Y (millennials) live.
• Recommended book: Thomas L. Friedman – Hot, Flat and Crowded.
• Sensors are Appel’s pets, his views are developed in CENSI. Sort of M2M communications to make this world a better place. Green isn’t just about green datacenters, please!

Gert Suur, Daily Media:

• I wrote about them several times before, so here are just the interesting bits.
• Essential to the business model and big differentiator to similar media center boxes: free to the consumer; no subscription fees; independent from any cable company or IPTV provider; extreme ease-of-use, including single sign-on and PIN for VoD rentals; content providers (also very small ones) can interact directly with consumers and don’t need a cableco (it’s a platform!); 1.5 Mb/s connection suffices (and 8 Mb/s for HD). COMMENTS: the Netherlands were a box-less nation for a long time. The advent of digital TV changed this, over the last few years. Now however, people may end up with a multitude of boxes: a STB from their cableco, a media center and (next year or so) a femtocell. Why not integrate all that hardware, and use it for new services? I think Daily Media should pitch this idea at box manufacturers, including Cisco, Tandberg and Ubiquisys. Also: cablecos very unjustly fear VoD income cannibalization; they control the margin themselves, Daily Media doesn’t necessarily make a difference.
• No hard disk, so no piracy issues (no room to store any downloaded content), no security issues, very fast re-start should the box crash (which it actually did during the demo).
• Innovation: a PC-based version sans box will be coming; Traffic Radio (started by entrepreneur Erik de Zwart) has launched, but is not available on cable; direct sales to consumers could start to happen once the box has HDMI connectivity. COMMENT: I strongly believe they should do the latter, taking their future into their own hands. Several people in the audience were interested in getting one, forcing Mr. Suur to reply: “It’s not for sale.” How do you like that?!?