Wednesday, March 30, 2011

Google Fiber for Kansas City (Kansas), more towns to follow

Google Fiber has chosen Kansas City (Kansas) for its first (or: second) project, but the first Google Fiber for Communities (not be mistaken with nearby Kansas City, Missouri). Here are some details, and here are the specs, which are few, so far:
Google adds: "We’ll also be looking closely at ways to bring ultra high-speed Internet to other cities across the country."

Short look back at the entire project:
Looking forward, what is left open?
  • Technology: PON or Active Ethernet?
  • Business model: wholesale only? Passive assets only, or active network as well?
  • Service provision: who will sign up as service providers?
  • More generally: Google plans to find out the best way to do the job.
  • More cities: theoretically, another 350,000 people could be brought into the project.
  • Google also plans to find out what the effects (economic, social, etc.) will be.

Tuesday, March 29, 2011

With email newsletters like these, who needs spam?

One of the possible implications of the AT&T/DT deal could be that American telcos would become interested in acquiring European incumbents. Except, they won't. Because:
  • They are not very interested in cross-border deals to begin with. In the 20th century, companies such as SBC picked up strings of stakes in European telcos, but that was during the privatisation phase. Any foreign interest will most likely be focused on Latin America, Canada, or perhaps even Africa and CIS countries.
  • Regulation will deter them.
  • There will be more important investment cases, such as buying out DT from AT&T and Vodafone from VZW. But also FTTH and LTE.
You have to remember further that the T-Mobile sale from DT's perspective has everything to do with the April 2009 profit warning. In other words, it is a one-time event and will not set a new trend.

And so it is fun to read two contrary commentaries landing in my mailbox from respected consultancy firms. Who's going to teach whom a lesson?

Brand X:
Until quite recently, the US was generally seen as being somewhat backward in comparison with Europe when it came to mobile. Those days are definitely over. The US now "gets" mobile – in a big way. In developed economies, the growth in mobile is no longer being driven by telecoms, but by software and the internet. US companies have long been the dominant internet services brands (Google, Facebook, Amazon, etc), and the recent Nokia/Microsoft announcement put the cap on North American dominance of the market for smartphone platform software. Having shown Europeans how to win in internet services, and in smartphone software, perhaps the Americans' next lesson for Europeans may be how to win in consumer mobile.
Brand Y:
An evil person might say that T-Mobile employees that had the experience and knowledge about how T-Mobile had experienced massive competition in Germany and Holland from successful competitors might not want to share that information - because nobody wants to admit that they got thoroughly beaten up by their much smaller competitors! In conclusion: The American T-Mobile venture has been scrapped, those that ought to be held responsible for this sad turn of events will not be accused of anything and the shareholders will once again realise that they have invested in a company that despite all their experience and knowledge could not perform.
With email newsletters like these, who needs spam?

Sunday, March 27, 2011

How to value Skype at $3.5bn and LinkedIn at $7bn

We compare Google's valuation to Skype's and LinkedIn's. For Google, we look at market expectations, for Skype and LinkedIn we use the IPO-filings information. Our valuation metric is a variation of the PEG ratio, i.e. we compare price to some form of earnings (gross result and result after product development/R&D).

First Google. We see that the ratio of price (EV) to earnings (gross) to growth is around 0.33, while price (EV) to earnings (after R&D) to growth is 0.40.

Now, we look at Skype. To arrive at a similar valuation and assuming a value of USD 6 billion, you can see that we need to put in some very solid revenue growth (26%) and a limited product development cost rise (36%). If we use growth rates that appear to be more plausible, valuation needs to be brought down to USD 3.5 bn in order to match Google's valuation.

Finally LinkedIn. We look at two scenarios: in the first, we assume that growth will fall off from the extremely high 2010 levels. We then see that we need a USD 7 bn valuation to match Google's valuation. Alternatively, we assume a USD 2.4 bn valuation (which is what currently states) and play with the growth rates to match Google's valuation. What becomes apparent is that growth can be allowed to drop much further from the 2010 levels.

  • Skype: a USD 6 billion valuation appears to be a bit high; USD 3.5 bn seems to be more appropriate, based on the assumptions above.
  • LinkedIn: a USD 2.4 bn valuation seems to be quite low; the company could be aiming for something in the order of USD 7 billion, again: if the assumptions above hold.

Saturday, March 26, 2011

There's more sense to 4K than you might think

Last year, at the Mediapark Jaarcongres 2010, Frank Kresin talked about 4K. Afterwards, I asked him what kind of bandwidth requirements would go with it. It appeared to be: 8 Gb/s for uncompressed and just under 1 Gb/s for compressed material.

Last week, DisplaySearch published its TV forecast, stating: "We anticipate seeing the first product release by the end of 2011, with very small volumes in 2012". Herman has been doing the math, and concludes that 4K would require 240 Mb/s.

Whatever the case is (it all depends on the assumptions for the precise numbers of lines, pixels, frames per second and compression), the high-level conclusions are:
  • 4K is coming. And if you want to say: forget it, people hardly see the difference between SD and HD, then I would add: that is precisely why we need 4K! HD is fine, but we need a bigger distinction to SD.
  • 4K will benefit 3-D. To be sure: 4K is the evolution of HD, not 3-D. But, as DisplaySearch puts it: "There are other reasons to introduce higher resolution, even where it is not viewable. The most obvious is for passive 3D glasses. Doubling the number of lines is necessary to restore 1080 lines to each eye, and would overcome the main objection to passive 3D".

Tuesday, March 22, 2011

The UBB deception

Monday, March 21, 2011

DT could buy E-Plus for ... EUR 10 billion

Much of the logic of selling T-Mobile USA to AT&T, from Deutsche Telekom's perspective, is network roadmap: how to get to LTE, as a relatively small player? By selling, DT will be able to reduce debt (EUR 13bn), buy back more shares (EUR 5bn) and have a EUR 10bn left, albeit locked up in AT&T shares.

There must be parallels in many countries. Take [Germany].

T-Mobile USA [E-Plus] is having difficulties in financing the inevitable LTE upgrade. Hence, a long period of merger speculation around Sprint [O2]. Now however, T-Mobile USA [E-Plus] is sold to the the market leader, AT&T [Deutsche Telekom], for 7.1x 2010 EBITDA.

(OK, the comparison fails in two areas: there has never been a conflict of technology between E-Plus and O2, the way it exists between T-Mobile USA and Sprint. And Deutsche Telekom doesn't need E-Plus to get to nationwide coverage.)

So how much would E-Plus cost? At 7.1x 2010 EBITDA, the price would be close to ... EUR 10bn! (EUR 9.8bn, to be more precise, leaving EUR 200m for those poor starving bankers).

Sunday, March 20, 2011

Acquisition targets in the Netherlands

It's deal time. And what could that mean for the Netherlands? A short summary:
  • International operators could do a little portfolio management and decide they don't need an NL asset: Deutsche Telekom (perhaps increasingly likely now), Liberty Global, Vodafone and most of all Tele2, which requires from its operations an ability to be a Top 2 (hence the name) player. Of course business providers such as BT, Verizon, AT&T, Orange are in a different game, as are Colt, Easynet (private equity owned).
  • Private equity investments: Ziggo (possibly heading for an IPO) and CanalDigitaal (could be attractive to any challenger on the TV market: KPN, Tele2, possibly T-Mobile or Vodafone).
  • Cable providers: CIF focuses on passive network assets, but has a majority stake in CAIW. How about an IPO for this multi-MSO service provider? And then there are 20+ MSOs, in which CIF is presumably interested, but Ziggo and UPC as well. And perhaps even Reggefiber.
  • Fiber assets: OBR (Rotterdam) and LomboXnet (Utrecht) could be targets for Reggefiber or CIF. Reggefiber itself has only one way to go: to KPN.
  • Other: Scarlet (owned by Belgacom), Solcon (privately owned) and a long list of newcomers on the FTTH market.
KPN, the hoovering company, is a little unclear in its strategy right now. Under CEO Ad Scheepbouwer it has acquired someting like 40 companies (most importantly: Telfort, Getronics, Tiscali NL, iBasis). Recently, focus was moved from ISPs to MVNOs, but also Atlantic Telecom (business services) and NL-ix (Internet exchange). In the meantime, several assets were sold: fiber and business in Belgium, fiber in Germany, towers in the Netherlands. All quite helpful in reaching the free cash flow target.

So what could be next for KPN:
  • Sell E-Plus and Base, by the same logic that DT sells T-Mobile USA: there is a step change coming for the roll-out of LTE.
  • Sell more passive network assets. CIF is dying to buy them.
  • Sell Getronics. Sort of a u-turn, but perhaps focus is moving to network assets (but not passive assets).
  • Buy WiFi assets.
  • Buy more MVNOs, esp. those focused on the business market.
  • Buy CanalDigitaal (see above).
  • Buy out Reggeborgh from Reggefiber.
To round off: one type of asset is out of reach of KPN: cooperatives. This could be the way forward on the FTTH market, but in light of the above (NL-ix), it implies that Ams-IX may be a desirable target, but cannot be folded into KPN.

AT&T buys T-Mobile USA: will the deal survive FCC scrutiny unscathed?

AT&T buys T-Mobile USA for $39bn, o/w $25bn in cash (may be raised by $4.2bn, as long as DT's stake is at least 5%) and $14bn in shares (for an 8% stake). The EUR equivalent is 28bn, of which EUR 13bn is for debt reduction (31%) and EUR 5bn for extra share buy-backs. The price implies a valuation of 7.1x adjusted 2010 EBITDA.

Deutsche Telekom further refers to its continued exposure to the US market and the attractive AT&T dividend. AT&T defends the deal by referring to the extra spectrum and the increased ability to blanket the US. They also refer to an 'impending spectrum exhaust'. And they are happy to report that T-Mobile USA will be 'part of a US-based company'. AT&T expects a synergy run-rate of >$3bn from 3 years after closing, total synergies will exceed the purchase price

Here are some first thoughts, some off of Twitter, for which a hat-tip to Dean, Keith and Brough:
  • What will DT do with the proceeds, i.e. the remaining EUR 10bn (locked up in AT&T stock for 1 year after closing)?
  • What does this imply for other markets, particularly those where DT offers mobile services only, such as the UK?
  • If 4G/LTE was a dealbreaker for T-Mobile USA, what does this imply for other companies still undecided on their 4G roadmap, such as E-Plus (KPN) in Germany?
  • Bad news for LightSquared, which is bound to lose a wholesale customer.
  • What will Sprint do?
  • How will the DoJ and FCC respond?
  • AT&T may have to give up spectrum.
  • If AT&T pulls this off, Verizon Wireless will be enabled to make further acquisitions as well. Which would be bad news for Vodafone: no long-awaited dividend re-installment.
  • What would the break-up fee be? - UPDATE: $3bn + some spectrum + a roaming deal.
  • Integration may lead to bad service for the next 12 months.
  • Competion will be reduced, the market may develop oligopolistic traits.

Tuesday, March 15, 2011

Connected TV 2011 Conference, April 27, Utrecht

Our Connected TV 2011 conference is taking shape. Here is the latest speaker line-up, with some buzz words for what they will be speaking about:

Content, DRM, standards:
  • DLA Piper: DECE/UltraViolet
  • NPO: HbbTV
  • Chello DMC
  • Irdeto: ActiveCloak for Media
  • Reggefiber
End-to-end (back-office) systems:
  • Nokia Siemens Networks: Ubiquity Multiscreen TV Platform
  • Alcatel-Lucent: Mediaroom
  • Cisco: Videoscape
Client-side solutions:
  • Amino: Freedom, tier 1 euro telco
  • Technicolor: MediaEncore
  • Intel: Groveland
  • Aprico
Real-world cases:
  • Oregan: Onyx, Telefonica
  • Philips: Net TV
  • Wyplay: SFR, Vodafone Spain
  • Vodafone Spain

Where there's a cap, there's no competition

AT&T plans usage caps from May 2, 2011 for its DSL and U-verse subs: 150 or 250 GB/month, and an overage fee of 20c per GB. Similar policies are installed by Canadian ISPs (but with much more expensive overage fees), whereas BT is abolishing caps/FUP.

Dave Burstein was quick to point out that installing caps has little to do with cost (see previous post: The Incumbents' solution is in search of a problem). So which are the true reasons?
  • Make an extra buck. There's nothing wrong with that, except that AT&T is not being fair about why they are doing this.
  • Fight OTT. This write-up reveals that IPTV (U-verse) usage does not count toward the cap. By installing a cap, users may be led to think twice before using Netflix or Google TV. If you are used to leaving the TV on all day, this turns into a real problem.
  • Insufficient competition. In a competitive market, there would be little room for installing caps just to make an extra buck (see above). Put bluntly: caps prove that competition is insufficient.

Sunday, March 06, 2011

The Incumbents' solution is in search of a problem

Deutsche Telekom, France Télécom, Telecom Italia and Telefónica (hereafter The Incumbents) have issued a report called ‘A Viable Future Model for the Internet’. It reflects their well known stance on the “they are using my pipes for free” case, but it also offers a thorough review of options for how the Internet could be financed. O, and it was published by A.T. Kearney (which I hope earned a little bit less than 25 million Ozzie dollars for this report). While being a thorough report and an interesting read, there are some passages that we have a problem with. For inquires you can try to mail to

First of all, here are the ‘structural problems’ that are perceived by The Incumbents:
  • ISPs (called Retail Connectivity Providers in the report) have not been able to monetize traffic growth. They are the victims of the very popular flat rate Internet. There is a disconnect between traffic growth and revenue growth. “… for growth to continue, the next few years must see a realignment of who captures the value and who funds investment in the Internet value chain” (page 1, a nice twist to Ed Whitacre’s above quote). It is also stated that “video traffic , much of it free to the end user” (page 1), while the end user pays a lot of money to his ISP.
  • OTT providers (called Online Service Providers in the report), who hardly contribute to the cost of the network, are not incentivized to use the Internet efficiently.
  • Network congestion is inevitable.
Second, the four possible remedies proposed:
  1. Raise retail prices.
  2. Introduce a data-conveyance charge for OTT providers.
  3. Enhanced quality services over the public Internet.
  4. Enhanced quality services based on bilateral agreements.
It becomes immediately clear from the above how The Incumbents position themselves in this issue. Fundamentally, they look upon OTT providers as publishers who have found an extremely cheap distribution platform, while The Incumbents pay for the networks. At the same time, OTT providers earn insane margins. The report also stands up for rights owners and speaks out against piracy, implicitly linking heavy usage to illegal file-sharing.

As the report discusses option #1, it finds an important objection against usage-based tariffs: “the problem is that end users often do not have full control or awareness of the actual traffic they are downloading. Software updates tend to download in the background automatically; animated adverts are not actively requested …” (page 26). This we agree with. Also, the writers think that price increases are not achievable, due to competition, and if they were, penetration and the uptake of new services could be negatively affected. This, we do not agree with.

As to option #2, the proposed charges are EUR 0.05 per GB for fixed and EUR 3.03 per GB for mobile networks. It is suggested that OTT providers could simply “… adjust their commercial models to cover the higher costs – doing so for example, with higher advertising fees or subscriptions ” (page 30). This is of course totally unrealistic: getting rid of the flat fee issue simply by pushing it to Google, saying that they should raise their rates. Check out Benoit’s post to see how this can’t work (or would they have their charges off by mistake by a factor of 10 or 100?).

Options 3 and 4 are about creating ‘Internet fast lanes’, or a ‘two-class Internet’, which has all sorts of net neutrality implications – which we do not need to replicate here.

Here are a few problems we have with the report:
  • The report is based on the assumption that congestion is inevitable. However, despite incredible growth, congestion hasn’t really been an issue yet. In other words, the solutions provided are in search of a problem. As Telegeography shows us periodically, capacity investments are ongoing and often outpace traffic growth. Also, the report appears to be sloppy in many places about which part of the Internet it focuses on: is it the long haul, the middle mile, backhaul or the access network?
  • A big deal is made out of using ‘the network’ efficiently, but if we are not mistaken, it becomes clear in just one place what this actually means: “Free video sites in general may find the conveyance charges are a strong incentive to develop more efficient traffic delivery (better compression techniques)”. So, one crucial and ‘structural problem’ is just about compression? Also, it seems entirely unconvincing that OTT providers would not care about compression.
  • The report fails to show that the linear value chain (figure 3) needs repair. In fact, it has served the Internet very well, with traffic flowing from left to right and money in the other direction. We also wonder why the Sony or the Philips analogy is not addressed, which so neatly nails the problem with solution #2 (above). (It must be acknowledged that the nature of OTT service provision weakens the analogy, because in the electricity world, there is no analogue to the OTT provider.) One could say that it is The Incumbents’ good right to try making OTT providers pay, but it would be at odds with how the Internet is organized today. Further, they could use their market power to bully these charges onto OTT providers, but that would appear to be sheer abuse of market power.
  • OTT providers (aggregators, not producers) pay a lot to distribute the content to servers all over the world.
  • OTT providers are described as the ones generating traffic ('traffic senders'), completely ignoring the fact that it is the end user who is in fact requesting the traffic. Put differently, without OTT providers, nobody would need a broadband connection in the first place. The Incumbents owe their entire (very lucrative) broadband income stream to the OTT providers.
  • Investors know that looking at margins alone doesn’t provide the whole picture. Some sectors are characterized by high margins, others by low margins. It depends on the nature of the business, which involves volumes as well.
  • The Incumbents are ignoring a minor detail: risk. For the non-investors: risk has nothing to do with our everyday-life use of the word; in investment theory, risk is about the standard deviation of returns. Just that, nothing more. Now take a look at Apple’s revenue growth and compare that to the extremely stable revenue profile of network operators. Remember further that The Incumbents are vertically integrated, with completely different risk/return profiles for the different layers (passive, active, services). Which is why there is increasing focus among newcomers on just one of the layers. OTT providers are no exception, even though they (Google in particular) do have extensive network assets.
  • If ISPs lack pricing power, sooner or later the market will sort this out. Many ISPs would be bound to go out of business. The fact that they don’t proves that broadband is still a high-margin (!) business. If ISPs complain about a disconnect between traffic growth and revenue growth, their natural response should be to raise prices, not turn around and look to the other end of the vale chain to extract money from.

Friday, March 04, 2011

Another worthless cable/DSL comparison

NLkabel, the Dutch cable lobby group, ordered a study from iPing Research into speed and pricing of cable versus ADSL. Of course, they claim that cable beats ADSL and that it is cheaper too.

On the face of it, it looks like a completely worthless exercise - which can be expected from a lobby group. This is where it goes wrong:
  • Cable and ADSL are compared. That should have been either Docsis x vs. xDSL (i.e. equipment), or HFC vs. copper (i.e. medium). And then there are differences in loop lengths to further complicate a solid comparison.
  • Broadband is available is speed tiers. What does 'average' mean? Exactly nothing.
  • When you look at actual speeds, it will be no surprise that people requiring the highest bandwidths take a cable subscription, because the fastest cable tier is faster than the fastest ADSL tier.
  • Current DSL research points to maximum speeds of 800+ Mb/s, beating any available Docsis 3 based offering. (Why not add a worthless claim ourselves, comparing lab conditions to actual offers?)
  • Why compare cable to ADSL and VDSL, while leaving out FTTH? Anyone?
  • Broadband is offered as part of a double or triple play, so pricing is not available in many cases.
Of course, NLkabel continues its fiber claims, while speed is mainly determined by bottlenecks (and they have a 300 meter bottleneck, not counting a handful of amplifiers). There is also a claim that 50-120 Mb/s over HFC is available to 98%, which is a fluid claim as well. Hell, 1 Gb/s is avaliable to 100% of the people - just need to do some digging!

I wouldn't be surpised if research shows that men drive their cars faster then women. I also wouldn't be surprised to see that men tend to buy BMWs, while women buy a Mercedes. But does that mean that a BMW is faster then a Mercedes? Or that 'the average BMW' is faster than 'the average Mercedes'?

Some people need a lesson in logic and reasoning.

Tuesday, March 01, 2011

PhotoRocket: the Facebook killer?

We are not fans of (the photo sharing site) Facebook. But what is a social network anyway, other than a set of personal homepages with communication between them? So why do we need a social network destination site at all?

This review of PhotoRocket may point the way in moving beyond traditional (2010 style) social networking. PhotoRocket is for sharing too, but it's not meant to be a destination site. It's simply an app. With some imagination, that could liberate us from advertiser traps such as Facebook.