One day all telecom operators will have a native IP backbone. The big question is when and how to achieve this migration – technically and commercially. Accelerating to a profitable IP future is the objective of the i3 Forum, whose inner workings I witnessed up close and personal in June.

For those unfamiliar with this powerful organization, a quick précis: It is an independent group of 37 carriers serving the underlying connectivity needs of more than 1.5 billion consumers in 100 countries.

But it was with some misgiving that I travelled to Warsaw to attend the group’s inaugural Technical Workshop, where i3 Forum members and more than 20 representatives of the vendor community gathered to talk turkey about international IP interconnect issues.

After all, I’ve suggested in this blog and my column in Capacity Magazine that carriers failing to migrate to IP will like the dinosaur, soon face extinction.

Read the rest of this entry »

The best way for a telco to succeed in today’s market is to go insane. No really, I mean it. I’m not talking about the type of gibbering insanity that relegates sufferers to a locked room with a dose of Largactyl.

I’m talking about insanity of a different sort. Specifically, I’m advocating multiple personality disorder.

This is where an individual displays several distinct personalities, with unique ways of dealing with their environment.

You may recall the 1957 film, The Three Faces of Eve, starring Joanne Woodward. She portrayed Eve White, a downtrodden housewife, Eve Black, an extroverted socialite and Jane, a nice young woman, all existing in the same body.

The performance won Woodward an Oscar, and I believe that goodies are in store for telcos that can act out a similar tour de force

Read the rest of this entry »

The irrational exuberance of the Enron era led many communications service providers (CSPs) to financial ruin, and paying penance for past excesses consumed the first decade of the 21st century. But Yankee Group argues that it’s time to shed the hair shirt.

Despite a crowded competitive environment, new business models are emerging to lead CSPs into a vibrant future. However, CSPs must have the courage to transform. Hard decisions about what is core and non-core are required.

Earlier today I hosted a webinar which explored the ever-evolving CSP business model. Thanks to everyone who attended and brought their questions.

The webinar runs about an hour: audio (mp3) and slides (pdf).

Farewell to the Noughties, a decade sandwiched between two crises: The dotcom bust and the current – but sputtering – downturn.  In that time, Europe accomplished much: The Euro was adopted, DSL went mainstream and telcos went NGN.

Not least, consumers woke up to the pleasures of mobile content, although it’s questionable whether MNOs will ever see a fair return for their expensive 3G licenses. Roaming charge crackdowns and market saturation haven’t helped financials either.

Time again to put a nebudchadnezzar on ice? There’s plenty under the tree for 2010:

1. Ethernet will be everywhere. Ethernet is in the LAN, it’s in the WAN, it’s transforming mobile backhaul economics, and it’s converging the datacenter. Fiber remains best, but clever vendors (see Hatteras, Actelis) are delivering copper-bonded Ethernet in the first mile. And new Ethernet exchanges (see CENX and Equinix) aim to speed order to cash with their interconnect services. Want a unifying communications fabric? Well duh!

2. The CDN bubble will burst. Telco CDNs can offer compelling features, but how many service providers can the market sustain, even if video traffic is exploding? Many partnerships are already in place: Tata Communications with BitGravity, Verizon with Velocix, Deutsche Telekom with EdgeCast and Global Crossing with Limelight Networks and EdgeCast. If you’re not in the game now, you’ll need deep pockets to buy in.

3. The cloud’s hot air will expand. Resilient, liquid (and probably Ethernet-based) connectivity is going to save the outage-prone cloud. To invest in cloud services enterprises require robust network as well as applications-specific SLAs, as well as network redundancy, say Yankee Group enterprise surveys. Offering on-demand VPN connectivity to cloud services (on a wholesale or retail basis) could help defuse concerns about their security and resilience.

4. Equipment vendors will want to be your new best friend. The ratio of CAPEX to revenue currently stands at 12.6 percent among European operators, according to Yankee Group analysis. It’s not going to recover much. That’s why European equipment vendors like Alcatel Lucent, Ericsson and Nokia Siemens Networks are on a charm offensive with managed services propositions and aims to transform telco business models. Listen to their pitch. And talk to Huawei:  With a new SDP partner program and growing software division, it’s got more in its arsenal than cheap kit.

5. Smart wholesale will become sexier than dumb wholesale. Get big, get niche or get out. Embrace revenue-sharing models with non-traditional partners. And work mobile angles: International remittances, GRX to IPX interconnect, content transcoding, white-label mobile UC and M2M are among many rich avenues of investigation.

Best wishes for the New Year – and decade – look forward to continuing the conversation!

“What the hell is cloud computing?” After a year, those infamous words of Oracle CEO Larry Ellison still resonate. The definition of cloud computing is hazy at best, and many companies remain wary of the technology over concerns about infrastructure, security and regulation.

Cloud computing has unique potential to save the enterprise cost, reduce complexity and provide highly available service to the end-user or client. With such compelling benefits, companies should look to understand cloud better—what it is, what it isn’t and what it will be.

Earlier today, Agatha Poon and I hosted a webinar where we defined cloud computing, explored the capabilities and challenges of the technology and advised enterprises what cloud can do for them.

The webinar runs about an hour: audio (mp3) and slides (pdf).

Volcanoes didn’t erupt, nor did the seas boil, but it’s a Biblical event for the global telecom industry when a major operator announces that it’s pulling out of international voice. That just happened: In an estimated $1.5 billion deal, BT is outsourcing its international voice termination business to Tata Communications:

• Upward of 6 billion minutes are involved, according to my estimates: That adds neatly to Tata’s existing 24 billion;
• With 30 billion minutes under management, Tata tips the scales to become the world’s biggest international voice player;
• Tata handles BT’s international direct dial (IDD) and voice termination for all except a clutch of European countries, and becomes BT’s primary partner for UK IDD traffic.

A break with tradition. This is what I call ‘smart wholesale‘ for both BT and Tata Communications. As Yankee Group has predicted, telcos are breaking with tradition and remoulding their business model. Going forward, that may not involve direct management of commoditized services or networks.

Indeed, BT’s decision is not a sign of weakness in my view. Poor financial results have certainly triggered deep cost cutting across the group. But I believe this marks a bold assertion about BT’s strategic priorities. Managed ICT, cloud and sustainability services come to mind as prevailing core competences for the UK incumbent.

“BT wants to focus on the customer perspective; it doesn’t want to worry about how to manage the back end,” says Srinivasa Addepalli, senior VP of corporate strategy at Tata Communications. “But this is a core business for us.”

Wholesale is reborn. Bucking the recession, wholesale is on the uptick. Savvy wholesalers are set to grab a major proportion of the $145 billion that telcos will spend on outsourcing and managed services over the next five years. And BT is not the first to rethink international voice, it’s just the biggest:  KPN, Swisscom, TDC and Tele2 have already struck deals with wholesalers including Belgacom ICS, Deutsche Telekom ICSS and iBasis.

Let’s also not ignore the ongoing voice revolution in which Tata now has a forcible say. VoIP represents about a third of Tata’s originated voice traffic and almost half of the voice traffic on its backbone, and these proportions continue to rise.

Meanwhile, Tata is also eager to court mobile operators. It is much involved with industry groups seeking to improve voice and content interconnect across fixed and mobile networks. This includes the GSMA-backed IP Internetworking Alliance with its potentially disruptive IPX model.

Competition is reinvented. BT Global Services will continue to cross swords with Tata for consulting and managed services deals targeting communication service providers in other competence areas. Not least, BT will still compete with Tata in the multinational enterprise segment in India and elsewhere.

But so what? The key to success is choosing specific battles to fight, and selecting allies where most logical in order to focus on the core business, whatever that might be.

As Adepalli says, “Other telcos spending tens of millions on international voice are going to ask themselves: Why keep doing this?” Why indeed.

Yesterday’s outage of Amazon’s Elastic Compute Cloud - ironically, as I was giving a talk on cloud’s contractual perils – highlights the need to scrutinize the small print. Just as more organizations explore cloud opportunities, cloud risks may outweigh benefits.

Cloud contract terms, and crucially, service level agreements (SLAs) fail my tests for contractual risk. In my upcoming report, ‘Cloud 99.99,’ I’ll detail problems with specific contracts, suggest tools to mitigate risk and discuss how leading innovators are finding remedies. Here’s a taster:

1. What SLA? SLAs are not uniformly available, even from well-known brands, according to my vendor analysis, from Amazon to Zoho. Self-certification of ‘best efforts’ is common, but legally toothless.
2. Uptime is mediocre. Good luck finding 4, let alone 5 ’9′s (99.999% uptime). And even if you do, unlimited maintenance events won’t be in the calculations.
3. It’s not their problem. Besides demanding full indemnity for any damages (which is typically not reciprocal), get out clauses for liability include physical service demarcation points (eg: not beyond the server farm) and network congestion.
4. They won’t pay up. Forget liquidated damages. Penalties for non-performance are usually limited to contract termination or at best service credits.  These are gated in value, and usually, it’s up to you to argue the claim (within a specific timeframe).
5. Prolonged complaints aren’t welcome. Arbitration clauses are usually missing.
6. Foreigners beware. It’s not unusual for the contract’s governing law to reflect where a vendor is incorporated. That might be thousands of kilometers away from you (eg: California when you are in the U.K.), and alien to you in legal practices. But some contracts also specifically exclude the possibility of invoking the United Nations Convention on Contracts for the International Sale of Goods. This was purposefully designed to provide a common legal lingua franca in case of disputes involving parties in different jurisdictions.
7. Terminate at your peril. So, to hell with them. But how much time have you got to extract your data from the vendor? Miss the deadline and your valuable data will be expunged.

The good news? These issues only strengthen the case for a closer relationship between cloud vendors and communication service providers (CSPs). What’s been described as a parasitic dynamic fraught with rivalry is actually symbiotic. CSPs have the know-how that many cloud vendors lack, particularly in dealing with enterprises. Cloud vendors – particularly in the SaaS realm - are creating innovative productivity tools. As I’ll argue, each needs the other badly to monetize the cloud on a volume basis.

Not least, there is an urgent need for intermediaries to broker, aggregate, secure and monitor disparate cloud services – particularly as standards, service levels and interworking remain in flux (see Deutsche Telekom’s Zimory for a compelling example). In future, I believe that enterprises will pay trusted intermediaries to act as operational integrators and SLA managers across private, public and hybrid cloud contexts. Bottom line: The view’s great from the cloud, but don’t ignore today’s earthbound concerns.

Wholesale is often viewed as the butt end of the telecom industry. But signs of health are legion: The boom in new submarine cables, the growth of mobile and fixed traffic (both voice and data), and the global sprawl of carrier hotels and data centers. All good stuff for those reselling digital commodities into the telecom supply chain.

Of course, here I’m speaking of traditional wholesale: Capacity, minutes and space. Still, this is now a big boys’ game. Only control over monstrous volumes offers adequate margin protection.

When we last met, iBasis CEO Ofer Gneezy inadvertently reminded me of model Linda Evangelista. She famously stated that models don’t get out of bed for less than $10,000. Ofer claims that in international wholesale you need at least 5 billion minutes to be taken seriously and – crucially – be profitable.

He’s probably right – if you’re into dumb wholesale. But there’s a big difference between what’s smart and dumb in today’s vibrant wholesale market.

Read the rest of this entry »

For Geoffrey Chaucer‘s pilgrims, arriving in the cathedral city of Canterbury in Kent ended a journey through the garden of England. But for Professor Shyqyri Haxha, Canterbury marks the beginning of a trip to far wilder parts.

Until December 2008, Dr. Haxha was a lecturer in telecommunications at the Canterbury-based University of Kent. Today, he is CEO of Post and Telecommunications Kosovo (PTK).

So what strategy is the new executive pursuing in one of Europe’s poorest and youngest economies, I asked?

His number one goal: connecting all Kosovo’s schools to the Internet with fiber – and he’s aiming to do it before year end.

Surprisingly, Dr. Haxha already has the funds; PTK is cash rich. He hopes that an upcoming RFP for the estimated €80 million project draws qualified bidders unafraid to work in this corner of Europe.

Land-locked Kosovo remains an uncomfortable hangover of war in the Balkans. Under UN administration for a decade, this disputed region of 1.8 million inhabitants declared the Republic of Kosovo in February 2008.

That status is still not widely accepted. Indeed, neighboring Serbia views Kosovo as part of its sovereign territory.

Political limbo makes life tricky for Kosovo’s telecom operators. Kosovo still lacks a national dialling code from the ITU, the UN agency governing allocations.

Instead, Kosovo routes international traffic via Cable & Wireless-controlled Monaco Telecom. It’s an arrangement that doesn’t come cheap. Monaco Telecom keeps about 40% of the value of each telephony minute.

Then there’s the issue of piracy. Kosovo is dotted with clandestine cell sites where rogue operators are effectively stealing voice minutes.

Not just locally-generated, but also international roaming traffic is up for grabs: Half a million Albanians holiday in Kosovo every year (PTK hopes to win a piece of this action by taking a 30% stake in Albania’s 4th mobile license).

As a drawn-out NGN backbone upgrade with Alcatel Lucent concludes, PTK’s legitimate competition is hotting up. Telekom Slovenije has ploughed almost €200 million into newcomer Ipko Net.

Now the government is looking to privatize PTK – currently Kosovo’s most profitable company – but how this will be done is unclear.

So with all this on his plate, why is Dr. Haxha prioritizing fiber to schools? Dr. Haxha asserts that he’s serving his primary customer: More than half of Kosovo’s population is under 25. It’s Europe’s youngest.

But education is poorly resourced, and investing in infrastructure that can help upskill the young offers a very direct way to improve Kosovo’s social and economic conditions.

Nevertheless, the choice of fiber access is surprising, particularly since neighboring FYR Macedonia chose WiFi mesh for a similar USAID-funded project.

But Dr. Haxha is resolute: “Fiber offers a future for the long term.” He’s also sending 40 engineers for training at his former university.

In a prior Yankee Group feasibility study for a nationwide WiFi/WiMAX network in the Balkans, we concluded that device affordability (eg: laptop, smartphone) was a major stumbling block, irrespective of connectivity platform.

Equipment subsidization for individuals and groups is often a necessity, as Dianne Northfield and I argued in a report outlining the ‘motive ingredients’ needed to effect connectivity-driven social transformation. And perhaps Nicholas Negroponte could also help Kosovo, as per Yankee Group CEO Emily Green’s recent discussion with the One Laptop Per Child chairman.

Because his wife and children remain in the UK, Dr. Haxha still pays BT for a copper-based home broadband line. That gives him a personal stake in the current discussions about Digital Britain.

His view on the broadband investment case? “Any economy needs fiber if it’s going to speed ahead.”

I’m visiting the world’s third fastest broadband country after South Korea and Japan. That’s right, I’m in Romania.

Here, a government stimulus package did not create a high-speed broadband infrastructure. Instead, groups of tech-savvy neighbors just started stringing fiber optic cabling between buildings and trees. They became micro-ISPs by default.

Today, Romania’s 800-odd reƫea de bloc and reƫea de cartier (apartment block and neighborhood networks) are consolidating, but the patchwork of home-grown fiber remains.

Perhaps it is not pretty to look at, as these pictures suggest. But so what? Almost half of Romania’s existing fixed broadband subscribers now enjoy connectivity at speeds vastly exceeding 5Mbps, according Akamai traffic statistics. The problem is that broadband penetration doesn’t extend far from main urban areas.

Considering the Pirate Bay debacle, it’s also worth noting that Romania’s neighborhood fiber networks were built explicitly for peer-to-peer file sharing. That’s partly why many Romanian broadband packages are tiered by geographic access.

For about $15 a month you’ll get a local (neighborhood) bandwidth speed of 100Mbps, a metropolitan (city) speed of 50Mbps and an ‘Internet’ (national & international) speed of 10Mbps.

True, cable companies RCS&RDS and UPC and are trying to curb Romania’s considerable torrent habits. They are buying up the micro-ISPs and integrating mobile and TV propositions. But Romania’s unique genesis has bred a content-thirsty consumer irrespective of fixed or mobile platform.

This will drive a tripling of Romania’s broadband penetration between now and 2012, according to Yankee Group’s global Anywhere Index of broadband lines.

Let regulators try to prescribe how high-speed broadband infrastructures are built and managed. But Romania’s continuing journey proves that there is indeed no one route to – or from – Rome.