Pile them high, sell them cheap

Despite covering only 20 per cent of households in Denmark, Clearwire affiliate Danske Telecom claimed a sixth of all new broadband subscribers in 1H07. The key is a strategy that owes more to supermarkets and budget airlines than the telecoms trade.

Danske Telecom is a joint venture between Danish investment fund P/S BI IT Venture, which owns about 54 per cent of the company, US wireless broadband pioneer Clearwire, which has 38 per cent of the shares, and CEO Peter Jerry Sorensen who owns the remaining eight per cent.

The company has large spectrum assets: two 3.5GHz licences of 2x28MHz paired FDD spectrum, 106MHz in total. One of those licences is nationwide, the other covers about 80 per cent of Denmark. Danske also has 3.6GHz, 10.5GHz and 26GHz spectrum licences.

It also operates its own nationwide fibre and microwave network that provides backhaul capabilities for its wireless subscribers as well as serving a separate fixed-line business division that offers 100Mbps symmetrical services to enterprises.

Danske launched wireless broadband services in its 3.5GHz airwaves in Denmark’s three largest cities- Copenhagen, Aarhus and Odense-in October 2005, using Motorola’s pre-WiMAX Expedience solution and operating under the Clearwire brand.

Since then it has expanded to four more of Denmark’s largest cities and, as of mid-February 2008, has an installed network of 212 sectors on 70 sites, covering about 20 per cent of Danish households, equating to roughly 40 per cent of the population (about 1.5 million people).

Despite having a footprint restricted to the urban areas of fiercest competition, in a market that tops the OECD charts for broadband penetration (34.3 per cent in 2007), Danske has held its own, claiming a broadband market share of just under five per cent in the areas where it offers services and where incumbent TDC, TeliaSonera and Telenor offer DSL.

That share leaps when measuring new additions: in the six months to July 30 2007 Danske claimed 16.5 per cent of all new broadband subscribers in its territories. As the only WiMAX player of any scale, Danske Telecom sits on 94 per cent of the WiMAX market in Denmark.

Danske had shipped 19,766 subscriber modems by mid-February and, for the six months to the end of 2007, turned EBITDA positive, just two years after launch, an impressive performance, particularly as Danske has already made the bulk of its investment in infrastructure, spending about $50m to date.

This steady progress has been built on a discount telecoms model that emphasizes cheap service plans, simple plug-and-play CPE, minimal customer acquisition expenditure, opening wide the network capacity, and packing the customers in.

No Frills

“We have a discount strategy like Ryanair or Easyjet,” explains Danske’s CEO Peter Jerry Sorensen. “We have the best price-per-Mbps propositions for consumers: 500kbps download and 192kbps upload for 59 Krone (E9.25) per month; 1Mbps for 139kr (E18.66); 1.5Mbps for 169kr (E22.7).”

Business customers pay more, but it is from the consumer market that Danske gleans 90 per cent of its volume and 80 per cent of its revenues.

When it investigated the reasons for customer churn, Danske found that people on the highest price plans churn three times as more as people on the lowest price plans. “The key parameter is price,” says Sorensen.

He argues that only 15 per cent of total telecoms revenues are ‘discount’, while the equivalent share of commercial aviation or supermarket revenues is more like 45-50 per cent.

There is therefore plenty of room for discount operators to expand, and so Danske has tackled big cities, undercutting DSL operators by offering cheap, no frills broadband access and voice services.

“We have our own nationwide fibre network, and our own access network, and we build on that a service network and VoIP, which we provide through third parties,” Sorensen explains. “We can put together a Surf and Talk combined package. Other services the customer decides. We don’t do TV packages or triple-play.”

Ninety-eight per cent of Danske customers purchase their modems and service packages online, many having been channelled to the site through targeted media campaigns. “The combination we think works best with the lowest cost is print, radio and Google,” says Sorensen.

The marketing spend is balanced by savings made elsewhere.

“We have no sales people. One of the secrets of our success, especially our EBITDA success, is our low cost of operations,” explains Sorensen. Just 22 people manage Danske’s WiMAX operations, just over one person per thousand subscriber modems shipped.

Those modems are plug-and-play models that enable consumers to self-install and are, says Danske, delivered on the same day they are ordered. They are also portable. Sorensen attributes much of Danske’s success to the simplicity and low cost of the plug-and-play technology, which further reduces operational and customer-acquisition costs.

Packing the network

When you are selling at discount prices, you need to not only reduce your overheads but also sell in bulk and maximise your available resources.

“The more capacity you get out of a sector the better the financials,” states Sorensen, adding that the best way to make money in WiMAX is to stuff each sector full of customers.

“Some of our sectors are running 500+ users which gives great payback. Those sectors pay back in less than four months,” says Sorensen. (The average is about 75 users per sector.)

To maximise sector capacity Danske has developed a suite of network traffic monitoring and management tools called WiMOSS.

WiMOSS monitors each sector of the network on several parameters, including the number of customers, modulation, uplink and downlink saturation, and ARPU.

Danske can identify, for example, whether certain CPE are causing interference in the network, or which customers are bandwidth hogs.

“The real challenge in networks comes after you get 300 customers in a sector, what the levels of interference, modulation and performance are,” Sorensen explains.

If a handful of people are hogging sector capacity, and therefore degrading other users’ experience and jeopardizing revenues, a Dynamic Quality Control tool will cut them down to size, temporarily reducing their bandwidth, and informing them of their anti-social behaviour. Persistent offenders will be eliminated.

“The alternative is just to build more sites, thereby increasing capex and opex,” says Sorensen, pointing out that 92 per cent of Danske customers use less than 5GB per month, utilising 46 per cent of network capacity, while just seven per cent of users take up the rest of the network capacity.

Given that both sets of users are paying much the same, and Danske is not pitching flashy, expensive value added services to the heavy users, it makes sense for Danske’s business model to weed out the bandwidth hogs and pack the network with more moderate users.

“You can generate 3.5 times more revenue per sector using this tool,” claims Sorensen, adding that it costs about E15 per month per sector to run, “probably less than the cost of the electricity to run the BTS.”

Move to mobile

WiMOSS notwithstanding, Danske will invest in further infrastructure in its seven cities in 2008, focussing on increasing existing sector capacity rather than expanding coverage. Small towns or rural areas are “not a good business proposition” because they don’t offer the numbers to pay for infrastructure.

“We are getting more customers in those seven cities and we don’t get many requests from the rest of Denmark,” Sorensen reports. “In Denmark people are moving into the cities.”

Because Danske’s strategy is based on these urban areas alone, it is building blanket coverage across these cities. It is therefore in a position to move rapidly to the next stage of its evolution as an operator, to fully mobile WiMAX services.

Sorensen acknowledges that since Danske began operating in 2005 it has become squeezed on the one side by rapidly accelerating DSL speeds and on the other by the increasing capabilities of mobile data networks.

“We have to either increase speeds to compete with DSL or increase mobility to compete with 3G. Increasing speed is difficult because of capacity limitations, increasing mobility means increased coverage and investment. It is a difficult dilemma,” he says.

WiMOSS and the discount telecoms model offer alternately sophisticated and simple solutions to this dilemma, but these solutions are partial, and Danske’s future lies in taking WiMAX mobile.

In fact, the company has been preparing for mobility services for some time. Indeed, the Expedience equipment can handle hard handover between base stations at speed and Danske has developed a business product for trains and a car kit that includes a rooftop antenna and in-car modem.

“I would call our current network a pre-mobile WiMAX service,” argues Sorensen, acknowledging however that Expedience is not enough for a truly mobile WiMAX network. Like other European operators, Danske has been awaiting the availability of 802.16e equipment in 3.5GHz.

Now it is selecting a mobile WiMAX vendor and is on the verge of conducting field tests with the aim of a “friendly customer launch in the second half of 2008, and a full commercial launch in 2009.”

Danske has the infrastructure in place and it has the requisite chunks of spectrum, which Sorensen proposes to use for separate but parallel networks.

“We can run both the Expedience platform for nomadic WiMAX services as a competitor to the low-end market for DSL and at the same time move into the mobile WIMAX market and compete with the mobile players. You don’t have to take down one network to build a new one.”

Once again, network capacity is the key to Danske’s proposition, as Sorensen argues that the capacity of mobile WiMAX will trump that of HSPA:  “Our competitors will be mobile data operators Telia, TDC, 3 or Telenor and we think they will experience the same problems everyone running mobile networks has with capacity, saturation and unhappy customers.”

Notwithstanding his willingness to confront the 3G establishment, Sorensen also talks of a “need for partnering” in the telecoms sector as technologies converge and multimode, multiband devices emerge.

“We have had some discussions with some operators about roaming between different technologies. They may not be ready for this now but are thinking about it and realise this will be a reality maybe two or three years from now.”

In the meantime, however, one suspects Danske will have a fight on its hands, but with its streamlined operations, a proven discounting model, and WiMOSS in its corner, it may also have a fighting chance.