Your browser is not supported. Please update it.

7 May 2007

Earthlink running out of options as it scales back on city networks

Earthlink is scaling back on its commitment to the city Wi-Fi market, amid rising losses from its new business ventures and nosediving business in its traditional dial-up market. Its statement that it will concentrate on fulfilling existing municipal contracts rather than seeking aggressive expansion in this area points to a cold – and overdue – dose of reality seeping into the US city wireless sector as a whole, but it also raises some critical questions for Earthlink. With the city networks and its Helio MVNO – the twin cornerstones of its growth strategy – both bleeding cash, does it have the time and momentum to create a broadband business upon them? If not, though, where can it find a model as profitable and broad ranging as its dial-up operations were until recently? One answer might be to emulate AOL, the only provider larger than Earthlink in the US dial-up ISP market, and concentrate on offering services over other companies’ networks (see separate item on Clearwire) – but Earthlink lacks the brand, the software and the financial backing of its rival. Yet if it is to stick to providing its own access links, it needs to find a way to bypass the stranglehold of the broadband incumbents – the main point of going the Wi-Fi route – while proving to its shareholders that it is capable of making profits out of wireless.

Earthlink said it will continue with current municipal projects but will review the results of those it is currently building out – in Philadelphia, Anaheim, Milpitas and New Orleans – before deciding whether to take on any new projects. It will aim to drive up usage in existing networks and halve capex on municipal projects. So far Earthlink has about 2,000 monthly consumer subscribers to its city Wi-Fi services, which executives estimate cost an average of $40 per household to deploy.

The future of the flagship San Francisco project may also be in doubt – Earthlink won this high profile contract in partnership with Google last year, but the city’s Board of Supervisors has delayed a final confirmation vote until July amid rising opposition from various factions, and apparent weakening of interest from the search giant. Looking at the cities where it is active already, Earthlink is “not yet able to establish that comfort level” that the investments are really profitable, chief financial officer Kevin Dotts said in an interview. He added that the company would look at factors like topography, concentration of households and alternative revenue opportunities to determine where – if anywhere – it might bid in future, and that any such bids are likely to be only for large cities such as Chicago, where Earthlink currently has a proposal lodged, taking on other majors such as AT&T. As with other wireless models, the US is once again seeing a trend that promised to extend low cost broadband access even to rural or low income areas being pulled back to focus on the lucrative urban centers.

Nobody could accuse Earthlink of not trying hard enough to try to change this picture and help take part in a new-look telecoms industry. In the face of the dying of its dial-up business, it was faced with having to lease access to DSL lines from the incumbents, in order to move to wireline broadband, and with the lower profitability of those broadband services, compared to the heyday of dial-up. Wireless, and the prospect of higher margin mobility, seemed to be the answer to both challengers, and it adopted a multi-faceted and creative strategy, with metrozone Wi-Fi and its joint venture with South Korea Telecom (first fruits, the Helio youth-oriented MVNO) as the key components. It has also added to this roadmap plans to bid for licensed 700MHz spectrum and to create a technical blueprint for other metro operators, as well as expanding wholesale and hosting activities. With a combination of WiMAX, the MVNO and the metrozones, Earthlink believes it could achieve widespread coverage at relatively low build-out cost ‘ compared to DSL or 3G ‘ and future network upgrades would support a full wireless triple play including television, echoing Sprint Nextel’s plan to use broadband wireless to take on the major telcos.

The idea was that the metrozones would deliver various revenue streams in the short term, to bridge the gap until the company could gain licensed frequencies and use the city networks and Helio as leverage to move into fixed/mobile convergence and other operations – for instance, it is starting a sideline business offering T1 alternative connections for schools, city services and business in cities where it offers Wi-Fi connections; and it plans to franchise its metrozone system to other WISPs in cities where it is not planning direct offerings.

The latter could become more important in the strategy as Earthlink looks to cut back on its own capex and risk. Under the proposed scheme, franchisees would gain Earthlink volume pricing and potentially could share its back office infrastructure and offer its services. Earthlink would not, however, finance the build-outs, and could insist on a first right of refusal should the WISP decide to sell the network. The company said, when it announced the plan last year, that there are about 750 major cities in the US, with 31m households, covering an area of 17,000 square miles, that are ready for municipal wireless. “Earthlink doesn’t have the resources to do this deployment and could not do it in an aggressive time frame. We need partners to build the infrastructure, own it, and allow roaming.”

A mixture of wholesaling and franchising may be safer short term routes to revenue and profit than consumer access, especially given the cut throat price wars that free Wi-Fi and other trends have driven in broadband. But when large companies like Earthlink rethink their commitment to city networks, it has a negative impact on the whole sector, and many of its would-be partners may also be getting cold feet. This could worsen the apparently disappointing performance of the municipal wireless strategy so far, and reinforce current shareholder thinking that Earthlink should cut back its capex – even though pulling back its plans to just a handful of cities reduces it to being a bit player, with no hope of replicating the broad coverage it had in dial-up.

The killer issue for Earthlink is that it is now “not sure” the municipal business can be profitable. Like many smaller city Wi-Fi players, it seems to have been attracted by the relative ease of build-out, and the opportunity to achieve broad footprint at low cost, and ignored the extreme difficulty of making profits from a network in unlicensed spectrum, where the barrier to entry for competitors is low and where there is almost limitless potential for price wars. Though Earthlink plans to make its money from higher value services such as enterprise access and faster consumer data rates – leaving partners like Google to generate advertising revenue from free or best effort plans – that strategy places the operator in an awkward middle position. On the one hand, to justify charging a decent price for services, when there are free tariffs on offer, the carrier must promise certain levels of service quality and functionality, which are very hard to guarantee on license exempt spectrum, and require higher investment in the network capacity. On the other, steering clear of the free or low cost access market, while sensible from a profit point of view, excludes the operator from the majority of the user base, and the opportunity to upsell them in future, giving companies like Earthlink a smaller target market than they had in dial-up. And by steering clear of free services, it remains in a head-to-head with the wireline broadband incumbents, forcing them to undercut them even in the valuable enterprise sector, or to offer additional value such as mobility, which in turn, requires higher capex commitment and the acquisition of licensed spectrum.

In the end, a market set up like that of the US is likely to break down between large providers with the capability to deliver a quad play over a large footprint; experts in generating revenue from free or very low cost services, mainly coming from the advertising model; and smaller specialists in municipal networking that work closely and specifically with local authorities. This last category – which includes successful start-ups like MetroFi – will aim to shift the focus of muni Wi-Fi from free access and bridging the digital divide, to creating networks from which local governments can benefit directly, and to which they will be prepared to commit their custom and a certain level of revenue, and even some upfront financial contribution to build-out. MetroFi last year pulled away from free ad-supported city network deals, saying it could not see a profit model, unless the municipal authority is prepared to commit in advance to service levels and financial contracts. While an attractive model, the shared risk system would be insufficient to satisfy the revenue levels required by a company with the scale of Earthlink, although it may well represent the approach it will take in any future city network engagements.

In the meantime, however, its overall goal is to become a broadband player on a close-to-national scale. Earthlink is leading the lobbying of the FCC to change rules that allow the incumbents to refuse access for competitors to their broadband networks, but it has to assume that the pro-telco mentality of the current regime will continue. This makes wireless the best route for would-be challengers to AT&T and Verizon, but to succeed on any real scale, there is no cheap route to wireless. Sprint and Clearwire are spending hundreds of millions on spectrum and equipment for creating a 4G-style quad play wireless network and they are still carrying significant risk. Despite its clear vision, Earthlink will not be able to join their ranks without making similar financial commitments, and its shareholders may not tolerate this. It does not have the big vendor backing and favors that the two WiMAX operators have attracted, nor the start-up mentality of Clearwire. It will be watching the drain on resources and the investor nervousness that Sprint is suffering, as it sees its traditional business under pressure and a steep path to deliver its next generation. Earthlink has those problems writ large, and now is pulling back even from the limited capex risks – compared to those of Sprint – that is has been prepared to take to get its foot into the wireless broadband game.

Although the main casualty of the uninspiring quarter seems to be the municipal strategy, the biggest current drain on the finances is Helio. Although faring better than many other new MVNOs in a similar catchment base, such as Amp’d, Helio had only around 70,000 subscribers at the end of 2006 and last week dropped its support for offering 3G broadand wireless services based on Sprint’s EV-DO network. This increasingly shows Helio pulling back from its ambitious launch targets of providing a pre-4G service heavily oriented to content and Korean-style multimedia, and instead relying on the old stalwarts of low cost voice and text. Helio generated $30.4m in revenue in the first quarter but made a net loss of $63.1m and despite the back-off from EV-DO, will incur significant new costs this year, and will require a great deal more than the $440m its parent companies initially pledged. Earthlink expects Helio will generate net losses of $330m-$360m in 2007, and reach positive cashflow in 2009. Earthlink’s share of Helio’s loss in 2007 is expected to be $160m-$180m, compared with $85m in 2006 – losses it can ill afford with its core business declining in size, despite strong cashflow, and its other new initiatives either failing or slow to take off. It seems likely that Earthlink will have to sell some or all of its stake in Helio to another player this year, but pulling back from all its new initiatives is hardly a recipe for shareholder confidence or growth – nor for driving Earthlink’s hoped-for creation of a new approach to broadband services that would give it a key position in the quad play, and a way to get out, finally, from under the shadow of the telcos.

Earthlink first quarter results:

‘ In its first quarter, Earthlink saw revenues increase by 4.7% year-on-year to $324.4m, and a loss of $29.96m.
‘ A decline in consumer services was offset by growth in the business sector.
‘ The total number of subscribers decreased 1.3% to 5.27m compared with a year earlier, while churn among consumer subscribers decreased slightly to 4.5%.
‘ Average revenue per consumer subscriber fell sharply, by 3.7% to $18.13.
‘ Total business service revenues increased 162% year-on-year, largely because of a 120% increase in average revenue per business subscriber.