On Monday, the Wall St Journal revealed that Harbinger plans to push ahead with “a multibillion-dollar plan to build an international satellite-cellphone business” which would “complement existing cellular networks with satellite coverage, and…use new chips that could fit inside affordable, mainstream phones, keeping costs down for consumers”. This is quite different from the original plan of SkyTerra, TerreStar and ICO to build out Ancillary Terrestrial Components (ATCs) – basically a new terrestrial cellphone network using their satellite frequencies. Instead, subscribers will simply rely on their existing cellphone networks and only use the satellite services of these companies as a roaming partner when they are in uncovered areas. From that perspective, the new plan is much more similar to the business plans of Iridium and Globalstar in the late 1990s, that business travelers would rely on satellite networks to fill in the gaps in cellular coverage. For example, here’s a description from the Economist in June 1998:
“By far the largest number of subscribers is likely to come from the “cellular roaming” market. These are users of land-based cellular phones who want to be able to extend the range of their handsets when they are travelling through areas of poor or incompatible coverage. MSS subscribers will be equipped with a dual-standard phone that will switch to a satellite when a ground connection is unavailable (Iridium’s first offering is a soon-to-be-superseded $3,000 half-kilogram brick). Subscribers will pay a higher standing charge to their normal cellular operator and a premium on MSS calls. Numbering will not change and unified billing will be standard. This week Iridium said it had recruited 200 distribution partners among cellular companies.”
Of course there are many advantages that the new and very capable satellites being built by SkyTerra and TerreStar will offer over the 1990s technology of Iridium and Globalstar. Most obviously, the extra power and sensitivity of their satellites will allow the satellite service to be added to mainstream cellphones with little or no penalty in size and weight, as opposed to the ‘brick’-sized handsets produced by Iridium and Globalstar in 1998 and 1999. In addition, SkyTerra, TerreStar and ICO have signed agreements with Qualcomm to incorporate satellite technology into Qualcomm’s next generation cellular chipsets, which are likely to be used in a wide range of handsets.
However, there is a major difference between the principal sources of revenue for an ATC and a satellite roaming business plan. In the ATC case, a cellular operator would pay to lease the satellite spectrum to provide terrestrial services over a new terrestrial base station network, thereby enabling it to add capacity or new broadband services to its network. Satellite services, while available, would be a minor component of the overall revenue stream for the satellite operator. On the other hand, a satellite roaming business plan relies on the satellite services themselves to generate revenue, with perhaps some incremental benefit to the cellular partner through reduced churn, if the satellite service is sufficiently compelling to subscribers.
Even more importantly, the decision maker who will produce these revenue streams is very different: in the ATC case, it is simply a matter of convincing the cellular operator to lease the spectrum, whereas in the satellite roaming case, the end user must decide to buy the satellite service. Many people who were intimately involved in the launch of Iridium and Globalstar’s services remain convinced that it will be very difficult to explain the limitations of satellite service to a mass market: those services were sold as enabling coverage “anywhere”, and so there were numerous complaints about the inability of satellite service to work reliably in buildings, cars and urban areas. For most people, their experience of cellphone coverage limitations is in precisely these areas: in the Bay Area there are 375K riders of BART each WEEKDAY (where coverage in the underground parts of the transit system has only recently started to be deployed) compared to less than 200K visitors to Pinnacles National Monument each year (the location where we most recently spent an extended period of time outside cellular coverage). Remember also that even the new phones almost certainly won’t switch beween terrestrial and satellite modes in the middle of a call, so will likely drop an ongoing call if the user moves through a cellular (or satellite) deadzone. As the Wall St Journal explained in July 1999:
“At its core, Iridium is struggling with an incongruity between its design and its market ambitions. It was originally intended for millions of globe-trotting business travelers, and it was launched with a $180 million world advertising campaign last year aimed at that market. But when Motorola began operating the system on Nov. 1, the Iridium handsets weren’t powerful enough to work within buildings or urban areas. As a result, a vast network intended for a mass market was usable only by niche groups, such as mariners, oil-rig workers or the military. Iridium faces a tough struggle to cover its huge costs in such relatively small markets.”
Indeed there are about 150,000 Iridium and Globalstar satellite phone subscribers within these niche markets in North America at the moment, generating about $100M in retail service revenues per year (excluding international users like the DoD). New smaller, cheaper handsets from SkyTerra and TerreStar should increase the size of this “professional” MSS niche significantly (including amongst “police, fire and ambulance personnel”). In addition, a low cost “satellite backup” service might appeal to several million consumers, particularly in earthquake or hurricane-prone areas such as California or the Gulf Coast, if it is explained properly: as an emergency service for use outdoors in the event that other communications are unavailable. In order to achieve this level of take-up, cellular carriers will not only have to sign roaming deals with the satellite networks, but also ensure that satellite connectivity is included in the phones they sell and support large scale distribution of the phones themselves. Even then, it may be hard to explain properly: there were reports after Hurricane Katrina that first responders were unable to get their satellite phones to work, and it was later discovered that some were trying to use the phones in a basement conference room or inside the Superdome. Although gaining several million subscribers would be a great achievement for the MSS sector, in view of these challenges we remain skeptical that there will ever be “vast global demand for the network [Harbinger] envisions”.
In contrast, we are more positive about the long term potential of ATC: cellular operators will ultimately need more spectrum to cope with the surge in wireless broadband data demand and will use up the stockpiles of 700MHz and AWS spectrum which they have purchased in recent years. At that time ATC will be one of the most obvious sources of supplementary spectrum, and there is no technical reason why it can’t be made to work. Indeed many of the developments being put in place to enable satellite roaming (such as the Qualcomm chipset) are precisely those needed as a pre-requisite for ATC deployment. The only problem is how long it may take before major cellular operators feel a pressing need to use MSS spectrum for their terrestrial operations – it is likely to be several years off at a minimum. Indeed, if WiMAX struggles, then Clearwire’s spectrum may be sold off to other players, pushing out the timeframe in which ATC might be considered even further into the future.