Femtocells not as sexy as hyped

There is currently a great deal of hype about femtocells and their ability to change the economics of providing mobile wireless services.  Based on our research, we believe that femtocells will initially be too expensive and, since the standards for deployment are currently still in development, growth of the femto market will lag behind that of other comparable technologies.

Femtocells are small, low-power, low-capacity base stations designed to operate in a residence or small business.  In general, femtocells offer service providers two opportunities.  First, it allows the improvement of coverage to the home by offering customers services designed to replace their landline or Internet-based voice service like Skype.  Second, femtocells can create operational efficiency and cost savings for the service provider, both because it may allow the use of fewer macro-base stations and because it uses a broadband connection to backhaul traffic so it takes fewer backhaul resources from the service provider's existing macro base stations.

While this technology still faces technical challenges, including coordinating potentially thousands of femtocells with the rest of the mobile network, the economic challenges seem to center around the cost of the femtocell itself and how much revenue and/or cost savings can be achieved by employing femtocells in the network.

Regarding pricing, much of the industry seems focused on reaching a price point of $200 per unit without any subsidy from the service provider.  From our interviews with systems vendors and silicon manufacturers, we believe that such a price will be difficult to reach in the near term given the complexity of the radio, baseband, and processing components required for a solution.

Nevertheless, the market has shown us in a number of instances that prices in that range are a typical starting point for customer premises equipment and high-volume home networking equipment like wireless routers.  Moreover, each of these product classes declined in price by an average of 30% per year over the first five years of life.  Highly commoditized, standards-based product replicas - many based on original design manufacturer products originating in Taiwan - facilitated such price declines.

At this point, we have assumed that the initial average selling price of a femtocell will be $235 and that this will decline by an average of 20% per year and settle to $90 in 2012.  Therefore, we forecast that the growth of femtocell shipments will initially lag behind that of comparable technologies but will more than catch up in the later years of our forecast.

The revenue opportunity associated with better in-home coverage is another key factor in the potential price of a femtocell.  T-Mobile USA's [email protected] service provides a recent example of a service provider addressing this opportunity.  This service uses voice-over-wireless LAN via unlicensed mobile access technology in the home and in hotspots in conjunction with VoWLAN-capable handsets.  T-Mobile USA now offers unlimited domestic calling at home and in hotspot locations for $10 per month for a single line.

If service providers that deploy femtocells intend to compete with T-Mobile's service, they will need to accept incremental ARPU of about $10 per month.  For the sake of comparison, consider that when DSL and cable modem services were first introduced, ARPU ranged from $40 to $50 per month or up to five-times higher than what a femto-based service might deliver.

While for forecast purposes we have considered femtocells to be a single-function device, another alternative that service providers and equipment vendors are considering is the incorporation of the femtocell into other broadband and home networking equipment.  An example is the LiveBox from France Telecom, which Sagem and Thomson supply.  This device incorporates a broadband modem, Wi-Fi, and VoIP in a single box, allowing FT to offer multiple revenue generating services from a single device.  We estimate that FT purchases the LiveBox from its supplier for about $90, about twice the cost of an ADSL router.


We forecast that shipments of femtocells across all technologies will reach 170,000 in 2008 and grow to 27.4 million in 2012, nearly all of which will be W-CDMA-based.  However, this growth lags behind that of other technologies due to our assumptions that a femtocell will cost more and will generate lower incremental ARPU. Nevertheless, we forecast that femto revenues will experience very strong growth, rising from nearly $50 million in 2008 to over $2.7 billion in 2012.

Greg Collins is vice president of the Dell'Oro Group