Since we last wrote on the topic in September, skepticism about the future of in-flight Internet services has become even more widespread, and recently disclosed usage data from Aircell has not been particularly impressive – roughly 100K sessions per week (of which only a fraction are paid for), equating to about a 5% take rate on equipped aircraft.
The good news is that Aircell is now touting the “operational applications” of in-flight Internet: the obvious corollary being that it is going to try and extract some money from airlines to pay for these benefits, as we suggested it would have to back in September.
The bad news is that the business case for Row44′s Ku-band service looks even more questionable than we had suspected, and it faces a near term deadline (we understand January 2010) from Southwest to secure $100M+ of funding for its planned fleetwide rollout. We have been told that the Southwest-Row44 agreement calls for Southwest to pay Row44 a fee of $0.25 per passenger flown on each equipped aircraft, whether or not they use the service, and Southwest will then mostly likely give the connectivity away for free. With Southwest carrying about 170K passengers per plane per year, that would mean Row44 receiving just over $40K per plane per year (about $22M per year in total once fleetwide installation is complete), which it hopes to supplement with advertising revenue. However, we are doubtful that a dramatic increase could be realized from advertising: for example according to a recent article, in-flight magazines generate an average of about $1M per airline per year in gross advertising revenue, and a large airline such as Southwest would presumably therefore generate in the high single digit millions of dollars from its magazine. Given the lack of technology (and power outlets) required to read the magazine, then even if Southwest gives away the Row44 service for free, usage would be far less than the 80% of passengers that read the in-flight magazine, and we would view it as unlikely that advertising revenue could add more than a few million dollars to Row44′s income.
More to the point, a free service will put an unsustainably costly load on the Row44 network: we believe this was originally designed with an expectation of loading 100 planes onto each transponder (which can provide 18Mbps of capacity), but if 25% of passengers used the network for streaming video, and other high bandwidth applications (remember that these were the primary selling point of Row44′s solution compared to Aircell), then it is quite possible that 1 transponder would be needed for every 20-30 planes. With each transponder costing about $1.25M, Row44 could find itself coming close to spending all of the revenue from Southwest on bandwidth and never making any margin to even begin to pay for the $100M+ of equipment that it would have installed.
In this context, it is far from clear that a sustainable business model is available for large scale Ku-band passenger communications deployments (although a limited Panasonic service on Lufthansa could be viable, assuming Panasonic has some form of revenue sharing agreement with Intelsat and initial installations rely on the old Connexion antennas). Certainly it appears that Viasat, which was the primary equipment supplier to Connexion-by-Boeing (and was rumored to be in pole position to secure a deal with Lufthansa, prior to its recent shift to Panasonic) is emphasizing the lower cost of Ka-band capacity over a Ku-band only model for mobility services. Intriguingly, even Inmarsat may agree that Ka-band is the future: we understand that it has recently issued an RFI for one or more Ka-band satellites, which are likely to be part of its planned roadmap for future government and/or aero services (e.g. UAVs).