The much-discussed air-to-ground auction that will determine the future of Internet and cellular access on domestic U.S. flights is underway: Some surprises have already taken place. As I write this, the auction's third round is about to open, the total bid is just $3.7m--far below what I anticipate this will go for--and AirCell has been dismissed as an individual bidder.
I call the winner and the price: Verizon at under $8m. I expect this could have run to $30m or more if Connexion by Boeing had stayed in the running, and I thought that AirCell had the finances to make a good run at it (especially with bidding credits they received of 15% for their business size).
(Update: Round 4 is over. AC BidCo LLC (see below) was the provisional winner in round 1 of one of the exclusive 3 MHz licenses (C; see end of article), while Acadia Broadband and LiveTV were 3 MHz winners in round 2. In round 3, we see Verizon take 3 MHz (license F) for $3.55m and a tiny firm take 1 MHz (E) for just $210,000. Round 4 ended with Unison Spectrum bidding $3.7m for the 3 MHz (F) license. Another round will start Thursday at 10 am Eastern.)
An article in today's New York Times talks about the overall market and carrier constraints on adding in-flight broadband, but Ken Belson missteps when he wrote, "The auction will not advance the on-board use of conventional cellphones, which use other frequencies and are still prohibited in the air." This is not quite right. While the FAA and FCC have not agreed to allow the use of cell phones in flight at any point in the near future, all companies involved in in-flight broadband said that they had picocells to work with major cell frequencies designed to integrate with whatever downlink they develop for pure Internet access. So the auction will advance the technical possibility of in-flight cellular even though many other regulatory, safety, and social issues are still to be resolved.
You can follow live bidding at the FCC Integrated Spectrum Auction System by selecting Auction 65 from the Auction Information popup menu in the Public Access area (lower left). Bidding was in four rounds on Wednesday ending at 5 pm Eastern. Bidding starts Thursday for eight rounds at 10 am Eastern, ending at 4.40 pm Eastern.
The FCC details on AirCell note that a change they made in their filing was grounds for dismissal from the auction because the change "constitutes a major amendment" as it "includes changes in ownership of the applicant that would constitute an assignment or transfer of control..." This turns it into a new application, and thus it is past the deadline. But this doesn't put them out of the running; that change of ownership was already part of their potential bidding strategy. They put in one bid as AirCell and another as part of a potential changed entity that has emberged.
The company had an agreement (filed with the FCC along with supplementary materials) that details how AirCell would become a subsidiary of a holding company which would hold 100 percent of AirCell and 100 percent of AC Bidco LLC, which remains in the bidding process. Shareholders of AirCell become majority shareholders of AC HoldCo LLC in that scenario. Because of their revised filing as AirCell qua AirCell, that agreement must have been put into effect. Thus AirCell is still bidding in just a slightly revised structure and without the bidding credits they would have received on their own of 15 percent.
Only nine bidders were identified by the FCC (DA 06-907, Appendix A in PDF form) on April 28 as qualifying for today's auction: AC BidCo LLC, Acadia Broadband, L.P. (which apparently overpaid $100,000 on its upfront payment), AirCell (now dismissed), AMTS Consortium LLC, Intelligent Transportation & Monitoring Wireless, LiveTV LLC (JetBlue), Space Data Spectrum Holdings LLC, Unison Spectrum LLC, and Verizon Airfone. Inflight Online dug up a little data about AMTS, Intelligent, and Unison. AMTS and Intelligent both report controlling interests by Warren Havens, according to an FCC order released yesterday denying a motion by those two entities.
If Verizon wins, they had previously pledged to roll service out fairly quickly. There are two licenses up for grabs in any of the three configurations of spectrum that are being bid on. The configurations offer either two licenses with overlapping 3 MHz chunks of spectrum (for up and down links), or two different configurations of one license with 1 MHz and one license with 3 MHz.
The companies I spoke to a few months ago highly deprecated the overlapping 3 MHz licenses, designated A and B in the auction. The C/D and E/F pairs are much more likely to win because there's one winner with 3 MHz, which is the minimum required for true in-flight broadband. 1 MHz could be used for ancillary services, but isn't sufficient for Internet access.
Is this spectrum also valued on a $/mhz/pop basis? What would a population assumption be with regard to number of potential customers in the air over the US on any given day or month?
[Editor's note: I don't know the basis on which bidders calculated spectrum value, but we can do a back of the envelope job here.
In 2005, there were 11.5m departures and 738.6m passengers, which counts one person per flight as a passenger. That's over 30,000 flights per day and what must be over 100 people per plane, although given different segment lengths, that's hard to calculate.
Some numbers show a minority of passengers travel on business. So let's be conservative and say that 25 percent of people on 30,000 flights a day might have laptops. Let's say that only 10,000 flights a day are equipped with broadband gear--smaller planes and some airlines won't ever install it.
So it might be that about 125m business passengers per year (very very roughly) are potential customers. That's a smaller group of people, but that represents the number of flights they take, so it's a good match to per-session use.
Can you get an average of a few bucks times 125m per year? If so, this could be a $100m to $500m/year business.
They've found with Connexion by Boeing that's there is a fairly high uptake of usage even at their quite high prices required for the satellite network that carries their data to and from ground stations. Would someone pay $9.95 for a 4-hour flight to Chicago for broadband speeds of 1.5 Mbps or higher (their piece of that)? Very likely if they need to get work done.
What's key is whether there are unlimited monthly rates for regular travelers, which would increase uptake and recurring revenue, and how they price a la carte services. If you get smart carriers, and Verizon as a winner would be a smart carrier, you could have seamless Wi-Fi and EVDO service across an entire trip with just a few moments without Internet access during ascent and descent.--gf]