Aircell agreed to put substantial service in the state of Alaska as part of its arrangement with the airline, with a commitment to have it ready by early 2011. Aircell will eventually have a Canadian network, too, run by a Canadian partner that will aid in Alaska coverage.
As I predicted, Google won't be sucking down Wi-Fi signals in its future Street View efforts in some countries: After the debacle of Google first saying it wasn't collecting data from Wi-Fi networks, only scanning for readily available public information, and then discovering and admitting it had stored information, the company is taking a different tack.
It's restarting Street View photography in Ireland, Norway, South Africa, and Sweden, but vehicles won't have Wi-Fi hardware on board, and the software has been vetted by a third-party to ensure there's no component that might have collected Wi-Fi data still installed (even though removing the hardware might be seen as enough).
I thought that the likely outcome for Google for its missteps was likely a very tiny amount of money in the forms of fines or voluntary settlement figures, but no criminal charges nor more than a technical slap on the rest--so long as Google agreed to stop scanning Wi-Fi signals, even if it promised to stop collecting data.
By being seemingly forced to exit the Wi-Fi positioning business, Skyhook Wireless reaps the biggest rewards, in that it will be the only worldwide provider of such information.
However, Google also uses the Android platform to collect Wi-Fi positioning information--something also employed by Skyhook Wireless, as News.com reported a few weeks ago. Every time a mobile devices sends a snapshot of the Wi-Fi environment to a Google or Skyhook server for lookup, that information further refines location data for subsequent users.
But mobile-submitted data isn't enough. For one thing, most of this data isn't tagged with reliable GPS coordinates when sent to the server--the intent of sending to the server is to obtain latitude and longitude in the first place. Skyhook and, formerly Google, drives with precision GPS receives and high-gain antennas to seed and re-seed their databases.
Meanwhile, in Australia, the country's privacy commissioner has found Google broke the law in sucking down data, even though such data was being publicly disseminated. The Sydney Morning Herald quotes commissioner Karen Curtis saying, "Any collection of personal information would have breached the Australian Privacy Act."
But I fear this sends the wrong message. Curtis says, "Australians should reasonably expect that private communications remain private." Not quite. If you're sending information unencrypted when the facility to protect that information is readily (and freely) available in the hardware you purchased, then you are sending private information in a public fashion, and shouldn't enjoy any expectation of privacy. Setting the bar that publicly broadcast information ensures privacy protections seems a bit rich.
Nevertheless, Google has apologized to Australians. Expect more apologies to be forthcoming.
Starbucks launches its free North American in-store Internet today: It was quasi-free before (two hours with a single purchase ever since December's changes), and now you pay nothing to use as much Wi-Fi as you want.
Will *$, as we affectionately call the company here in Seattle, be filled with snoring loafers and couch squatters? Likely not. The company found an average of an hour's daily use by those who took advantage of the two-free-hours offer previously.
Starbucks will give away Wi-Fi at nearly 800 outlets in Canada: Starbucks announced today that it would provide free Wi-Fi service at its company-owned stores in Canada, in addition to the nearly 7,000 US locations gaining no-fee Internet access. The service becomes completely free in both countries starting tomorrow, 1 July 2010.
Let's run the numbers. Aircell charges from $5 for a red-eye session to $35 for a monthly unlimited usage pass across the network. Let's assume the average transaction fee is $11.
Aircell is nearing 1,000 planes equipped with Gogo Inflight Internet, and one assumes as that number has gotten higher, revenue has increased at a higher pace. So let's average the year's use across 800 planes (assuming several hundred more planes by year's end).
Take $95m and divide it by 365 days. That's $260K per day. Divide that by $11, and that gets you 24,000 sessions per day. Divide that by 800 planes and you get 30 sessions per day on average per plane. Some planes are flying short hops (Virgin America's West Coast routes), and others are heading cross-country on 5-6 hour missions. Planes are being turned somewhere between two and five times per day, so divvy up those 30 sessions per day per plane into flights.
That would means somewhere between 5 and 15 sessions per equipped plan per flight. I'm making a very rough calculation, but it makes sense.
As usage goes up, average revenue per user will increase, but eventually taper off: frequent travelers will opt for the $35/mo rate (probably negotiated lower through corporate bulk purchase deals).
By the number In-Stat is estimating and my logic above, it seems that achieving $500m/yr in revenue could be achieved by 2014 both by doubling the current size of the fleet, and increasing casual and business usage.
Aircell and its partner airlines have rarely disclosed much in the way of exact usage, session numbers, users per flight, or other financial or session data.
At Ars Technica, I explain femtocells: In a long article at Ars Technica, I explain what makes femtocells tick, and whether they wind up as a good deal or not for consumers. I've been skeptical for years about femtocells because they are a tricky value proposition for carriers to explain.
"Our network is great, but because it's not, pay us extra money for this thing that we've advertised we can already do."
Not a great sale. Verizon, with the best network for voice in the US, sells its femto for $250, and it's just intended for improving reception. I don't hear complaints about it. People who live in places with poor coverage know that they get great coverage elsewhere, I suppose, and suck it up.
Sprint splits the difference, selling it ($100) for both unlimited calling (with a monthly fee of $10-$20/mo) and pure coverage ($5/mo). Its deal is best.
AT&T charges a relatively low price ($150), but because of complaints about its network coverage and quality in some urban areas, gets the most criticism as it's not precisely what people want. If AT&T coupled the femto with a decent calling plan, it would be more of a sell. AT&T wants $20/mo for unlimited North American family plan calling, which is only slightly cheaper than unlimited calling without being tethered to a femto.
Atom Feed (xml)