IFF the take rate is high enough. It’s completely unannounced and until the Kansas City details came out I didn’t believe it myself. Sergei Brin and the team at Google want to move ahead on a plan to run fiber to 10’s of millions, maybe hundreds of millions of homes. That’s implied by the $70/month price, which is high enough to be quickly profitable if 20-30% of homes sign on. Google’s research says they’ll get that high a rate, but no one is sure until they actually offer the service to 170,000 homes in Kansas City.
Jason Calaconis (a friend) writes “Everyone who would pay $70 for Google Fiber, raise your hand,” and predicts “[95% of folks raise their hands ].” I hope he’s right, but it’s not certain. Around the world people offered more than 10 or 20 megabits have very rarely been willing to pay much more for it. Overwhelmingly where given the choice of speed for more money, most people have gone for the cheaper service. Very, very few people except heavy downloaders have been willing to pay much more for over 10 meg.
With two HD channels fitting in 7 meg and almost no sites effectively downloading at more than a meg or two, it's rare for most people to max out 10 meg. For the small amount of time they do, many aren't willing to pay much extra.
I have empirical examples of surprisingly low high speed take rates from Sweden, France, Britain, Hong Kong, Japan, Korea, and California. A difference of $10 to $30 discourages 3/4ths of the customers. Google is looking for $70/month for a gig, compared to 10 meg cable at about $50 and 3 meg DSL at about $35. Cable in Kansas City can easily offer 20 or 30 meg for the same price and possibly 50-100 meg. It’s all DOCSIS 3.0.
A $50 difference between 10 meg cable and 50-100 meg DOCSIS 3 has been standard in the U.S. and the take rate on the higher speeds has been so low no company will reveal it. Cablevision is coming down to $70 and we’ll soon know if it helps. All of these prices are approximations and available with different deals and bundles, of course, so none of these comparisons are exact.
“Mark my words: Google Fiber is not a test, it's a takeover plan,” Jason writes. But Google’s way is to test, I learned from Steve Levy’s excellent book “In The Plex,” well worth the $16 from Amazon. So it makes sense they are waiting for confirmation from actual orders before pulling the trigger.
Some less reported datapoints about Google fiber:
- The box has two terabytes, enough for 500-1000 hours of HD video and/or a fine home music collection.
- It also has eight tuners, so it’s almost impossible to have a problem with too many simultaneous shows.
- HBO and other networks are refusing to sell to Google at a price Google considers reasonable.
- It will not be close to the originally announced open access model, Stacey Higginbotham reports in an important article http://bit.ly/PpVsw0. Scott Canon of the KC Star confirms that directly from Kevin Lo of Google http://bit.ly/PltraJ. Jeff Baumgarten quotes Steve Effros open access “disappeared before they even dug their first trench" http://bit.ly/MzbPIU
Harold Feld asks “They're really up to date in Kansas City. But have they gone as far as they can go?” Not nearly - wait till they turn on the public WiFi from every box.