AT&T’s announcement today that it is “considering” offering speeds of “up to 1 gigabit” in “up to” 21 cities is not particularly substantial in itself. But it marks a major turning point.
Google has successfully re-defined “broadband” to mean a gigabit per second. A gigabit is 1,000 megabits. That is about 980-994 more megabits than most connections today (in the 6-20 megabit range if everything is running right at 3 AM when no-one else is online). It is also a lot more than the 20 megabits Randall Stephenson (AT&T CEO) was straight-facedly telling investors was “all people needed” just a few short years ago. I was in the room when he said that once and the sad thing is most people there just nodded and accepted it.
I wrote a post a few months ago about why/how “gigabit” would become the new access standard (“Why Accept “Meh”-ga When You’ve Can Do GIGA?). It also explains why AT&T’s next line of defense (“Don’t worry investors! This will only be in selected, high-value cities”) is just another un-tenable holding position.
The short version is that a two-tier broadband market is politically un-tenable. It is, in fact, downright laughable. Broadband is already seen as a “necessary” utility. It will end up provided to rural and urban, rich and poor alike. Just like other utilities – electric power, and – wait for it – telephone service!
The genius of Google was to set the bar at “a Gigabit” via their (small but highly publicized) rollout in Kansas City, Kansas and now Austin Texas (AT&T’s headquarters is in Dallas). Most importantly, Google showed that you could build and price a Gigabit network for about the same cost as a megabit network.* This was always obvious, but no-one had made it clear enough that the emperor(s) had no clothes.
That is what is so important about AT&T’s announcement. It is the corporate equivalent of the legs-crossed-hands-to-the-crotch-and-chest moment when they realize that everyone can see them naked. The jig is up. The fat lady has sung. The tide has gone out and they weren’t weaning shorts. Its over. Stick a fork in-em. Book ’em Danno.
The Austin build was particular genius of Google. It is a university town, a tech hub, gets excess attention from the chattering classes (SXSW festival if nothing else), and it is right in AT&T’s backyard. AT&T HAD to respond in Austin. They would have lost the city (and tremendous face) otherwise.
AT&T’s problem is that all the other kids will start asking for a Gigabit too. Especially with Google prodding them along. Remember, Google wins even if AT&T ends up doing the build.
That dynamic is what promoted AT&T’s announcement. They have given up pretending and starting to chase Google. In this phase, AT&T can probably get some decent concessions from the local notables to be first in the line. But every city they add makes it harder to justify NOT adding another city.
I expect the eventual outcome will be a return to some sort of subsidized build-out. That might smack of “socialism” but its exactly how we built the electricity grid and the original phone grid.
I’d also expect the incumbent telcos will probably try to spin off their local lines (keeping wireless) to escape that regulation. Like with the original AT&T spin-off of the RBOCs, the smart investor will probably get a better return from the spin-offs. Boring, predictable cash flows are good. Especially with a guaranteed (regulated) return. Even if you do have to do that icky, dull, digging up of the streets. Besides, betting against telco managements is usually a pretty sure thing…
I’ll finish with a huge round of applause for Google. They have catalyzed something the country desperately needs (better, universal) broadband. And they did it brilliantly. Well done.
* Cost of Gigabit vs Megabit: The incumbent telcos have done a great job of masking this clear, indisputable, obvious-when-you-think-about-it fact. Consider;
- Today’s corporate LANs run at 1 gigabit. If we can get a cost-effective gigabit to every desk in a street-block-sized office floor, why can’t you do the same in a street-block-sized (ahem) neighborhood block?
- Heck, you can buy a Netgear 8 port gigabit Ethernet switch at Best Buy for $52.99 (on sale now!). That is $6-7 a port. A Netgear 8 port 100 Megabit switch (10x slower) is $41.99 ($5-6 a port. The telco-equivalent gear is different, but not THAT different. If anything, the mega/giga cost differential is even lower. And an extra few dollars cost per house is totally meaningless spread out over the 30 year life of the fiber.
The reason the telcos never offered a gigabit is because they would have to have offered it to everyone. Much better to fabricate the perception of scarcity.