Commentary: AT&T’s Culpability In The “Bandwidth Shortage”
There ‘s been an awful lot of controversy surrounding AT&T lately:
- ‘Operation Chokehold ‘, the quasi-protest quasi-organized by (presumably) quasi-serious Fake Steve.
- Reports/rumors of AT&T planning to incentivize users to use less bandwith; possibly by charging additional fees for unlimited bandwidth or eliminating the unlimited plan “ the rumors abound.
- Gizmodo posting documentation stating that AT&T has been dropping capital expenditures on network infrastructure ever since the iPhone ‘s introduction.
A pretty bad picture for AT&T being painted here, to be sure, but one thing sticks out in my mind “ what expectations DID AT&T have when the iPhone was introduced?
When the iPhone was introduced, there were only a limited number of apps, no App store, and only one app that did any kind of streaming “ YouTube “ and at low quality (bandwidth) at that. There was no Pandora, no Google Earth, no MLB At Bat. In fact, I ‘m pretty confident that the discussions that Apple and AT&T had at the time likely centered around the expectation that the iPhone would consume somewhat more bandwidth per customer than, say, a Blackberry “ but I don ‘t think they ever imagined in their wildest dreams that a) the iPhone would take off as fast and furious as it did and b) that one day, apps that eat bandwidth like there ‘s no tomorrow would be on just about all of them.
Consider that the stock original iPhone:
- Was Edge only (although they assuredly knew a 3G version was planned)
- Had one streaming video app and no streaming audio apps (save for links from Safari)
- Had no VOIP apps
- Did not readily support uploading pictures and other comparatively large files
- Did not permit iTunes downloads except over WiFi
As a networking professional, I understand capacity planning. At launch, the iPhone must certainly have seemed ‘safe ‘ to AT&T ‘s networking folks “ they probably expected it to sell roughly as well or perhaps a bit better than your average Treo; in other words, from a cap-planning perspective, nothing to get terribly excited about. Unless you watched YouTube 24×7 (and really, two years ago, most videos on YouTube weren ‘t terribly enticing), your data usage was likely pretty low. Sure, you emailed, surfed the web a lot, and looked up stuff on Google Maps; but that was low bandwidth, intermittent traffic. That ‘s what mobile data was before the iPhone, and that ‘s what AT&T was likely counting on.
The App Store changed all that.
Fast forward again to the present, and now there are apps that are constantly using bandwidth that are being run for hours at a time. Compare the bandwidth used for an hour ‘s use of Pandora versus an hour of web surfing and it ‘s no contest; now multiply that times the number of iPhone owners and you begin to get an idea of the huge difference between what AT&T likely expected and the reality that they now face. And it ‘s not just Pandora, of course, but Facebook and Twitter and Google Earth and all the other popular apps that use data.
Think about how often an iPhone was actively used in late 2007 (once the novelty wore off) versus how often it was put away, then compare that to now. If you ‘re like me, you are indeed using a lot more bandwidth now because there ‘s a lot more to DO online now with the iPhone. Back in 2007, we couldn ‘t listen to or watch a 3-hour long baseball game — or listen to Sirius or XM all day at work. I ‘m willing to bet that AT&T was willing to bet that we wouldn ‘t be, either.
‘So, build up the infrastructure! ‘ comes the next pointed barb. Sure, that ‘s the obvious solution “ throw more hardware and bandwidth at it. The concept is simple, but, as they say, the devil is in the details. Undertaking something of this magnitude is measured in years, not months. Even with an unlimited budget, AT&T still must secure permits, identify and arrange for sites, coordinate frequencies, test and physically install equipment, and the other myriad tasks associated with retrofitting old cell sites and installing new ones. It ‘s far from an overnight process; rather, it ‘s extremely time consuming. Further, we ‘re discounting any obstacles, such as problems securing permits, NIMBY-ism over cell site locations, etc.
So while AT&T can fairly shoulder the blame for voice call issues (which are pretty much inexcusable since every customer in a given area is affected), I ‘m not sure that we can hold them totally accountable for not having all the bandwidth required to handle both the mammoth success of the iPhone nor the advent of the App Store. Yet.
Check with me in two years.
And yet they keep selling these phones (and other smartphones), despite the fact that they haven't caught up with the infrastructure. Of course they're to blame. If not, they would have stopped greedily selling them until they had improved the network.
Actually, they were probably contractually obligated to do so.
It doesn't take 3 years to improve network load issues if you're serious about it. To your own point 3G was never a surprise, but neither was the concept of 3rd party apps. They may have taken hold faster and in greater volume, but a good network architect should have had a fall back plan knowing that this could very well be possible. The decline in capex (if it's true) is not only irresponsible but is a statement about the integrity of the organization. They are clearly saying one thing to their consumer base and doing the complete opposite. Not a company I want to do business with. Apple deserves some blame here for allowing the exclusivity to continue as even if a contract does exist the service level of AT&T almost certainly should allow for breach with some type of out clause. Apple should not only be demanding the improvements but actively engaged in the project and monitoring progress. Clearly AT&T cannot be left to operate on the honor system. All I can say to AT&T is, "Good luck repairing this brand". Once they lose exclusivity nobody is going to touch this network if you pay them to. If I were on the board, I would fire the Chairman/CEO and everyone below that has any responsibility for the wireless division. This is a clinic on how to destroy a successful organization.
Actually, the apps almost certainly caught them by surprise; until October 2007, Apple was insisting that the only 3rd party content would be via Web apps. The SDK was released in March 2008, and Pandora (for example) wasn't released until July; and many of the other "bandwidth hog" apps were released roughly a year ago. Data usage on AT&T ramped up exponentially after the release of the 3G, which was in July 2008 — same as Pandora. (http://www.techcrunch.com/2009/10/21/how-the-ipho…
So figure by this time last year they knew they had a problem on their hands. They've obviously been working on it, but in places like San Francisco they've had a 2000% increase in data traffic. (http://digitaldaily.allthingsd.com/20091117/thank…
To think that this kind of capacity is added nationwide in 12 months is to be living in a fool's paradise, my friend.
But, if you think the grass is greener, let's look at Verizon's LTE rollout. In Dec 2008, they announced that they wouldn't begin deployment until 2010.
(http://www.engadget.com/2008/12/10/dick-lynch-ver…
Then recently, they announced that it will be completed "by 2014" with up to 30 markets rolled out in 2010.
http://text.broadbandreports.com/shownews/101893
..And the LTE equipment will coexist with CDMA, a much less tricky affair than mucking with your in-use equipment.
But hey, grab a pitchfork and a torch if you must – if Verizon gets an iPhone next year that's LTE capable and you can't use it, I presume you'll call for their heads as well?
While I think some are too harsh I think you are being too soft on AT&T. Agreed things like this take time etc but they could have done a much better job than they have. Some areas are fine but some are very bad, that points out bad planning and monitoring of what needs to be done to the network. Plus if you look at their pricing they charge more for texting than they do for data. If data is such an issue why is it cheaper than texts?
Yes and no. While I am giving them some rope on dealing with the unexpected surge, I give them no free passes when it comes to voice, or the 3G rollout in general.
You mention that some areas are "very bad" – if one of those areas is San Francisco and you are referring to data, you should be aware that data traffic has increased a whopping 2000% (not a typo – two thousand percent) in 2009.
http://digitaldaily.allthingsd.com/20091117/thank…