Fred Vogelstein has an interesting piece in Wired about AT&T's uncomfortable relationship with Apple and the origins of the now widely used shorthand "#attfail":
Wired: Inside the iPhone Network Meltdown
It's got some interesting insights, particularly how pissed off Steve Jobs was about idiot boy Ralph de la Vega promising tethering was coming "soon" before a deal had been worked out with Apple on pricing.
But it misses a big point. It buys into the idea that AT&T was "shocked, shocked, shocked" that customers actually consumed a lot of data on their iPhones and simply couldn't keep up with demand. It details that AT&T wanted to ramp down certain iPhone features, like tethering and video, in order to help it keep up with data needs.
But this is bullshit. Vogelstein, probably trying to be balanced, buys into AT&T's lies about it's data problems. Any idiot could have seen the iPhone was going to use a lot of data. The biggest problem for AT&T is not handling data (though it clearly doesn't want to pay for the infrastructure necessary for quality service), but that it wanted to suppress features so customers continued to buy wi-fi and also to give it time to force in a tiered pricing system.
As Vogelstein points out, AT&T made record profits from it's wireless division thanks to the iPhone. The proper response to data needs would be to seriously invest in infrastructure, something AT&T only reluctantly did. And if that isn't enough, a modest raise in iPhone rates could have been justified.
iPhone customers biggest complaints are about dropped calls, which are not due to data demand, and lack of signal, which also has nothing to do with the amount of data being consumed. Speed, which is also not that great on AT&T, is the real victim of data demands, but most customers are willing to wait longer to get what they want. You loose a call, you get pissed off. You can't make a call, you get pissed off. You have to wait a little longer to download your Google map, not as big a deal. Data demand has NEVER been the real issue with the iPhone. Being able to make phone calls, THAT'S an issue. If you buy a cell phone in a major city, like New York, you expect to be able to make calls on it. No one has been yelling because they can't download pirated movies fast enough.
Meanwhile, AT&T had more important goals than servicing iPhone customers desire to simply make phone calls on their $100 a month two year minimum plans. They used iPhone profits to build out their cable and wi-fi business (which the wireless business competed against). iPhone customers ended up subsidizing all those McDonalds that offer free wi-fi. (And those $19.95 initial monthly cable service deals.) If people started using their iPhones for tethering (unless there was tiered pricing or it was cost prohibitive) and for video (therefore not needing AT&T's cable offerings) the wireless business might cut into AT&T's other, less profitable enterprises.
In short, AT&T was betting from the beginning against the iPhone and a wireless all you can eat network. They want a wired network (that they can control better) and a tiered wireless network that is too expense to really use for all their customer's needs. They want people to pay two bills.
AT&T's justification for tiered pricing is that most people DON'T use much data (which of course is a lie). The problem is not that AT&T didn't realize that it needed to build a much better wireless network to service customers, the problem is that AT&T doesn't want customers to have access to a really great wireless network. They want customers money, but they don't want to provide a service good enough to cut into their other business, which is wired. Also, they don't want customers to have a lot of options, which was what the lobbying against net neutrality was all about. AT&T spend it's early iPhone years hoping it could get Google to pay it extra money for simply providing access to it's internet sites. After buying out all those little regional companies, they wanted a deal to lock up the internet forever.
AT&T (and the other US telecoms) realize that wireless is dangerous. For all the bitching about finding places for towers, and needing spectrum, the real problem is that as technology improves, it will become easier and easier for competing companies to offer competing wireless services, and easier for products like the iPhone to switch carriers. If people abandon wired networks, AT&T will not only have wasted billions in buying up old wired networks, but will have real competition.
So the problem is not that AT&T didn't anticipate the future. AT&T doesn't want the future to happen. They are like feudal Japan when it decided they didn't like guns and preferred swords. That didn't work out very well in the long run and AT&T's giant bet against wireless will turn out to be a bad move too.