AT&T (T -0.47%) had a fairly mediocre year, in that its stock price looks to finish 2015 in roughly the same place it began it.

That was partly due to increased competition in wireless and somewhat due to the industrywide struggle to fight off cord cutters in pay television. The company actually did quite well in the mobile-phone part of its business, adding subscribers even as T-Mobile (TMUS 0.30%) and Sprint (S) offered its customers lower prices.

AT&T did less well in pay television, where it lost customers (though not many) through the first three quarters of the year. That was to be expected, given that consumers, especially younger ones, have become more open to the idea of cutting the cord and dropping pay TV altogether.

Though the company held its own in wireless and pay television, it didn't do as well in the one segment where the overall industry has been growing: AT&T had a poor year in its Internet business. While its competitors were almost all universally posting gains each quarter, AT&T followed a slight increase in subscribers in Q1 with bigger losses in Q2 and Q3.

Subscribers (lost)/gained

CompanyQ1Q2Q3YTD
AT&T 69,000 (136,000) (129,000) (196,000)
Verizon 41,000 (25,000) 2,000 18,000
Comcast  407,000 179,000 320,000 906,000
Time Warner Cable 328,000 189,000 246,000 763,000
Frontier 17,000 29,000 27,000 73,000

Source: Leichtman Research Group.

What happened?
Put simply, the major cable companies ate up all the growth in broadband. That left AT&T, which still serves many of its customers using slower DSL service, struggling for an answer.

Bruce Leichtman, president of Leichtman Research Group, spoke in a press release about cable-dominated broadband growth in the third quarter, which followed how the earlier parts of the year had gone:

"While major providers now account for nearly 90 million broadband subscribers in the U.S., top cable providers added subscribers at a faster pace over the past year than they did over the prior year. Over the past year, cable companies accounted for 103% of the 2.93 million net broadband additions."

Why is this happening?
While it makes sense that the pay television world is getting smaller because of cord cutters, the broadband universe should grow for the same reason. Cord cutting requires an Internet connection -- a good one if possible.

That has been a boon for cable providers that have offset their television losses by adding broadband customers. For AT&T, it's clearly a question of not offering the top-tier, highest-speed service in some markets, as in Q3, LRG's numbers showed that along with Verizon (VZ -0.09%), the telcos lost 432,000 DSL customers while gaining 305,000 for their respective fiber products, U-verse and Fios.

That's happening largely because the telcos built out their Internet business over the now-inferior DSL lines. AT&T (and Verizon) are working on fixing this, but for now, a customer who wants broadband for cord cutting reasons would probably not want DSL. In addition, in many cases AT&T requires a landline before it even offers DSL and it seems unlikely that in the company's DSL-only markets that most people looking to save money would opt for a landline.

Can AT&T fix this?
If cord cutters are driving the broadband expansion, they are pretty universally saying no to DSL when they have a choice. To hold onto customers or gain them, AT&T simply has to continue to build out its network to offer more of its service areas a better product.

That's something the company is working on. It's investing heavily in adding capacity, and there's no reason to believe that, in the long term, it won't be able to reverse the fortunes of its ISP business. It pretty much needs to do that, because there is every reason to believe that even with its purchase of DirecTV, pay television will continue its slow decline for years to come.

AT&T needs top-tier Internet service in all its markets. As soon as it has that, there's no reason it can't use its ability to bundle pay TV, Internet, wireless, and even landlines to reinvigorate its subscriber growth.