Friday, March 11, 2016

U.S. Cable TV Companies Gain 106% of All New High Speed Access Accounts in 2015

U.S. cable TV companies now are grabbing all the high speed access net gains, data from Leichtman Research Group suggests.

The 17 largest cable and telephone Internet service providers acquired more than 3.1 million net subscribers in 2015. But cable TV firms gained 3.3 million accounts in 2015, some 106 percent of all additions.

The reason is that the largest telephone companies lost about 185,00 accounts in the year. For those of you interested in inflection points, that loss markets the first year that the telcos ever have lost high speed access accounts.

There is a caveat. AT&T and Verizon are gaining fiber-access accounts, adding 1,481,000 U-verse and FiOS subscribers in 2015.

But telcos have another problem: they are losing all-copper digital subscriber line accounts. In 2015, AT&T and Verizon lost 1,708,000 DSL subscribers.

That performance, one might note, points out a problem for telcos upgrading their networks. Quite often, next generation services and products essentially cannibalize existing products, with no significant net gain in revenue.

The top cable companies had gained 89 percent of all new accounts in 2014, and 82 percent of all new accounts in 2013.

ISPs
Subscribers at End
of 2015
Net Adds in
2015
Cable Companies


Comcast
23,329,000
1,367,000
Time Warner Cable
13,313,000
1,060,000
Charter
5,572,000
497,000
Cablevision
2,809,000
49,000
Mediacom
1,085,000
72,000
WOW (WideOpenWest)
712,500
(15,300)
Cable ONE
501,241
12,787
Other Major Private Cable Companies*
7,945,000
260,900
Total Top Cable
55,266,741
3,303,387



Telephone Companies


AT&T
15,778,000
(250,000)
Verizon
9,228,000
23,000
CenturyLink
6,048,000
(34,000)
Frontier^
2,444,000
101,500
Windstream
1,095,100
(36,500)
FairPoint
311,130
(8,785)
Cincinnati Bell
287,400
17,500
Total Top Telephone Companies
35,191,630
(187,285)



Total Broadband
90,458,371
3,116,102

No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...