U.S. Internet Access: What Would it Take for AT&T, Verizon to Take 10 Market Share Points?

The largest U.S. cable TV companies have 64 percent share of internet access accounts in the United States, according to the latest data from Leichtman Research Group. But there also is an 80/20 rule at work: the firms that drive most of the activity are Comcast and Charter; AT&T and Verizon.

Charter and Comcast have 81 percent of the cable internet customers. AT&T and Verizon have 67 percent of the telco internet access customers.

Between them, Charter and Comcast got 93 percent of the net account additions in the cable TV internet access provider segment. And while AT&T gained marginally, while Verizon lost marginally, nearly all the telco ISP losses came from CenturyLink and Frontier Communications.

In other words, though cable ISPs continue to get virtually all the net gains in accounts, AT&T and Verizon are roughly flat, in terms of subscriber installed base, while it is the rural operations that are losing share to cable rivals.

There might be some larger implications. Assume Verizon and AT&T get about 40 percent share in their markets, with cable getting 60 percent. No matter what they do, how easy will it be for AT&T and Verizon to nudge up to about 50 percent share? And how could they do so?

Basically, AT&T and Verizon likely would have to be able to match cable speeds and product features, as well as offer lower prices cable refuses to match, or otherwise change the value-price bundle in some other way relevant for consumers.

Ratcheting up speeds to match cable likely is the less difficult precondition, as costly as network upgrades might be.

Gaining a sustainable pricing advantage over cable is more difficult, as AT&T and Verizon cannot control the cable reaction. And it is by no means clear that  cable competitors would accept lower market share to protect their profit margins.

Leaving those issues aside for the moment, assume that AT&T and Verizon were able, somehow, to grab 10 points of market share, in part by upgrading to gigabit speeds, with a path to 10 Gbps.

The upside from such an upgrade is about 10 points of market share in internet access. What is a point of share in the consumer market worth?

Assume an internet access account taken from a cable supplier represents about $50 a month in revenue, or an annual $600 worth of gross revenue. If so, one percent of share gain is about 945,322 accounts in the overall market.

The issue is that none of those firms operates fully nationwide, and do not compete solely with the other two firms (Comcast and Charter on one hand, AT&T and Verizon on the other).

But as a simplifying assumption, assume AT&T, Verizon, Charter and Comcast collectively represent about 72 million accounts, and that the share changes would happen only across that 72 million installed base.

In that case, one percent of share change represents 720,000 accounts. Also, since the market is a zero-sum gain, five percent of gain by telcos means five percent loss for cable, for a total net swing of 10 percent, and a new share structure with cable at 55 percent and telcos at 45 percent.

That suggests, broadly, the the upside for AT&T and Verizon, to gain five percent share of the installed base, is really about 3.6 million accounts. At $600 for each account, annually, that implies something on the order of $2.16 billion in incremental revenue for AT&T and Verizon, with AT&T gaining about 69 percent of that.

The implications of a full 10-point change in market share, resulting in a 50-50 split of the market, is $4.3 billion in annual revenue and net swing of 7.2 million accounts,  again assuming that AT&T/Verizon only face Comcast/Charter in their markets.

As a practical matter, the potential for installed base share change between those four firms likely is less than that, since none of the four firms actually faces a zero-sum situation across the cable-telco industry segment divide.

The point is that the revenue upside for internet access gains arguably is less than $3 billion in annual revenue for AT&T, some $1.3 billion for Verizon, if AT&T and Verizon were able to take half the internet access share in their fixed network markets.

Compare that to the cost of upgrading 18 million passings to get those 7.2 million new accounts. Recall that both AT&T and Verizon presently have customers on about 40 percent of passings. So it is necessary to upgrade all passings to capture half the new customers.

At $700 per passing, that implies a network investment of about $12.6 billion, plus activated account investment of perhaps another $2.1 billion, or about $14.7 billion total. Marketing or customer acquisition costs would be incurred as well.

Such customer acquisition costs can run about $2000 per new account, including direct marketing costs and the cost of promotional pricing and incentives. That could add another $14.4 billion in operating costs, for a total of $29 billion.

You might consider that a reasonable bet (spending $29 billion to harvest $4.3 billion in additional revenue). But you also can see why AT&T and Verizon are hopeful about 5G-based fixed wireless access, which might offer capital investment about half what fiber to the home costs.

ISPs	Subscribers, 3Q 2017	Net Adds, 3Q 2017
Cable Companies
Comcast	25,519,000	213,000
Charter	23,603,000	285,000
Altice	4,020,900	16,500
Mediacom	1,194,000	9,000
WOW (WideOpenWest)	730,000	2,400
Cable ONE*	519,062	(2,662)
Other Major Private Company**	4,860,000	15,000
Total Top Cable	60,445,962	538,238
AT&T	15,715,000	29,000
Verizon	6,978,000	(10,000)
CenturyLink	5,767,000	(101,000)
Frontier	4,000,000	(63,000)
Windstream	1,017,400	(8,400)
Cincinnati Bell	307,900	800
FairPoint^	301,000	(3,193)
Total Top Telco	34,086,300	(155,793)
Total Top Broadband	94,532,262	382,445

source: Leichtman Research