The Fixed Network Business Case: An Illustration

There’s an interesting point about competitive high speed access markets raised in a new analysis of U.S. high speed access lines, produced by IHS isuppli.

Noting  that 80 percent of U.S. households now are connected to the Internet, some 70 percent using high speed access, IHS notes criticisms of the level of competition in the U.S. market.

“For instance, while an overwhelming proportion of the population had broadband Internet access through cable, DSL or fiber, actual uptake by consumers through subscriptions was much less: 40 percent at the end of 2012 for cable, and 23 percent via DSL,” IHS iSuppli notes.

“This means that less than half of U.S. households with access to cable actually subscribed to it, or that only a quarter of the country’s homes able to connect to DSL broadband chose to do so,” IHS notes.

In the monopoly days when voice was the only product sold to consumers, only one provider was lawfully authorized to supply that service and nearly every household bought voice service, there was scarcely a difference between revenue per connection and revenue per connection.

The same could have been said about investment per line and investment per customer. The figures of merit were very closely related.

All that changes in a competitive market.

Consider a simple example where two facilities-based providers compete in a single local market, each getting 50 percent market share of paying customers.

In that scenario, where each provider builds an access network and each garners half the customers, there is a huge difference between revenue and cost per customer, and revenue and cost per connection.

In other words, to serve four homes, if each provider invests $1000 per home, total investment is $8,000 for the four homes, $4,000 spent by each provider.

Under monopoly conditions, total plant investment would have been $4,000.

Assume take rates are 100 percent. That means cost per customer is $2,000 for each home served (the paying customers have to allow the service provider to recover full network investment).

But assume take rates are only 50 percent. Each gets one customer out of four, so network cost per customer grows to $4,000.

Therein lies the danger for all next generation networks: financial returns are highly dependent on the level of competition and customer demand for the products available on the network.

Consider the video subscription business, where cable operators have 54 percent share, satellite providers have 36 percent and telcos have 10 percent.

Roughly speaking, cable’s network has about 50 percent stranded assets, satellite providers have nearly 40 percent stranded assets (though their network makes that less crucial than for a fixed network or terrestrial provider) and telcos have 90 percent stranded assets.

Or, consider fixed network voice services, where about half of U.S. homes actually buy voice services. Telcos have about 66 percent share, while cable TV operators have about 33 percent share.

Of 100 homes built out by both a telco and a cable TV operator, 50 buy voice services. Some 33 out of 100 homes buy voice from the telco and 17 buy voice from the cable operator. That’s an awful lot of stranded investment.

The only reason the fixed network access business still works is that providers are about to sell three or four units of service to the customers they can manage to attract.

Consider a 100-home area where 50 percent of homes buy voice, 70 percent buy high speed access and 90 percent buy video entertainment. Assume satellite is a major competitor only for video.

Assume telcos have 10 pecrent share of video, 66 percent share of voice and 40 percent share of high speed access.

Assume cable TV companies have 54 percent share of video, 33 percent share of voice and 60 percent share of high speed access.

That results in 70 units sold by the telco, 107 units sold by the cable operator and 32 units sold by the satellite providers, in a neighborhood of 100 homes. Overall, cable sells a bit over one service per home (average), while telcos sell one service to 70 percent of homes.

It’s tough to build and make money on that sort of network. In reality, telcos and cable providers do not actually sell one service to each home, but rather two or three services to one home out of three.

Fixed Network Provider Stranded Asset Risk
Service	Households	Market Adoption	Customer Potential	Telco Share	Cable Share	Satellite Share	Telco Accounts	Cable Accounts	Satellite Accounts
Voice	100	0.5	50	0.66	0.33	0	33	16.5	0
High Speed Access	100	0.7	70	0.4	0.6	0	28	42	0
Video	100	0.9	90	0.1	0.54	0.36	9	48.6	32.4
Total Units Sold							70	107.1	32.4

Carriers Pursue Different 4G Business Models

There are several different revenue models driving fourth generation Long Term Evolution spectrum decisions.

For some carriers, the issue is simply a need for additional capacity.

Though carriers expect and hope that some incremental revenue can be generated, the strategic issue simply is that more spectrum is required, with or without the expectation of immediate net revenue growth from "new services."

The ability to support current customers and demand is reason enough to add more capacity.

In a few cases, 4G has additionally had at least a temporary perceived advantage. Sprint thought its embrace of WiMAX was that sort of decision, allowing it to move first and uniquely in the 4G services market.

In the United Kingdom, EE was able to launch 4G in former 3G spectrum, before the holding of 4G auctions or the building of the new networks.

For some attacking carriers, 4G is a weapon to continue a price assault, as Free Mobile is doing.

Longer term, some carriers also think 4G could enable unicast video services.

Still, contestants sometimes have paid too much to acquire assets. That was true of some Western European operators who, in retrospect, paid too much for 3G spectrum.

Service providers have become wary of such excesses, and auctions for fourth generation Long Term Evolution spectrum, while high in some cases, have not reached levels seen in a few 3G auctions.

But even “rational” decisions can be uncertain, in highly competitive markets.

In the case of a few 3G auctions, one might argue the problem was not the high cost of spectrum, but the low amount of new revenue, and the delay in “discovering” important new apps, which wound up being mobile email, and then mobile Internet access and then personal Wi-Fi hotspots.

In that regard, the danger with LTE auction prices is not so much the absolute cost of new spectrum, which is in some ways a necessary precondition for remaining in business in the future.

The danger is that that increased outlays for spectrum are not fully offset by new revenue streams, or at least higher revenue streams, even if important new apps do not appear immediately.

In markets where customers assume their is a price premium for getting access to LTE’s faster speeds and lower latency, the revenue model is clear. The bigger problem is markets where LTE investments are made, but there is no pricing premium.

That isn’t to say there is no value. Faster speeds and lower latency might well discourage customer churn. But if all major competitors in a market can offer LTE, then the value as a marketing platform is reduced.

Faster networks have value, to be sure. But it is hard to measure the competitive value if all service providers also offer faster speeds.

In the past, supporters of LTE have argued that the real immediate value is the greater bandwidth efficiency of 4G, compared to 3G, used to support data bandwidth. Nor is there doubt that ownership of more spectrum, in markets where demand for data access is growing, has value.

But is has to be disconcerting that LTE access carries no price premium at all, for at least some providers.

French mobile service provider Free Mobile has added an LTE 4G high-speed broadband service to its Free Mobile package without raising the price, in a market where its chief rivals charge a premium for using the 4G network.

Iliad says its monthly Free Mobile subscription remains unchanged at 19.99 euros a month including 4G, without a long-term contract. The no incremental charge policy has a clear logic, of course.

Illiad expects it can use its pricing advantage to further attack the bigger carriers, taking market share. So in the case of Free Mobile, the revenue model is “new subscribers,” not incremental revenue from offering customers a faster network.

Recently, mobile operators in India have faced the problem of revenues insufficient to allow recovery of investments, as well, unrelated to the particular problem of spectrum costs.

India's Bharti Airtel, for example, bought Zain's African operations in 2010 for $10.7 billion, considered a high price at the time, to gain access to the high growth potential in Africa. But competition has lead to low retail prices.

Bharti's average revenue per user grew eight percent from 2012 levels to 192 rupees in India and fell 10 percent in Africa.

For the moment, though, the big incremental decisions turn on spectrum investments.

Since no company invests capital without anticipating a return, every buyer of LTE spectrum has an explicit revenue model. In some cases, there also is an operating cost model (LTE offers lower cost per bit).

But what one hopes is not the case that investments in LTE, or investments in fiber to the home, are not ulimately a matter of “you get to keep your business.” In other words, as many fiber to home investments have largely been made to trade market share with a key competitor, it is not always the case that a traditional return on investment thesis drives the decision.

Instead, the investments are defensive in nature, more a strategic investment than a traditional “high return on investment” decision.

Verizon Acquires Content-Delivery Firm EdgeCast Networks

In one sense, the Verizon acquisition of EdgeCast Networks is a simple way of gaining revenue and customer base in complementary businesses Verizon Digital Media Services sees as important, namely content delivery, transport services and hosting services. 

Verizon Digital Media Servicesprovides quality of service protection for studios, broadcasters and retailers.

Eventually, EdgeCast also could help Verizon in other ways, as Verizon itself could be a big user of such quality assurance features. 

EdgeCast competes with other content delivery networks, such as Akamai, and some would argue CDNs are an example of how a transport services provider can add value to its transport offerings, as well as adding new "services" revenue. 

EdgeCast has 6,000 accounts, including Pinterest, Twitter and Hulu, for example.

To Attack U.S. Mobile Pricing Structure, Sprint and T-Mobile US Will Have to AddressTheir Own Cost Structures

If a mobile service provider wants to attack prevailing retail prices in a serious way, it also has to attack its own operating and possibly capital costs.

That's the only way to sustain lower prices (and hence lower revenue) over time, while maintaining long-term profit margins at a level that allows the firm to survive.

And time works against even a successful attacker.

Given enough time, and enough success, a contestant in a market attacking with “low prices,” and operating with a better cost structure, eventually will find itself with fewer of those advantages. Consider Southwest Airlines, which revolutionized air travel and best exemplified the “discount carrier” concept.

In 2004, there was a 36 percent cost gap in terms of operating costs per available seat mile for the three largest US network airlines versus Southwest Airlines, a study on airline costs found.

But the cost gap has narrowed. By 2011, the “mainline” or “full service” airlines had gotten costs down as well. Southwest still had an advantage, but it was a narrower advantage.

Southwest's cost per average seat mile was still 30 percent lower than even the most efficient of the network carriers. For a brief period in 2008, the cost gap was greater than a 70 percent.

More has changed. Southwest now is the largest U.S. domestic airline, and by some reckoning, its labor costs are higher than the “mainline” carriers. Keep in mind that fuel represents 28 percent of total operating cost, while labor represents 24 percent.

At least so far, it would be hard to find an exact analogy in the U.S. or global telecommunications business. But, in broad outlines, legacy service providers have been chopping away at operating costs, while upstart providers, ranging from over the top app providers to metro fiber access providers have been building business models designed to attack retail prices by economizing on cost parameters.

Some might argue that robust wholesale access policies in Europe have allowed many contestants to enter markets and compete precisely because they can operate without the cost of building their own access networks.

Cable TV operators in the United States arguably have been able to build profitable voice and data access businesses because cable TV operators, while building their own networks, have chosen an approach that--by luck more than design--works well for broadband access and IP services.

Cable TV operators also have had lower operating costs, compared to legacy telephone network operators. As the Southwest example indicates, though, advantages can erode, or even flip, over time.

Whether or not that is possible, and to what extent, in the U.S. or other markets, is unknown. Fixed networks remain capital intensive affairs. And, to some extent, “service quality” is related rather directly to hiring enough people to handle the customer service load.

However, as Southwest Airlines and the U.S. airline industry experience already has shown, given time and will, legacy providers can reduce the operating cost advantage over lower-cost, competitors.

At the same time, the longer a competitive service provider remains in business, and the bigger it gets, the more its costs will tend to grow.

If a SoftBank-lead Sprint or T-Mobile US assault on U.S. retail pricing is to be sustainable, it also will have to be accompanied by creation of a cost structure that sustains reasonable profit margins over time.

How that can be done is the issue. Sprint says Network Visions is one such example. 

Motorola Modular Phone Prototype "Almost Ready"

The first prototype of a modular smart phone, likely including the exoskeleton and at least a few module variations that can be assembled to create a working smart phone with various features, is almost ready.

Though commercialization is a ways off, Motorola's Project Ara, working with Phonebloks, looks to be first out of the gate with such a device architecture. 

Project Ara is an open source project run by Motorola that aims to create a modular smart phone, where crucial components, such as screens, processors, batteries and radios are components that can be put together on a custom basis.

In principle, modular smart phones might be attractive for several reasons. Users could customize their hardware features to some extent, as they now customize the look and feel of their screens and have personalized app loads.

That should allow for the possibility of lower-cost devices as well, as devices are custom-built the way Dell used to assemble PCs only after they were ordered.

Less waste would be another advantage, since a device would not have to be thrown away when a major upgrade was required. Perhaps a module swap would do.

Such devices also would to a greater extent be easier to repair, as a module swap would be possible in some cases, not replacement of a whole device.

The whole idea is to create a smart phone that is completely modular, so each component, including the display, processors, battery, storage, camera, Wi-Fi and Bluetooth, for example, all are all separated into discrete blocks that all attach to a main board and are secured by just a couple of screws.