Shared Infrastructure for Small Cells?

Service provider thinking about infrastructure sharing always is intimately and directly related to their perceptions of business advantage. Actors will favor sharing when economic or business advantage can be obtained, and will oppose it when there is perceived harm.

New questions will arise as small networks become essential for 5G networks. In many markets and scenarios, it will be argued that only a shared infrastructure approach will work.

Networks that are dense, with large numbers of small cells, and virtualized baseband signal processing, will require a huge number of new radio sites, backhaul links and power sources. In markets with multiple suppliers, there is “a clear argument for a single, shared network,” argue Rethink Wireless analysts.

On the other hand, as always, larger suppliers will think hard about any shared infrastructure proposals that allow competitors to compete more effectively. In the U.S. market, that has been a major reason why larger incumbents have resisted and opposed mandatory wholesale requirements that are similar to shared infrastructure proposals.

As always, that might be an argument better received in some markets than others. Where one or two suppliers believe they have financial or other advantages favoring building and owning their own network, they are likely to act that way. In other markets, where no single provider believes it can reap advantages for building and owning its own small network, cooperation and shared infrastructure are likely to be better received.

Those third party networks might work much as tower companies now operate, offering colocation and backhaul to multiple mobile operators. There are some differences, where it comes to small cells.

Cable TV operators long have expected their dense high-capacity networks would allow them to become retail or wholesale operators of small cell infrastructure. That means multiple entities might believe they have advantage where it comes to access networks. AT&T, with its large fixed network footprint, is among them.

Most other fixed network telcos, including Verizon, have smaller in-region fixed assets to leverage. Sprint and T-Mobile US would be most likely to favor some shared approach, as they own virtually no fixed network assets. It is possible each of those firms, if acquired by a cable operator, would have less interest in third-party shared small cell infrastructure, unless their parents wished to consider it.

And even there, cable operators are more likely to partner with other cable operators to fill in the out of region coverage.

At the moment, one might argue the prospects for small cell infrastructure in the U.S market are less favorable than in some other markets.

Line Between Managed Services and OTT is Getting Harder to Define

Managed services  are not the “internet,” a source of some confusion about the proper limits of

policy intended for one or the other domains. Managed services such as business access services are specifically exempted from “net neutrality” regulations. In the consumer arena, subscription video or no-incremental-charge “over the air” services also are managed services.

There are clear business model implications. Consumers who watch “free, over the air TV” do not pay the provider for bandwidth consumed to deliver the services. Neither do buyers of linear subscription TV services.

The situation is a bit less clear for  “over the top” services, where there is no charge levied by the app provider, but the customer “pays” for bandwidth only in an indirect sense, for an internet access connection. But one business model “rule” is clear. Consumers do not expect to pay, and have not in the past, paid for bandwidth used to deliver their subscription TV services.

Debate and confusion are likely to grow, as legacy linear video providers increasingly move into on-demand, over-the-top services themselves. Among the reasons for confusion: the same physical facilities and bandwidth can support internet access and managed services alike. That is virtually always the case for single-fiber access (fiber to the home) using one wavelength and time division. In principle, the services could be logically and physically separated using two or more wavelengths on the same fiber.

In fact, it is clear that the existing TV subscription business model would not work at all, in most cases, if consumers had to pay for bandwidth charges in a direct sense. The implied full cost of a video subscription--assuming half of all account bandwidth is consumed when watching subscription entertainment video--could be higher by about half the cost of the whole internet access subscription.

That might range from a minor annoyance to a bit of a problem on a fixed network, where usage caps are generous and per-gigabyte charges are low. The same cannot be said for mobile consumption.

On a fixed network, a $100 video subscription, consumed on demand, would have to cost about $125, assuming a $50 fixed network subscription and use of half that bandwidth directly by the video subscription.

That might not be a model killer; just a key impediment.

Where the model breaks down almost completely is consumption of that same video on mobile networks, where bandwidth costs as much as an order of magnitude (10 times) more than on a fixed network.

If consumption of a mobile gigabyte represents about an hour to three hours of video, depending on resolution, then watching one hour of video could cost nearly $10 to $30. That is unsustainable.

If you want to know why some mobile video service providers consider  “zero rating”  so important, that is  why: if data consumption charges are required, the entertainment video model collapses.

There are lots of other ways the business model is affected. Bundling policies and access rights, for example, affect the business model, as when access providers require the purchase of one service to use a feature.

Comcast, for example, has released a beta version of the Xfinity TV app for Roku, allowing Xfinity TV customers the ability to watch their TV content on a Roku box. That fundamental principle--granting access to the streaming product when a customer has a linear product account, is fundamental for Comcast: a clear way to protect the legacy product while creating the new product.

Among the implications: access to the streaming product is a feature of the managed service subscription. Think of that as a strategy akin to adding steam engines to a sailing ship, a hybrid stage in the evolution from sail to steam.

Biggest Impact of Gigabit is Sales of Lower Speed Services, AT&T Finds

Success-based deployment of capital is one way access providers try and match incremental capital investment to incremental revenue. That is why firms from Google Fiber to AT&T build gigabit networks in neighborhoods, not whole cities; where demand for gigabit services and faster speeds is higher than average.

In its latest statements about take rates where it is building its fiber-to-home networks, AT&T suggests it is finding what other internet service providers have tended to find, when offering a range of speeds. Among the primary effects of launching gigabit service is that it spurs buying of services at lower speeds (40 Mbps, 100 Mbps, for example).

“After we launch our 100-percent fiber network in the new market, we're seeing about half of the new broadband customers buying speeds of 100 megabits per second or higher with 30 percent of the customers taking a gig,” says John Stephens, AT&T CFO.

In other words, 70 percent of customers buy speeds other than a gigabit per second, when it is possible for them to buy a gigabit access service.

"Free" is a Powerful Price Point

“Free” is a powerful price point, as shown by mobile account additions in India in October 2016, where Reliance Jio is grabbling most of the account growth in the whole market. It is a reasonable question how many of those customers--which seem to include a goodly number of mobile customers who added Reliance Jio as a secondary provider--will keep those accounts once they have to start paying for service.

Reliance Jio is going to make it very attractive to switch, continuing to offer very low prices, once the promotional pricing period ends in March 2017.

source: TRAI

Some Telecom "Moats" Might Actually Exist

An installed base of internet access customers is not necessarily a “moat” protecting an internet service provider from losing customers, but it really helps. So, apparently, do service bundles that offer more value and bigger discounts.

In mature telecom markets, customers simply do not switch providers all that often.

According to a 2014 study by Ofcom, the U.K. communications regulator, just eight percent of adults fixed network voice customers in the last 12 months. About nine percent of broadband access customers switched in the last year.

Just six percent of mobile customers switched providers over 12 months, while just five percent switched their subscription TV provider over the same time frame.

What that means, for any mobile service provider in a mature market is that only about one half of one percent of current customers chose another provider in any given month.

Likewise, Parks Associates consumer data show that almost 50 percent of U.S. mobile phone service customers did not change providers over the last 10 years. In other words, fully half the customer base virtually never changes providers, meaning that all switching behavior is concentrated on just half the total subscriber base.

According to Parks Associates, about 25 percent of respondents changed service providers only once in 10 years.

According to a 2016 study by U.K. comparison site uSwitch, only 11 percent of internet access customers switched providers over the last year.

Some 22 percent of customers report they have not changed providers in over five years, while 35 percent have never switched providers.

That one-year level of  churn is roughly the same percentage of U.S. mobile customers who switch from AT&T or Verizon Wireless over a year’s time as well, providing more evidence that, in a mature market, customer defections are less common than most might believe.

Also, most studies suggest that customers buying bundles churn less often, as well.

In the case of fixed internet access, it appears consumers also resist switching providers because they do not want an interruption of service while new providers install the new service.

Fully 35 percent of U.K. internet users who’ve experienced a period of internet access when they tried moving providers in the past say the thought of being without internet has put them off doing it again, according to uSwitch.com.

Of the 55 percent of respondents  who reported being without broadband between providers, the average length of downtime is 1.4 days. But 10 percent reported downtime of one to two weeks without service, while six percent had to wait longer than three weeks.

Internet users in London wait an average of 2.3 days for new service to start.

How Much Connection Revenue Will IoT Drive?

By 2020, nearly two billion devices with mobile connections capability, 2.7 billion devices that can connect to one or another low-power communications network and 3.3 billion devices with Wi-Fi capability will be shipped, or more than eight billion total devices in one year, according to IHS Markit.

If correct, and if all those devices are connected to networks, then perhaps 4.7 billion devices will likely represent new “public” potential network connections. The 3.3 billion Wi-Fi connected devices will mostly use existing public network connections.

Public network revenue will feature lower average revenue per device than public telecom network services providers have been used to when serving “humans,” however. If a mobile phone account might represent roughly $40 per device in monthly revenue, an IoT device might represent about $1 or $2 a month or less, depending on the amount of data any particular sensor has to transmit, over a month’s time.

That is one important reason why larger tier-one telcos will be focusing on IoT systems, services and platforms, not simply connectivity. Most of the revenue will be in devices, software and applications, not connectivity.  

source: IHS Markit

source: Cisco

source: Jasper.com

Telecom Failure Within 10 Years is Possible if Carriers "Do Nothing"

The telecom industry is on course to becoming “unrecognizable” within 10 years in a “do nothing” scenario, say researchers at McKinsey. That might seem alarmist. It is not, McKinsey seems to argue.

Over the past five years, the telecom business has entered a period of slow decline, with revenue growth down from 4.5 percent to four percent, EBITDA margins down from 25 percent to 17 percent, and cash-flow margins down from 15.6 percent to eight  percent, McKinsey says.

Just a few years ago, over-the-top messaging represented nine percent of revenue, OTT fixed voice 11 percent and OTT mobile voice two percent.

Forecast alternatives prepared by McKinsey & Company suggest that, in the most aggressive scenario, OTT messaging could be 60 percent share, OTT voice could be 50 percent of fixed voice revenue and 25 percent of mobile voice revenue.

In the “best” outcome, OTT messaging share of revenue would be 40 percent, share of fixed network voice 25 percent and mobile voice share could be seven percent of total.

Growth will continue globally, in newer markets. An additional billion middle-tier customers will likely be added, mainly in emerging markets, by 2025. But  revenue growth and profit margins will be low, McKinsey says.

Globally, the compound annual growth rate (CAGR) for traditional telcos is estimated at only 0.7 percent through 2020. For many telcos, largely in developed markets, the outlook is worse.

Telcos in Western Europe and in Central and Eastern Europe are facing –1.5 and –1.3 percent average growth, respectively, over the next four years, while those in North America are expected to barely tread water with growth at only about 0.3 percent, McKinsey says.

So twin challenges lie ahead: to “create a super slim and efficient core business” and then “to strategically define and aggressively pursue growth areas.” That is why internet of things, machine-to-machine services and even artificial intelligence or deep machine learning will be so important: they are among the key ways to find growth and also run the core business efficiently.

Network technologies need to become IT-centric and more software driven, allowing service providers to reduce baseline costs by 30 to 70 percent. So do not be surprised if carrier capital investment declines, even as next-generation networks are built. The new networks will cost more than the older networks, so the dramatic cost savings will be necessary.

source: McKinsey

source: McKinsey