Tuesday, May 31, 2016

What is the Killer App for Smart Cities?

Some now believe smart cities will be an early and substantial market for Internet of Things apps and services. The problem is that there are any number of practical implementations, without clear and sustainable business models, yet.

Of course, nobody yet knows. It is conceivable that connected car, wearables, health or agriculture and manufacturing or even home automation theoretically could emerge earlier.

Parking, air pollution, traffic management, monitoring of water pipe leaks, streetlight management or wastewater management are some of the areas believed to be fertile ground for smart cities initiatives. Others might argue less-exotic implementations, such as municipal Wi-Fi, also count as “smart cities” programs.

But there are many sources of inertia, including unclear payback or business models, technology platform confusion and the cost of equipping end users and networks with sensor capabilities. It is not clear what sustainable revenue models look like, nor are there absolutely clear killer apps to drive massive adoption.

Some amount of pump priming, in the form of government grants, might help, in a few cases. Still, beyond installing a general purpose communications network, it is not so clear that there are synergies between the various proposed “smart city” capabilities.

source: Compass Intelligence  

Banks Warming to Cloud Computing?

Large financial institutions traditionally invest heavily in new technology, but also are “conservative” about security issues and control. That some now seem willing to embrace moves to cloud computing might suggest change is coming.

Capital One is already in the process of shutting down five of its eight private data centers to move most of its data to AWS by 2018. Another "large" bank cited in the note said it's committed to shutting down all of its private data centers by 2020, although it's more likely to start with a hybrid of cloud and on-premise data centers, according to the note.

These kinds of changes will only speed up the adoption of the cloud by other sectors, too, expanding the overall pie for public-cloud vendors. Gartner estimates spending on public-cloud services to increase from $85 billion in 2014 to $180 billion in 2019, according to a note by Bank of America.

source: IDC

Verizon Will Not Wring Blood from a Stone

You cannot, as the aphorism suggests, wring blood from a stone. Neither, history suggests, can an industry in fundamental decline afford to boost its own costs of operation. To have any hope of maintaining cash flow as long as possible while the business declines, firms have to control costs.


And, as always within any ecosystem, one segment’s “costs” are another segment’s “revenue.” That is rarely pleasant, for workers, suppliers, collaborators, investors or managers, in a declining business or industry.


The new Verizon contract that ended a strike might provide an example. The new contract adds 1,400 jobs and provides a 11-percent wage increase. Union officials argue that Verizon has been understaffing fixed network workforces, a charge that resonates.


Verizon has been moving spending towards mobile, and away from fixed network operations, to match its revenue generation.


And wages for U.S. workers have been depressed for quite some time. Workers think they deserve more of the surplus. Verizon management knows it has to "harvest" the business. Both have a point.


Only in context do those developments make sense.


If you wanted a one-sentence description of how the U.S. fixed network business has been transformed over the last 15 years, here it is: “Wireline now accounts for less than 30 percent of Verizon’s total operating revenues, down from 60 percent in 2000, and less than seven percent of our operating income,” noted Verizon Communications CEO Lowell McAdam.


In other words, from 2000 to 2014, revenue from fixed networks was cut in half, while profit dropped more than that. At a very high level, that means any business in a similar situation would have to chop at least half its costs. And, unfortunately, enterprise cost includes employee headcount,  wages and benefits.


In the first quarter of 2016, Verizon fixed network revenue was down, while earnings were flat. To be sure, weakness in wholesale and global enterprise were bigger problems than the mass market segment, lead by FiOS revenue.


The point is simply that fixed network operations are generating a small, and likely ever-smaller, portion of Verizon revenue and profits. Under such conditions, harvesting is a rational strategy.


There is nothing nefarious about this. It is simply what a business must do when it is in decline.


In that light, a 10 percent or 11 percent hike in wages and benefits, plus the addition of more employees, will create greater pressure in the fixed networks business.


Indeed, some might speculate that selling or spinning off most of the fixed network business might make sense, if willing and able buyers can be found, and if regulators allow such dispositions.

Verizon would be better off as a smaller, mobile-focused entity, the logic suggests.

Some entities can operate FiOS or fixed networks more efficiently than Verizon can. Some entities might have different business models as well, allowing them to wring more value out of assets such as Verizon's, if they are free to tweak the cost model.

Unfortunately, employee prospects in a declining industry are a zero sum game, at best. Most often, negative growth is the requirement. If one accepts that Verizon's fixed network business is in decline, then more employees and higher wage bills are not sustainable.

It will not be pleasant. It never is.

More Cloud Means Fewer Moves, Adds, Changes: Bye Bye MAC Revenue

Disruption across the information technology and communications businesses is a virtual certainty as cloud-based computing grows.

And that applies equally to sales channels, software platforms, hardware suppliers and services operations. For decades, revenue models in some parts of the business have been built on “moves, adds and changes.”

As enterprises, mid-market and small business customers move operations to cloud-based alternatives, the revenue formerly earned by MACs can diminish.

Hutchinson Networks, a U.K. systems integrator and professional services provider, provides a case in point.

The fundamental problem: enterprises won't need systems integration and professional services if they shut down their IT operations and move to the cloud.

One example: Hutchinson used to support SMEs with their own data center operations, to support Microsoft Exchange, Active Directory, Sharepoint and phone services.

Increasingly, those businesses are going to Office 365, and don't need on-premises computing. That also means they do not require the support for owned facilities and systems that Hutchison Networks used to support.

The fundamental problem for some IT support operations is that more cloud computing means fewer MACs. And that means a diminished  business role for IT services companies.

Android N IVI for Connected Car

source: GSMA
Leverage in the connected car market is not at all obvious, and a new move by Google shows why. Any “in vehicle infotainment” system requires an operating system or platform that allows in-car systems to communicate with mobile networks.

source: Business Insider
As you might guess, there are many rivals in that area. Google and Apple, BlackBerry, Linux, GENIVI and Tizen are among the proposed OS approaches. Auto OEMs can build their own or source a platform.

Google already offers Android Auto. But it also offers Android N IVI, which allows the Android smartphone itself to supply the mobile network link and intelligent car functions.

Android N adds support for AM/FM radio, HVAC controls, Bluetooth links between the OS and the car’s dialing system, similar controls for media streaming, and the option for digital dashboard instrumentation clusters.

source: GSMA
Of course, and Android N IVI system will work closely with Android Auto, but that same functionality could presumably be achieved by simply syncing the driver’s Google accounts between the phone and the car, to share contacts, bookmarks and media.

To the extent that Android N IVI enables tethering, the size of the tethered and smartphone-based market segments will matter.




Why Google is in the Internet Access, Mobile, Linear Video and Voice Businesses

In addition to Google Fiber, which has Google operating as a fixed network provider of Internet access, linear video entertainment and soon, voice, Google Fi is a commercial mobile virtual network operator in the U.S. market.
But those efforts are not the only initiatives Google parent Alphabet has under way. After dabbling in municipal Wi-Fi and other Wi-Fi networks (train stations in India, for example), Alphabet also has several wireless initiatives under way.
Project Loon entails use of a balloon-based mesh network in the sky, using Long Term Evolution 4G signals to directly reach end user mobile devices. That approach necessarily entails working with mobile operators who are willing to allow Project Loon to use licensed frequencies.
At the very least, that entails Project Loon becoming a wholesale customer of one or more mobile operators in a market, and then acting as both backhaul and access network.
The precise nature of retail relationships might change over time. Initially, Project Loon might function mostly as an infrastructure partner for branded mobile operators who would have the retail relationship. But other business models are feasible, so it is just too early to say what might evolve later.
So far Project Loon  has tested such services  with Vodafone New Zealand, Telstra in Australia, Telefónica in Latin America, all three mobile operators in Indonesia and with Sri Lanka as well.
Since Google has asked for permission to test the concept in the United States as well, there is some speculation that Project Loon might actually operate over rural parts of the United States as well as many countries in the southern hemisphere.
Project Titan is the unit of Alphabet working on unmanned aerial vehicles, and Alphabet also is an investor in Elon Musk’s SpaceX, presumably because SpaceX plans to launch a huge fleet of low earth orbiting satellites for Internet access.
Some might be skeptical about how much success Alphabet might have in all of its various ISP and service provider roles.
But the fact remains that Alphabet is investing serious amounts of capital in a variety of programs that involve competing with cable TV, mobile, fixed network telco, satellite and ISP interests on a possibly substantial level.
Though some might question the priorities, Google cares about Internet access and app or web page loading speed for very practical reasons. Every incremental Internet user is a potential user of Alphabet and Google apps.
And the faster content loads, the happier users are, while increasing ad inventory at the same time. For a company that earns the bulk of its revenue from advertising, that matters. Facebook cares about universal Internet access and access speed for the same reasons.

Monday, May 30, 2016

OTT Revenue is Growing Fastest in the Mobile Ecosystem


By 2020, Mobile Service Providers Will Have Lost 10% of Ecosystem Revenue; App Providers Will Have Gained 10% More

It is getting harder and harder, all the time, to describe what is happening in the mobile ecosystem, for the simple reason that the ecosystem now includes truly-significant revenue from application, device and service providers.

Mobile service providers generally earn about 60 percent of identifiable mobile ecosystem revenue, including network infrastructure; components; apps, content and advertising; devices as well as access.

Over the top app providers earn about 10 percent of ecosystem revenue while device suppliers earn about 22 percent of ecosystem revenue.

GSMA believes that mobile access provider revenue share will decline to perhaps 50 percent of ecosystem revenue by 2020, while device share remains about the same and app revenue might grow to 20 percent of total ecosystem revenues.

In other words, GSMA predicts that the biggest changes will have mobile operator share of ecosystem revenue share declining about 10 percent, while app providers gain about 10 percent share.

The caveat is that the percentages could change more drastically should one or more app providers also become key factors in the access business.  

80/20 Has Many Implications

One should not underestimate the importance of the Pareto theorem, commonly known as the “80/20 rule,” where 80 percent of results come from 20 percent of instances.  

One example are financial returns from the Standard and Poors 500 index, over the period 1989 to 2015, when just 20 percent of stocks accounted for 100 percent of index gains.

Put another way, 80 percent of stocks in the index actually had a collective return of zero percent.


In communication or other markets, the Pareto theorem matters because it tends to describe the broad structure of markets, as well as the generation of revenue and profits within each market.

In the Indian mobile services business, just 17 percent of customers generate 60 percent of revenues, for example.

The fundamental principle is that effort and outcomes are non-linear. A small number of inputs or instances drive most of the outputs or results.

The practical implication for communications or app providers is that a relatively small number of decisions and priorities actually matter, where it comes to making a transition from legacy to next generation business models.

The corollary is that there are a many “good” or “useful” or “helpful” things any service or app provider can do, but which should not be done, to concentrate on the few areas where breakthroughs are possible.


The Pareto rule can guide resource allocation, the principle being that there is some allocation or resources that makes a person or an organization better off, while not harming existing persons or the organization itself.

In other words, at some point, additional effort will produce diminishing returns.

The Gini coefficient essentially follows Pareto distribution patterns as well, and describes national income inequality patterns as well.

In the United States, the number of homes without a broadband connection follows a Pareto distribution.


It illustrates the law of diminishing returns. The cost of building access loops generally follow Pareto rules, for example. The inverse of the Pareto distribution is that a small number of instances produce most of the “per-line” access cost.

In other words, a small number of remote locations represent a disproportionate share of network cost, based on cost per mile.



In practical sense, the Pareto theorem suggests that a small number of actions actually drive most of the actual organization results. 

In other words, the temptation to “do” any number of helpful things actually can be detrimental to strategic success, which requires intense concentration on a relative handful of decisions, investments and effort.

There are always lots of useful or helpful things a company might do, to support its business. Many of those things actually will deliver a measurable result. But most will fail to help a company make a strategic breakthrough.

So saying “no” to most of those helpful things can be a prerequisite for focusing effort on a few matters that can decisively change a company’s future.

Multiple Networks Needed to Support Smart Cities, Says GSA

source: ABI Research
Perhaps predictably, the GSA has authored a white paper on the central role of communications in smart city efforts.

Also predictably, the GSA, representing the interests of platform suppliers of many types, supporting rival IoT platforms, argues that an “all of the above” approach will be needed.

“Multiple networks are required to underpin the smart city,” GSA argues. “A cellular network is highly unlikely to be able to deliver appropriate connectivity for every smart city application, even if it is able to satisfy many requirements.”

“Even with the emerging narrowband IoT (NB-IoT) networks and partnerships with competing IoT LP-WAN providers, for the foreseeable future, a smart city will use multiple network technologies,” GSA argues.

source: GSA
An appealing a political statement for an organization that has to support all its members, the position also is unlikely to happen, on a wide scale, when IoT services and networks actually are commercially deployed on a mass scale.

By definition, and excluding the potential role for mobile networks, most of the proposed IoT networks are rivals and substitutes, featuring low data rates and low power consumption.

The GSA’s argument is not without a rationale. Underpinning every IoT service, application and network is a physical transmission layer of some sort.

What is less clear are the business models and “killer apps” that will dictate what services make sense and are sustainable, and what mix of public and private investment will be possible.

But at least some forecasters believe “smart cities” apps will be a leading segment of the market, driven at least in part by traffic management and other infrastructure concerns. Almost certainly, those hopes will fail to match the optimism.

The reason is simply that the “smart cities” category is an amalgam of many different potential markets, across many industries, some amenable to public investment, while others are contingent on investment by many private entities.






Friday, May 27, 2016

Hosted PBX or Hosted UC Might be Substantially Uncompetitive at Firms Representing Nearly 70% of Workers

When market researchers aggregate a whole bunch of different revenue streams together, and then call them all “one market,” it is a clear sign that each of the constituent markets is small. Only by aggregating can a market of reasonable size be shown to exist.


That arguably is the case for a host of services collectively lumped into the hosted UC or hosted PBX category.

In 2008, the combined business phone system, conferencing, messaging, unified communications, contact center, collaboration, rich media, email and calendaring markets collectively amounted to perhaps $29 billion globally.


By 2014 the market perhaps had shrunk to perhaps $19 billion, according to some estimates.


In fact, despite hyperbole about “huge new markets,” the truth is that the broad UC and IP communications businesses are not that big. In fact, one might argue that traditional business voice systems are a flattish business, while all manner of software-based services might be where the growth is happening.


Bottom line, is the market still growing? Probably, but only because so many categories are lumped together. As sales of business phone systems fall, hosted alternatives still seem to be growing, if rather slowly, on a global basis.


Hosted PBX revenues globally might only be in the $10 billion range, meaning the market is small, in most countries. If the U.S. market represents half the world total, then hosted PBX represents $5 billion or so in annual revenue. That is significant for some specialists, but a small bogey for the whole U.S. service provider market.


Also, a significant portion of reported “hosted UC” or “hosted PBX” revenue actually consists of SIP trunking, an access service, not a “voice” service. By some estimates, North American hosted PBX revenue--subtracting out the access portion--might be about $4 billion annually.


The ultimate issue is total potential market size, and what percentage of that total hosted services can address. Some estimate the total addressable North American market is about 140 million business lines, for all solutions. So far, hosted services have gotten about 12 percent of that total.


It still might be useful to estimate what percentage of hosted services adoption is reasonable, since, at some point, it still appears that “owning” phone systems, rather than “renting” services, makes sense.


Most workers (more than 70 percent) are at firms that have been in business six or more years. Roughly speaking, you can assume that roughly 70 percent of demand for “lines” and services therefore also exists at those firms.


Nearly 40 percent of all U.S. employees work for firms of at least 10,000 employees. Those are the employees working at firms most likely to benefit from “owning” switches rather than renting services, for the most part.


At least so far, most hosted PBX adoption has occurred at firms with fewer than 50 employees, nearly 70 percent, by some estimates. Roughly 17 percent of hosted PBX adoption occurs at firms with 50 to 500 users, while 15 percent of demand has come from firms with more than 500 users.


And that points to an apparent limitation. Only about 25 percent of business employees work at firms with 100 or fewer employees.


If the economics of owning versus renting services do not change much--if it continues to make financial sense for larger enterprises to won their switches, then a substantial part of the business communications market will remain largely “off limits” for hosted services adoption.


It will continue to make sense for small, young and fast-growing firms to choose a hosted solution. The issue is simply that most people work at large firms where hosted alternatives are not necessarily cost competitive, whatever “soft” advantages (moves, adds, changes) exist.




                   Distribution of Total Private-Sector Employment by Firm Age


Distribution of Employment at 16+ Year Firms by Firm Size Fig. 8: Distribution of Employment at 16+ Year Firms by Firm Size (1992 v 2011)














Whatever you think might be happening in the broad “unified communications or business voice” markets, one fact has not changed over the last decade: as important as the segment is for some suppliers in the market, the categories are relatively small, within the context of the whole communications market, or even within the business communications segment.

"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...