Monday, June 27, 2016

Rent Rather than Own: First Music and Video, Then Airbnb and Uber. What Next?

It is not yet clear how far the "rent rather than own" trend can spread. Cloud computing represents one aspect of the trend, allowing enterprises and content or app providers to rent computing services rather than owning computing hardware and software.

Public Wi-Fi hotspots might be cast as a form of renting, rather than owning, Internet access.

In consumer markets, audio and video content always has had a "rent" rather than "own" character for most of its existence. Prerecorded video had a big upsurge in the 1980s, but not has fallen hard, and now prerecorded music is suffering.

Airbnb has shown people have a willingness to commercialize latent "lodging" assets, while Uber shows a similar willingness to commercialized latent transportation assets. How many more industries might be similarly affected is the big issue.

Some might argue almost any industry where zero marginal cost economics can be obtained, will be subject to "Uberization." One way of looking at it is that such markets can be created where there is demand for some product that must be fulfilled in real time, by people using smartphones to signal that demand, and where smartphones (and the backend processing) allow assets to be moved to fulfill, in real time. 

Almost any digital content product fits the bill.  On-demand transport works. So the issue is what other on-demand fulfillment is possible, in how many other segments of industries.

In 2015, the U.S. radio industry generated about $17.4 billion in revenue from advertising, while prerecorded music, streaming services and downloads produced about $9.4 billion. In other words, streaming, downloading and packaged media generated about 35 percent of network-delivered music revenue (not including live performance).

What the data also suggests is that music “ownership” in a classic sense (albums, discs, downloads) is disappearing. Some would say that is an example of people choosing to rent rather than own, much as it might be argued people now spend increasing amounts of money on video streaming services, compared to buying DVDs or downloading content to own.


Fitch Ratings Sees Longer Term Constraints on ISP Investment Because of Common Carrier Rules

The U.S. Court of Appeals for the District of Columbia has upheld Federal Communications Commission rules on common carrier regulation of Internet access services, but an appeal is certain.

Fitch Ratings says “we believe there will be very little near-term effect on revenues or operating profits from existing services” as a result of the ruling, because of the certain appeal.

But Fitch also believes that unless the rules are overturned, the major telecom and cable operators are likely to constrain investments in potential new growth areas affected by the net neutrality rules.

“Ultimately, Title II rules could change the way Internet traffic is managed, as well as affect future revenue opportunities and business models,” Fitch Ratings argues.

Some analysts, and Internet service providers themselves, have argued that common carrier rules imposed on U.S. Internet access services would depress investment. Critics have argued that would not happen.

At risk are not simply common carrier rules that could include rate regulation and prohibition of some business models.

Title II regulation--despite the FCC’s current “forbearance” on imposition of wider common carrier rules--opens the door for much greater regulation of the Internet in the future.

As presently applied, the common carrier rules do not impose  rate regulation, tariffs and last-mile unbundling, but a future commission could decide to enforce the provision.

Additionally, the FCC believes it can, under those rules, ban "zero-rating.”

At a high level, Fitch Ratings says, a long term negative impact of common carrier regulation is “reduced opportunity for wireline or wireless operators to benefit from potential new business models that deliver targeted advertising.”

The winners? Google and Facebook.

Will Google Launch Own Line of Smartphones?

Channel conflict is an almost-inevitable occurrence in any industry where sales, manufacturing and intellectual property are distinct segments of the value chain.

In the Internet value chain, it frequently happens that app, device and service segments of the ecosystem are in some state of tension. That happens most often when leaders in one portion of the value chain conclude they can profit by entering other parts of the value chain. So far, Google is the foremost example, operating in the operating system, application, device and access services portions of the value chain.

But Google might be aiming to take a bigger role in mobile devices, a step some have suggested would be required if Google was to integrate end user experience on the Apple model, amidst what every observer would say it a tendency to fragmentation of the Android market.

If Google does enter the market for branded smartphones, the move will be risky, not just because of execution issues, but because it causes channel conflict with partners using the Android operating system to power their own brands of smartphones.

Never has the phrase "coopetition" been more apt. Up to this point, Google has run the Nexus program largely on the basis that it showcases what Android can do, on a handset built to showcase operating system features.

Now Google might be thinking of producing and marketing its own branded smartphones, on the Apple model, essentially. So Google might gamble that the damage to its operating system efforts will be offset by the upside from a device and apps ecosystem that actually showcases and optimizes Android operation on compatible hardware.

Smart Cities Business Models Still Elusive

Business models always are among the earliest tasks whenever new technology is coming to market, and that will not be different for smart city initiatives. Though some trials can take place without a clear business model--generating revenue or avoiding cost--sustainable mass scale operations will require some clear idea of how new revenues can be generated or existing costs avoided.

And while proponents will avoid making arguments that are essentially “build it and they will come,” that manifestly is where matters presently stand, as a large amount of new revenue or cost savings are hard to describe.

And while reducing air pollution, improving parking or reducing traffic congestion are worthy social and governmental goals, none of those goals are going to be met without some definite and sustainable ability to pay for the infrastructure and operations, unless tax revenues are so abundant such goals can be met by taxpayer subsidies alone.

Early on, most observers would guess that a smart city infrastructure could generate revenue or save cost, often from multiple sources, in the areas of traffic management, parking, possibly Internet access, advertising or power savings are feasible.

What remains is to demonstrate that the aggregate new revenues and savings are sufficient to justify long term operation, as it likely remains the case that there is, in fact, still no killer app.

That does not mean one will not be found, or that some combination of values will be sufficient to drive and sustain investment. But we are far from having such knowledge.

Juniper Research, for example, suggests incremental revenue--globally--will be generated by
two million million smart parking spaces in operation by 2021. Skeptics quickly will suggest that global infrastructure cannot be contemplated or sustained--even in dense urban areas--by revenue from two million parking spaces.

So most proponents will argue that multiple revenue sources will develop. All that is possible, but also highly complex, suggesting a classic “chicken and egg” problem. Extensive and expensive infrastructure is required for many different business models to develop, but the existence of those revenue sources is needed to justify the investment.

Saturday, June 25, 2016

Facebook and Google Now Looking Both at Rural and Dense Urban Internet Access

Terragraph is Facebook’s new effort to develop a high-capacity, fixed wireless Internet access system running on 60-GHz unlicensed spectrum.  Google, for its part, also has invested in fixed wireless.

At the same time, both firms are testing use of unmanned aerial vehicles for satellite Internet delivery, while Google’s Project Loon is testing use of balloon fleets for the same purpose.

Verizon and AT&T also are looking at fixed wireless.  

Facebook’s Telecom Infra Project meanwhile seeks to develop open source and therefore lower cost network elements and platforms for traditional telecom networks. And Google has other efforts underway in the access network infrastructure or commercial access business.

All those efforts have a common theme, but also distinct missions. Since both Google and Facebook have advertising-based business models, anything the helps connect billions more people creates the revenue opportunity both firms are based upon. In other words, Internet access and Internet users directly create the revenue opportunity.

For that reason, there is direct alignment of “Internet access availability” and “our revenue opportunity” for Facebook and Google.

At the same time, there are nuances. Terragraph and Google’s fixed wireless efforts seem primarily aimed at dense, urban areas. The Facebook Aquila UAV effort, as well as Google’s similar effort, plus Project Loon, are aimed primarily at rural areas.

Facebook’s current satellite delivery effort in Africa also is aimed primarily at rural users. Telecom Infra Project might ultimately provide lower costs in either mobile or fixed realms.

And, for some time, lots of app providers have been pushing for use of more unlicensed spectrum, as well as spectrum sharing, as ways to bring more low-cost communications spectrum to market.

Telcos are likely to prefer use of licensed spectrum. So will cable companies in mobile, eventually. But lots of other app providers are likely to look for ways to leverage lower-cost unlicensed or shared spectrum to underpin Internet access efforts, using several business models.

At least in part, Google has chosen to go direct, becoming an ISP, a mobile services and Wi-Fi hotspot supplier. Project Loon would act as a commercial partner for 4G mobile networks, but also is a commercial operation.

The other business model is open source. Facebook already has done so in the data center area. Now it is working to bring an open source approach to telecom network infrastructure as well.

Facebook, for the moment, continues to insist it does not want to become a commercial provider of Internet access. Indeed, its approach has been to create new platforms that take an open systems approach, usable by any entity.

The Telecom Infra Project already has major support from telecom service providers and network infrastructure suppliers. Presumably, Project Aries and Terragraph would be available for other suppliers to sell commercially.

Google’s UAV effort relies on partnerships with mobile operators who have licenses for spectrum Project Loon will use.

In some quarters, the bigger question is whether more firms such as Google might eventually decide there is a business model beyond open source, research and development. As telcos represent on access platform, and cable TV uses a different platform, might others eventually use additional distinct platforms, including networks based on use of unlicensed and shared spectrum?

Friday, June 24, 2016

No New Revenue for 5G?

From time to time, you will hear a consultant or market researcher make a startling claim about the next generation network (fixed, mobile or wireless), namely, that there are “no new revenues" to be generated by the new network.

It is a hard claim to refute in its entirety, for some logical reasons. Telcos replacing copper access with “fiber to home” or “fiber deep into the network” already know the issue.

If a service provider has customers served by a digital subscriber line connection, and then replaces it with a fiber-to-home connection, there is a net zero change in subscription volume: one replaces the other.

One might also make the argument that there is “no net new revenue” involved, but that likely is wrong for several reasons. That might sometimes be the case.

Quite frequently, a higher-bandwidth connection is purchased, at higher cost. Also, the fiber connection supports more-reliable video entertainment services (linear and streaming).

Or consider the matter of mobile next generation networks. Did 2G represent “no net new revenue, compared to first generation analog?” In one sense, over a period of time, voice service revenue possibly decreased, as prices for mobile service dropped.

Also, over a period of time, 1G voice revenues became 2G voice revenues. Sure, it is tough to calculate net revenue changes due to the network upgrade, as opposed to market price changes, which likely were more substantial.

But 2G made text messaging possible, so eventually, incremental new revenue was generated. Also, as voice prices dropped, consumers began substituting mobile voice for fixed line voice, increasing the number of subscriptions. So, yes, a next-generation network most often does lead to higher revenue, over time.

Some might argue 5G will not lead to higher revenues, compared to 4G, or that Internet of Things will not boost revenue.

It always is difficult to fully anticipate the business value provided by each successive generation of mobile networks. There always is a stated business case, of course. From the first generation to the second, the advantage was the transition from analog to digital, with the advantages that normally represents.

The shift from 2G to 3G was supposed to be “new applications.” That eventually happened, but not right away. First mobile email and then mobile Internet access were new apps of note, though the use of mobile hotspots also was an important development.

The shift from 3G to 4G generally was said to be “more bandwidth” supporting new applications.

Video apps generally have been the most notable new apps, compared to 3G, although user experience when using the Internet also is far better with 4G. And though it often goes unnoticed, 4G speeds have allowed any number of users to substitute mobile for Internet access.

More U.S. households now seem to be abandoning even fixed Internet access in favor of mobile access, as it now is common for households to rely on mobile voice (more than 46 percent of U.S. households now are “mobile only” for voice) , instead of fixed network voice, or over the top video entertainment in place of traditional subscription services.

In fact, because of mobile use, fixed network Internet access rates actually are dropping in the United States, having reached an apparent peak in 2011.

Eventually, skeptics will be proven wrong. In the early going, one might well argue that the incremental cost of running both 4G and 5G, plus the new capital investment, plus substitution effects (customers move from 4G to 5G) might not--initially--show net new revenues or earnings.

Over time, that will change, as 5G becomes the default network and customers migrate over. That, of course, means the 4G network is less-heavily loaded, and therefore drives less revenue, unless a new temporary purpose can be found for 4G that is distinct from 5G.

Eventually, the 4G network reaches end of life and is decommissioned. But there is a long period when multiple networks coexist, with evolving business models and profitability, over time.

That is not to say the discovery of new revenue models happens right away. Frequently that does not happen immediately. Some of us would argue it took some time before new apps--such as mobile email--actually became interesting new 3G apps and revenue drivers (direct or indirect).

It is too early to say, with precision, how new millimeter wave spectrum allocations around the world, to support 5G mobile networks and other applications, will increase the total supply of wireless communications bandwidth, spectrum and business models.  

Some might guess that the new millimeter wave capacity will equal or exceed all prior wireless allocations yet made. There are tradeoffs: distance for capacity, generally speaking.


Thursday, June 23, 2016

Can Service Providers Really Affect Customer Satisfaction Enough to Warrant High Investment?

Some things are hard to change. Perhaps some things are nearly impossible to change. So, as a matter of business priorities, how much should communications service providers invest in efforts to improve customer satisfaction? Maybe the answer is “not so much.”

Ecosystem providers who make a living selling solutions that are supposed to raise consumer satisfaction obviously will disagree. But you have to question the wisdom of spending too much to change consumer perceptions that might not improve all that much, no matter how much is spent.

Historically, few service providers and services ever rank anywhere but towards the bottom of consumer satisfaction surveys. Airlines typically have the same problem.

That might not be a good thing, but the amount of change that actually is possible is arguably questionable. In other words, beyond a certain minimal point, it might not be wise to invest too much in trying to improve consumer satisfaction scores, because there might be a sharp limit to the payoff, in terms of investment, compared to many other business processes that also require investment.

That is especially important when resources are limited and when a small number of actions arguably produce the bulk of value for any access service provider. It just makes sense to focus on the relatively smaller number of things that actually can influence business results.

And it arguably is hard to prove that investing in better customer satisfaction actually will produce very much change in business fortunes.

A Frost and Sullivan report commissioned by IBM argues that, by 2020, customer experience will overtake product and price as the key brand differentiator across all industries.

Access providers are “behind the curve,” Frost and Sullivan argues. “Research indicates that only one in six customers is an advocate for his or her CSP, and just 12 percent strongly agree their CSP listens to them and collects the right amount of information to meet their communications needs.

To be a contrarian, that might well be true. But with all the other problems access providers have, just how much should be invested to improve perceptions on such matters?

We have yet to see any evidence that a telco, cable company, mobile services provider or Internet service provider is able to score anywhere above the bottom half, or maybe even the bottom third, of industry consumer satisfaction ratings.

It is possible, in other words, that people simply are not satisfied with communications services for reasons that defy easy remedies. While it might be nice to outperform peers on satisfaction measures, it also is not clear that such relative advantages produce sustainable or measurable revenue lift.

Harsh, but historical data offers little support for the notion that an access provider is “loved” by its customers.

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...