What Has Changed for FTTH?

For more than two decades, U.S. cable operators have won the market share battle with telcos (net new additions) as well as the installed base battle (percentage of total customers). That appears poised to change, with telcos now believed to be possible installed base gainers. 

To accomplish that, telcos also would likely have to win the market share (net new additions) battle. We haven’t seen that in two decades (some might argue telcos never have won the market share battles) but it seems possible for the first time. 

So what has changed? Several things, probably. Some important tier-two telcos that had been capital constrained have now restricted to the point where they can afford to invest in new fiber-to-home facilities where they had not been able to, in the past. 

Tier-one suppliers also arguably have altered options. Verizon, which had largely halted FTTH deployments because of the business model, now sees different returns as a result of fiber deployment to support its 5G small cell deployments. One byproduct is a denser optical transport network that can change the incremental cost to provide FTTH. 

But market share or installed base can change in other ways directly related to that denser fiber transport footprint. In some cases, 5G fixed wireless can allow Verizon to gain share without full FTTH. If the issue is “bandwidth to the home” or “gigabit to the home,” then 5G fixed wireless might work, irrespective of the platform. 

AT&T has been deleveraging, and is the telco with the most room to upgrade its access networks to FTTH. 

source: Standard & Poors 

Lumen Technologies, on the other hand, recently divested itself of about half its total consumer access lines, to concentrate on its denser metro areas. 

It might seem paradoxical that perceptions of return from FTTH investment are higher than once was the case when three mass market services--each with high adoption--were possible with FTTH. With the decline of voice and linear video entertainment revenues, the fixed network business case for consumer services largely rests on internet access. 

Logically that should create a worse business case, as revenue mostly must come from a single lead application. But other parts of the revenue and cost model also are changing. Third party sources of funding sometimes are more lucrative (either from joint builds or bigger government subsidies). 

Divesting linear video reduces revenue, but also cost. Harvesting voice while concentrating on internet service provider operations might in some cases lead to lower operating costs. 

Also, though telcos failed to halt the slide in broadband market share over the last two decades, the growing need for more-symmetrical bandwidth now offers telcos a possible marketplace advantage over cable operators. 

Also, telcos increasingly are building models that rely on broadband for nearly all the financial return from a new FTTH build, based on steadily-improving efficiencies. Telcos with 5G backhaul networks now can leverage those other fiber transport investments to support consumer home broadband investments. 

Expectations about installed base share also help the new payback models. Where telcos once might have held only 30 percent share of the installed base, they now can reasonably expect to eventually take 50 percent share of the installed base, which changes the financial return. 

Up to this point the U.S. FTTH footprint has been rather modest. All together, telco FTTH probably today passes only about 29 percent of U.S. homes. That percentage will grow closer to half of all U.S. homes over the next five years or so. 

That still leaves telcos with a problem: they wil be able to sell FTTH-based gigabit services to only half of U.S. homes. What to do about the rest is the logic behind 5G fixed wireless. 

In 2021, for example, Comcast, the biggest U.S. cable operator, faced an FTTH competitor in less than 30 percent of its footprint. That obviously limits the amount of total share loss Comcast is exposed to, as cable trounces digital subscriber line platforms  in performance. DSL simply is not competitive with cable modem service. 

Then there are the strategic issues. Absent the upgrades to FTTH, can a fixed network service provider reasonably expect to remain in business? Increasingly, the answer is “no.” To the extent that internet access is the paramount driver of fixed network revenue, then FTTH either is installed or the telco faces bankruptcy. 

The argument then is not so much “we will make more money” as it is “we get to stay in business.” 

Are Happy Workers Really More Productive?

Are happy employees correlated with better financial returns? Are happy employees more productive? Most of us instinctively would agree with an article in the Harvard Business Review that “a decade of research proves that happiness raises nearly every business and educational outcome: raising sales by 37 percent, productivity by 31 percent, and accuracy on tasks by 19 percent, as well as a myriad of health and quality of life improvements.”

One new study--looking at firms deemed “best companies to work for” from 1984 to 2020, suggests it could be the case, at least for these firms. 

Other studies also suggest that happier workers are more productive, though the studies do not address the relationship between productivity and profit directly. 

The issue always is the “Hawthorne Effect.” The Hawthorne Effect refers to the tendency of some people to work harder and perform better when they are participants in an experiment. 

In other words, individuals may change their behavior due to the attention they are receiving from researchers rather than because of any manipulation of independent variables. They might report being more productive or happier simply because they are being studied. 

There are lots of other views, including the argument that happier people are more productive, irrespective of work conditions. In other words, they were happy independently of the work setting. 

Methodologically, one issue is that we cannot actually quantity “happiness” well enough to measure it. And some studies report just the opposite: unhappy workers also are productive. 

Also, the reverse might also be true: productive workers feel happy. It is productivity itself that drives feelings of happiness, not the other way around. 

According to Stephen P. Robbins, a careful review of the evidence finds the correlation between satisfaction and productivity in the range of +0.14 and +0.30. Essentially, no more than nine percent and maybe as low as two percent of the variance in output can be accounted for by job satisfaction self reports, which then is used as the proxy measure of happiness. 

So productive employees are more likely to be happy workers, rather than the reverse.    

There is evidence that more-profitable firms also have more happy employees, though it is hard to say whether the relationship is causal or only correlative. 

Highly-profitable firms are able to provide more benefits for their workers, more opportunities for promotion and nicer work environments, for example, so that alone might contribute to relative feelings of happiness. 

Most of us would instinctively believe that happier workers also are more productive. It might be the case. But it is difficult to prove the thesis. We need proxies for “happiness” that are relatively more objective than self reports. 

We need evidence that there is a causal--not merely correlational--relationship between work environment and “happiness.” We need to remove performance effects (Hawthorne). 

Also, there is little incentive for research about circumstances where “happiness” is unrelated or inversely related to productivity or firm profits, so that subject likely is under-researched. 

"You Get to Keep Your Business" is the Main Strategic Advantage of 5G

Lots of mobile industry executives have worried about 5G monetization. The real concern might not actually be so much whether there will lots of revenue, as there will certainly be.

The issue is more the cost side of the business model, especially capital investment to support dense backhaul networks and small cells. Even discounting new sources of revenue from network slicing, edge computing or internet of things, 5G revenue will mostly approximate 4G revenues.

Some will think that is an argument against deploying 5G. Strategically, 5G investment is necessary to keep the business a mobile service provider already has. In other words, lost market share, gross revenues and profits are at stake if the upgrade to 5G is not made.

That might seem like a paltry return, but that is the history of the mobility business. Every decade, a new next-generation network must be deployed, in large part simply to supply ever-increasing data demand and at the same time reduce the cost of supplying those bits.

Each mobile next-generation network is required as a primary means of accomplishing something else: generating new revenues to replace substantial lost revenues.

As a rule of thumb--as shocking as it might seem at first--service providers tend to replace about half of their existing revenue every decade. 5G provides one very concrete example, as did 2G, 3G and 4G. 

5G accounts will eventually replace half of all other accounts--primarily 4G but also 3G and 2G--over a decade. 

source: Statista 

The same process happened for analog mobile, 2G, 3G and 4G. By design, each successive mobile next-generation network displaces the older networks. A common pattern is for the latest generation to get about 50 percent share of accounts within five to six years. 

source: Strategy Analytics

The same process arguably also held for home broadband and fixed network internet access services. As typical speeds increase over time, customers upgrade. The product we called “dial-up internet access” was replaced by “broadband” access at single-digit or low double-digit rates. 

But speeds keep increasing. And customers gradually replace older services with newer services operating at higher speeds. 

source: Ofcom 

Access speeds  in the U.S. and other markets have grown by an order of magnitude (10x) every five years, since about 1990. 

source: iconnect007 

So, in a real sense, both mobile and fixed network internet service providers do replace at least half of all existing revenue every decade, looking only at a single product: consumer internet access. 

source: Openvault

For example, about a third of U.S. home broadband customers in early 2019 bought services operating at 100 Mbps to 150 Mbps. Two years later nearly half were buying services operating between 100 Mbps and 200 Mbps. 

source: Openvault  

In early 2019 some 54 percent of U.S. households purchased services operating at speeds less than 75 Mbps. By mid-2021, all customers buying services at less than 50 Mbps had fallen to 10.5 percent. 

Enterprise data services show the same trend. Where enterprises once bought X.25, they later switched to ISDN, also frame relay, then to T-carrier, followed by ATM, then witching to dedicated internet access and MPLS. Now many are moving to SD-WAN. 

The point is that the products enterprises purchased--and service providers sold--evolves over time. The same general pattern--replacement of half of current products--every decade also holds. 

So, to answer the question about where 5G revenue will come, it will mostly come from the same places it does now. There will be new revenue sources over the decade, including fixed wireless, sensor connections and edge computing revenues. But the mainstay will remain consumer and business mobile connections.

Sequential Hurdles Explain Low Rates of Transformation Success

A common rule of thumb is that about 70 percent of information technology projects fail to meet their objectives. Historically, most big information technology projects fail in some major way, failing to produce expected cost savings or revenue enhancements or even expected process improvements. 

Some would argue the digital transformation failure rate is the same. “74 percent of cloud-related transformations fail to capture expected savings or business value, ” say McKinsey consultants  Matthias Kässer, Wolf Richter, Gundbert Scherf, and Christoph Schrey. 

Of the $1.3 trillion that was spent on digital transformation--using digital technologies to create new or modify existing business processes--in 2018, it is estimated that $900 billion went to waste, say Ed Lam, Li & Fung CFO, Kirk Girard is former Director of Planning and Development in Santa Clara County and Vernon Irvin Lumen Technologies president of Government, Education, and Mid & Small Business. 

BCG research, for example, suggests that 70 percent of digital transformations fall short of their objectives. 

From 2003 to 2012, only 6.4 percent of federal IT projects with $10 million or more in labor costs were successful, according to a study by Standish, noted by Brookings.

Some industries arguably do better than others, especially consumer-facing businesses and industries. The success of e-commerce is one likely reason why that occurs. Compared to changing a whole business model, shifting sales partly to online mechanisms is relatively risk free.

Another common rule of thumb is that about the same percentage of digital transformation efforts also fail. Many reasons are cited. 

• Company Culture is not aligned

• CxOs do not really support it

• Silos

• Knowledge gaps about cause and effect

• Indecision or tepid initiatives

• Technology novelty not harnessed to business processes

• Expectations not in line with reality

• Not iterating fast enough

• Human capital mismatch

• Lack of continuity and consistency

• Unclear vision

• Business and IT execs do not agree on objectives

• Organizational inertia

• Lack of employee buy in

• Governance not aligned

Those concerns typically are found in every industry and firm that attempts DX, as this example from the banking industry suggests. 

 

source: Banking Circle 

There is a simple way of illustrating why innovation is so hard. If you have ever spent time and effort trying to create something new in the communications business, you know it rarely is easy, simple or uncomplicated to do so, and the larger the organization you work for, the harder it seems to be. That is because all organizational change involves power and politics, and changes will be resisted.  

There is a reason 70 percent of organizational change programs fail, in part or completely. Assume you propose some change that requires just two approvals to proceed, with the odds of approval at 50 percent for each step. The odds of getting “yes” decisions in a two-step process are about 25 percent (.5x.5=.25). 

source: John Troller 

The odds get longer for any change process that actually requires multiple approvals. Assume there are five sets of approvals. Assume your odds of success are high--about 66 percent--at each stage. In that case, your odds of success are about one in eight (.66x.66x.66x.66x.66=82/243). 

The more hurdles (approvals) required for a change to happen, the less likely the change will happen. Even when the odds of approval at any stage are 66 percent, the necessity of just five approvals will lead to seven of eight change efforts failing. 

source: BCG

Fermi is Necessary When Estimating Demand for a Product That Does Not Yet Exist

Every professional asked to estimate the size of a new market--especially for products that do not yet exist--has a Fermi problem. One has to estimate the size of something but with unknown or incomplete data.

It is not required when making forecasts about existing products with relatively well-established demand curves.

As a client once said, asking for an estmate of demand and a cost to create the network for a non-existing consumer product, “all I care about is that your projection is in the right order of magnitude. I don’t worry about precision less than that.”

source: Everyday Concepts

Disintermediation is the Single Biggest Disruptor of the Connectivity Business, Ever

Disintermediation has been an issue for all retailers and distributors since the internet emerged. Irt also affects the connectivity business. Disintermediation happens when a “middle man” or distributor or retailer is cut out of a supply chain. 

source: WallStreetMojo 

Consider what Facebook has explored about operating as an internet service provider, and what Amazon is doing as a supplier of unified communications. 

Ignore for the moment the fact that app suppliers no longer require a business relationship with an access provider (telco, cable company, satellite or wireless internet access provider or Wi-Fi network) to supply value to end users and customers. 

Look only at the disintermediation of the whole third party internet access provider role. 

Facebook--as do other hyperscale app and computing giants--operates its own wide area networks. But it also has been working on ways to reduce the cost of building access networks, through Telecom Infra Project, for example, which has produced the Terragraph wireless access system, the Bombyx  optical fiber construction robot, 

Facebook also has experimented with internet access from unmanned aerial vehicles and satellite delivery, as well as seeking to overcome the other gaps--cultural, social, linguistic--that have prevented people from using the internet. 

That Facebook has determined that neither UAVs or satellite access makes sense for Facebook as a way of supplying internet access directly to customers or end users is not the point. It was exploring disintermediation. 

Also, consider the new offering of Amazon Connect, which provides contact center capabilities that displace or compete with other contact center services or owned infrastructure. 

Amazon Connect, available from Amazon Web Services,  “offers high-volume outbound communications for calls, texts, and emails,” says AWS. “Amazon Connect now offers organizations a simple, embedded, cost-effective way to contact millions of customers daily for communications like marketing promotions, appointment reminders, and upcoming delivery notifications without having to integrate third-party tools.”

Contact center managers can more easily schedule and launch high-volume outbound communications by simply specifying the communications channel, contact list, and content that will be sent to customers. 

Those are examples of disintermediation at work in the connectivity business. The most challenging example has been the change of telecom architecture from closed models to open internet protocol. 

Connectivity providers no longer have gatekeeper power over applications that flow over their networks. Any lawful application can reach any customer on any public network, anywhere. That separation of app from access is profound.

But connectivity providers also sometimes face other, more direct, forms of disintermediation, as when a third party completely displaces one or more of the values (services and apps) once provided directly by a telco. 

Messaging, voice, unified communications, video entertainment and rival facilities-based networks provide examples. WhatsApp--owned by Facebook--displaces text messaging. Facebook Messenger--owned by Facebook--likewise displaces text messaging. Skype early on displaced international voice calls. 

Cloud-based unified communications providers displace enterprise phone systems. Streaming video services disintermediate linear video subscription services. Google Fiber displaces carrier internet access. Google Fi displaces carrier mobility service. Now Amazon displaces some unified communications services. 

You Cannot Sell What You Cannot Bill For

With the caveat that it has been decades since I was professionally required to think about or write about billing and operations support systems, these days often referred to as converged charging systems, some issues do not seem to have changed. 

The limitations of legacy systems still seem to be an issue. Flexibility, monolithic architectures, high support costs and real-time charging limitations still seem to be issues. The scalability of legacy solutions--seen as limiting the ability to create new products, services or features, remain issues.

The old adage that “one cannot sell a service one cannot bill for” seems to remain a key issue. 

“Legacy BSS systems constrain CSPs’ ability to make timely launches of new products and offerings, or even provide real-time usage and billing information, says Oracle. 

“This put severe limitations on CSPs’ ability to compete with over-the-top (OTT) providers and new entrants,” says Oracle. 

Up to a point, this all makes sense. In principle, the ability to charge for anything, any way, is an advantage. But unlimited capability also tends to come with unlimited cost, and the general rule is that a charging system instance cannot cost more than the item sold. 

source: Vinod Sharma 

A charging instance cannot even cost more than a fraction of the sale price and proceeds. In fact, in many instances, the cost of tracking a discrete event also has to be infinitesimally small, as there might be no direct revenue associated with that instance. 

In a practical sense, converged charging arguably is more relevant as a way of lowering billing instance and overall cost, while supporting real-time charging capabilities that might be needed for on-demand products whose use varies considerably, and whose charging model is based on some combination of usage, services, features, quality metrics, account discounts, payment model and other criteria a marketer might devise. 

These days, converged billing systems also are required by computing-as-a-service suppliers as well. 

The broader trend of billing systems requiring integration with operations support systems was an issue decades ago, and remains a live issue today. Perhaps the big difference is the heightened importance of supporting real-time rating, as more services and features can be created for on-demand use. 

One thing has not changed: service provider charging systems are oriented around customers with whom one has a business relationship. Connectivity provider systems are not designed around users with whom there is no formal business relationship. 

That remains a big difference between “fee for service” business models and “indirect monetization” models used by many application providers. 

App providers who have users and service providers who have customers both need to track usage and identity. But the need to charge is different. Indirect user monetization relies on knowledge of consumption or dwell time, so third-party monetization is possible. 

Some vendors will argue that converged charging systems are required so connectivity providers can “compete” with application providers. That arguably is not the case. A connectivity provider that owns an application with indirect monetization would use the different sorts of tracking systems needed to validate user engagement, based on “users” rather than “customers.” 

Or so it continues to seem.