Friday, February 16, 2018

What Does "Winning" Look Like for Telco Fixed Internet Access?

What does “winning” look like for telco internet access? Many tier-one U.S. telcos, for example, have about 40 percent market share. Is that winning? It depends on one’s perspective.

Share of 40 percent means one thing if share previously was 35 percent. It means something else if share formerly was 50 percent.

Many independent telcos other than AT&T and Verizon have been losing market share to cable operators for some years, and might well have less than 40 percent share. Cincinnati Bell seems to have been in that category, and seems to find that its fiber to premises program is allowing it to regain market share.

“Our results demonstrate that we continue to compete and win against cable with fiber,” said Leigh Fox, Cincinnati Bell CEO. The caveat is that Charter Communications has not yet launched its DOCSIS 3.1 gigabit service in Cincinnati. That will be the test of whether Cincinnati Bell can continue taking share from Charter.

“Competing” in this case means having 40 percent market share, and gaining about three share points in the year.

“During 2017, we added both video and Internet subscribers despite continued intense marketing and advertising efforts from our primary competitor,” said Fox. “In the fourth quarter, we added 5,400 Fioptics Internet subscribers and ended the year with approximately 227,000 total subscribers with penetration rates reaching 40 percent and ARPU increasing three percent year-over-year.

Will Mobile Users Still Rely on Wi-Fi in the Future?

You might not be surprised to learn that U.S. Android users consume most of their mobile data using Wi-Fi. But you might be surprised that customers on unlimited usage plans also rely mostly on Wi-Fi for data access. But that is what happened in January 2018, according to Strategy Analytics.

Android customers buying unlimited usage plans consumed only about 28 percent of total mobile device data using the mobile network, 72 percent on Wi-Fi.

That seems counterintuitive, if most users also are on 4G networks offering performance often better than Wi-Fi  (especially on public hotspots). In fact, behavior should already be changing.

“Customers are rational,” says Craig Moffett, MoffettNathanson analyst. “When pricing incentives favor Wi-Fi, customers use more Wi-Fi. When pricing incentives shift, so does behavior.”

In fact, some studies suggest that nearly 40 percent of U.S. “at home” access uses the mobile internet, not a fixed connection. In part, that might be because most internet sessions now happen on mobile devices. In part, that might be because many users do not buy fixed internet access, or do not pay for it.


Many of those users say they are on unlimited usage plans and therefore do not need to offload to Wi-Fi to save money.

Of course, it also is possible that self-reported usage actually does not reflect actual usage. Respondents might have been connected to a Wi-Fi connection at home, and not known it. The sample might be include an unrepresentative universe of respondents who do not buy, or have access to, fixed internet access, and must rely on mobile network access.




Still, Wi-Fi usage could fall in the future, as more users opt for unlimited plans, as 5G networks start to offer speeds equivalent or faster than fixed connections and as the cost of mobile access starts to near parity with fixed network prices.

Look to Google, Apple for Emergency Location Innovation

Here are two examples of innovation in the telecom industry that come from “outside” the industry. First, Amazon’s Alexa and the line of Amazon voice appliances has created a new platform for consumer voice. Basically, Alexa is becoming a voice-activated “home phone.”

The other example is emergency calling, where it is Google that is innovating in location services. In recent tests, Google tested emergency call location at 911 call centers with West Corp. and RapidSOS.

RapidSOS said its portion of the trial involved about 50 911 centers covering some 2.4 million people in Texas, Tennessee and Florida.

Location data in more than 80 percent of the 911 calls using Google’s technology were more accurate than the carrier data in the first 30 seconds of a call, according to RapidSOS.

Google’s data provided an average location estimate radius of 121 feet, RapidSOS said, while carrier data averaged 522 feet. Carrier data also took longer to reach 911 centers, RapidSOS said.

Google has said it hopes to deploy the technology broadly across the U.S. some time this year. Apple also is said to be developing such location technology.

There are lots of reasons why innovation, research and development have largely moved outside service provider purview. Profits to support such research no longer exist, for starters. Telco research also was outsourced to industry suppliers and in some cases to third party research outfits.

In that absence, and because of profound changes in ownership of key data stores, key device and app suppliers appear to be moving into the breach.

Is Telecom Like Airlines, or Autos?

Analogies sometimes are helpful when trying to understand the underlying dynamics of the “telecom” industry. In past centuries, telecom might reasonably have been likened to roads, pipelines, electrical or water utilities. They were considered “natural monopolies” not amenable to competition, with state ownership quite common.

In the competitive era, beginning nascently about 1985, new analogies were more apt. Some have likened competitive telecom to the airline industry . Both were highly regulated in the past, were then deregulated, are capital intensive, subject to scale economics, with global and local business models.

Both industries now rely on multiple revenue streams, where in the past revenue was generated only from customers buying tickets or making phone calls. Airlines now generate revenue by selling miles to affinity and reward program providers, while telcos are moving into advertising-- where business partners, not subscribers--are the revenue model.

Consultant Martin Geddes says both industries actually deal in abstractions, or should.  The “value” is destination arrival or application performance--an outcome--rather than the underlying resource (capacity, seats).

In other words, the performance (specificity of arrival time) is the outcome and value. High-value, time-constrained arrival is one class of service, while best-effort arrival is another class of service. It might cost one amount to be guaranteed arrival on a specific day, at a specific time; it might cost quite a lot less to “get there” with some possible delay (the next day, eight hours later, and so forth).

There are some barriers to full competition, including ownership of local access facilities in one case and landing slots in the other. Both industries have been reshaped by low-cost competition. But the nature of the competition has changed. Back around the turn of the century we might have thought the competition would come from new service providers.

As it turns out, the new competition comes from “over the top” app providers. To be sure, facilities-based competitors do exist, mostly in the mobile segment of the business. But the primary assault as been by application providers whose products simply offer product substitutes that work on any internet-connected device.

Think of the analogy to business partner (advertising) revenue: app providers often make their money from advertising, not subscriptions and direct payments by users (customers).

But the next set of analogies might be to privately-owned autos. Fleets of self-driving vehicles might eventually obviate the need for private ownership of automobiles. Transportation still is provided (by fleets of self-driving vehicles available on demand), but not by one means (car ownership).

Some people might object: “there still will remain a need for access.” That is correct. Networks still will be needed. What is not clear is that “ownership of my own network” is required. Netflix is a major, tier-one provider of subscription TV. But it owns not retail access networks. Neither does Skype, Google, Facebook or Amazon or Alibaba.

In that new analogy, access networks are like fleets of self-driving cars. They are available to provide transportation as needed (over the top apps).  But the present “owned automobile industry” (telcos and other service providers) will suffer, as people will buy far fewer automobiles.

That is the inevitable consequence of adopting the internet and IP networks are the next generation network. Use of apps is decoupled from ownership of networks. So long as any user has internet access, internet apps work.

To use the auto analogy: people (app providers) will not have to “own” cars to satisfy many  transportation needs. Vehicles (network access) still will be necessary. “Owning” the networks (access facilities) will not be required.

As it turns out, the airline analogy was the least of the telecom industry’s problems. The self-driving auto--and its ability to support fleets of on-demand accessed transportation--is the real problem.  

Thursday, February 15, 2018

Are Happy Customers Really Loyal? Are Loyal Customers More Profitable?

Loyal customers are “good” because they do not churn, conventional wisdom suggests. There are other ideas we tend to take for granted: Satisfied customers are loyal; customers who give a firm high “net promoter scores” are satisfied and loyal; loyal customers have longer “customer relationship lifetimes;” and are “more profitable.”

Sometimes those assertions are correct; often not. Customer satisfaction might not translate into customer loyalty, defined as unwillingness to desert a current provider for another supplier.

“What we’ve found is that the relationship between loyalty and profitability is much weaker—and subtler—than the proponents of loyalty programs claim,” say Werner Reinartz, Professor of Marketing at the University of Cologne, and V. Kumar,  executive director of the Center for Excellence in Brand and Customer Management at Georgia State University’s J. Mack Robinson College of Business.


“Specifically, we discovered little or no evidence to suggest that customers who purchase steadily from a company over time are necessarily cheaper to serve, less price sensitive, or particularly effective at bringing in new business,” they argue. The researchers find “no evidence” to support such claims.

“In none of the four companies we tracked were long-standing customers consistently cheaper to manage than short-term customers,” say Reinartz and Kumar. “In fact, the only strong correlation between customer longevity and costs that we found—in the high-tech corporate service provider—suggested that loyal and presumably experienced customers were actually more expensive to serve.”

In business-to-business markets, that is perhaps easy to understand.  n business-to-business industries, high-volume customers often exploit their value to get premium service or price discounts from suppliers.

That arguably is not so true, if true at all, in business-to-consumer markets. “At the very least, the link between loyalty and lower costs is industry specific,” they say.

“Long-term customers consistently paid lower prices than the newer customers did—between fiver percent and seven percent lower, depending on the product category,” said Reinartz and Kumar. “What was surprising was that we found no evidence that such loyal customers paid higher prices in the consumer businesses.”

In fact, they argue, loyal customers--business or consumer--are more price sensitive than less-loyal customers.  That might help explain higher churn in telecom markets. If consumers believe loyalty should be rewarded with lower prices, then raising prices after an introductory period runs completely counter to that belief.


Churn, to be sure, is a key issue for most service providers, as getting new customers costs money. By some estimates, postpaid subscriber acquisition costs run about four months recurring revenue, while a new prepaid account might cost about one to two months’ revenue.

What is not so clear is the direct linkage between service provider programs and customer behavior (lower churn, less call center traffic).  Should service providers aim to “delight” customers, or only prevent and fix problems quickly? Should providers invest in relationships or not? If so, in which segments? The research is not so clear on such matters.

Wednesday, February 14, 2018

AT&T to Attack Verizon MDUs in Boston

Anybody who thinks internet access competition is reaching a nadir might need to rethink those assumptions in the wake of new out-of-region assaults by Verizon, now countered to some extent by AT&T.

The important development is that markets traditionally lead by a telco and a cable operator, sometimes supplemented by competition from Google Fiber or other independent internet service providers, now will become markets where two tier-one telcos, a tier-one cable operator and often other ISPs also compete.

In other words, competition still is increasing, not shrinking, as some believe.
AT&T is offering internet access to multiple-dwelling units in Boston. The move is one example of a new trend in the fixed network business: large tier-one competitors moving out of region for the first time at scale.

Verizon, for its part, already has launched an assault on the AT&T market in Sacramento, and seems likely to attack another dozen or so markets as well. Most of the new 11 launch markets are out of region for Verizon, including Ann Arbor, Atlanta, Bernardsville (NJ), Brockton (MA ), Dallas, Denver, Houston, Miami, Sacramento, Seattle and Washington, D.C.

Of those markets, Washington, D.C.; Brockton and Bernardsville are inside Verizon’s fixed network footprint. The other markets are AT&T or CenturyLink domains.

Those moves outside the fixed network footprint by Verizon indicate thinking about growth prospects outside the core fixed network service territories, and probably also show that fixed wireless in the millimeter wave bands offers a new business case that did not exist before.

It remains unclear how much--if at all--the Verizon deep fiber architecture will play a key role in those out-of-region assaults. At least initially, the targets of opportunity likely will be locations reached by the metro fiber assets, using fixed wireless for access. That tends to suggest urban core targets of opportunity.

Essentially, AT&T and Verizon are becoming competitive local exchange carriers, at some scale, for the first time, attacking other tier-one telcos in their home markets.

The business case likely also includes the ability to use those assets to help with backhaul operations to support 5G small cell deployments as well.

T-Mobile US Deploys AI-Based Customer Care

T-Mobile US has deployed an automated customer care resolution tool supplied by Tupl.

The ACCR tool provides T-Mobile US customer care reps with detailed and easy-to-understand cause reports and technical resolutions, Tupl says. The ACCR tool is said to be 100 times faster and up to four times more accurate than legacy resolution practices support, automating as much as 90 percent of such interactions.

A TM Forum survey suggests service providers already are using artificial intelligence to support chatbots for customer service, network automation and service management, while 70 percent of service provider executives are conducting “proof of concept” trials.

At least so far, improving customer experience is top reason for AI interest, followed by reducing operating expenses.

The survey of 187 executives from 76 communications service providers operating in 51 countries, and 115 executives from supplier companies also suggests some problems.

Since AI is designed to augment human intelligence and processes, reaping value presupposes that organizations actually understand their core processes well enough to augment them, and can do so, as a practical matter. The history of enterprises applying information technology and reaping rewards is spotty, though.



Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...