With many cable operators reporting a need to double network capacity every 18-24 months to keep up with customer traffic demands, the industry is spending time and money contemplating how to meet future needs while also finding ways to cut costs and make networks more efficient.

Top technology executives from five major cable operators answered questions (sub. req’d.) from Multichannel News about their current broadband networks and their plans for the future. Some, like Mediacom, are aggressively adopting DOCSIS 3.1 cable broadband upgrades for their customers while companies like Cox and Comcast are deploying multiple solutions that use both traditional hybrid fiber-coax network technology and, on occasion, fiber-to-the-home to boost speed and performance. But at least one cable company — Charter Communications — thinks it can continue operating its existing DOCSIS 3 network without major upgrades for several years to come.

Cable Broadband Traffic Can Be Handled

“We’ve been on a pretty steady path of doubling our network capacity every 18-24 months for several years, and I don’t see anything that makes me think that will change,” said Tony Werner, president of technology and product at Comcast. “We’ve been strategically extending fiber further into our network to meet customer demand, and that effort, combined with our commitment to deploying DOCSIS 3.1 has given us a network that’s powerful, flexible, and ready for what’s next.”

J.R. Walden, senior vice president of technology at Mediacom was more aggressive.

“We have completed the removal of all the analog channels. That was the big step one,” Walden said. “Step two was to start transitioning high-speed data over to DOCSIS 3.1, so we’re not adding any more 3.0 channels, and reuse spectrum for 3.1, which is a bit more efficient. The whole company is 3.1, all the modems we’re buying since June have been 3.1, so we’ve begun that next transition.”

Walden added Mediacom is also trying to improve broadband performance by reducing the number of customers sharing the same connection.

“We average about 285 homes to 290 homes per node as an average,” he said.

Mediacom is also scrapping older technology on the TV side to open new bandwidth. The cable company is getting rid of MPEG-2-only set-top boxes so the company can transition its video lineup to MPEG-4. But even that won’t last long. Walden admits the company will then quickly start moving less-viewed channels and some premium networks to IP delivery.

Traditional cable broadband service relies on a hybrid fiber-coax network.

In its European markets, Liberty Global has adopted Converged Cable Access Platform (CCAP) equipment across its footprint. CCAP technology saves cable operators space and operates more efficiently, and supports future convergence of technologies that cable operators want to adopt in the future. CCAP has helped Liberty Global deal with its 45% traffic growth by making upgrades easier. The company is also using advanced features of CCAP to better balance how many customers are sharing a connection. The next step is adopting DOCSIS 3.1.

“Seventy to 80% of our plant will be DOCSIS 3.1 ready by the end of next year, giving us a path to even greater capacity expansion allowing us to continue to increase the available capacity across our access network, upstream and downstream,” said Dan Hennessy, chief architect of network architecture for Liberty.

Charter is prioritizing maximizing performance on the network it already has.

“Our priority is to constantly balance capacity against demand. It’s a never-ending quest,” said Jay Rolls, Charter’s chief technology officer. “We watch it very closely, and we’re very pragmatic about it — the volume of tools, metrics and ways to see what’s really happening, and invest accordingly, is really deepening in ways that matter.”

Is Fiber-to-the-Home in Your Future?

While some cable operators like Altice’s Cablevision are scrapping their existing hybrid fiber-coax networks in favor of fiber-to-the-home (FTTH), America’s largest cable operators are not in any hurry to follow Altice.

Comcast has expanded its fiber network closer to customers in the last few years, but sees no need to convert customers to FTTH service.

“I feel pretty strongly that the best path ahead is to leverage the existing coaxial network and DOCSIS resources to the fullest, then inch towards FTTH, over time Why? Because we can. We don’t have to build an entire network just to turn up one customer.”

The next generation of cable broadband service may depend on CCAP – technology that will cut operator costs and lay the foundation for changing the way video and other services are delivered to customers.

Cox has a 10-year Network 2.0 plan that will bring fiber closer to customers, but not directly to every home. More important to Cox is having the option to support symmetrical speeds, which means delivering upload speeds as fast as download speeds. In the meantime, network cabling Greensboro can improve your current connectivity and reliability, preparing your network for high-speed internet.

“We’re also thinking about the fiber investment and fiber deep as it relates to our wireless strategy, enabling some of our customers with a small cell strategy but also positioning ourselves to take advantage of that in the future, as well as thinking about fiber deep to benefit both residential and our commercial customers simultaneously,” said Kevin Hart, Cox’s executive vice president and chief product and technology officer.

Liberty/Virgin Media’s Project Lightning is bringing cable broadband and TV service to places in the UK that never had cable service before.

In Europe, Liberty Global’s “Project Lightning” network expansion initiative is building out traditional cable service in the United Kingdom. Most of the UK never adopted cable service, favoring small satellite dish service instead. Now Liberty Global is putting cable expansion on its priority list. But decades after most North Americans got cable service for the first time, today’s new buildouts are based largely on fiber optics — either fiber to the home or fiber to the neighborhood, where coaxial cable completes the journey to a customer’s home.

Charter admits the technology it will use in the future partly depends on what the competition is offering. Rolls says the company can eventually roll out DOCSIS 3.1, take fiber deeper, or offer symmetrical download/upload speeds presumably targeted towards its commercial customers. But he also suggested Charter’s existing network can continue to deliver acceptable levels of service without spending a lot on major upgrades.

“It’s a rational approach, where we’re trying to balance the needs, the available technologies, and the costs,” Rolls said. But he also suggested DOCSIS 3.1 isn’t always the answer to upgrades. “DOCSIS 3.1 has some pretty remarkable capabilities, but it’s not necessarily a hard-and-fast reason to not take fiber deeper, for instance [allowing for additional DOCSIS 3 node splits]. Different situations drive different capacity decisions.”

Walden agreed, and Mediacom customers should not expect more than DOCSIS 3.1 upgrades for the near future.

“[Fiber deep] is a bit further out, at least as a large-scale type of project,” Walden told Multichannel News. “I think fiber deep for multi-dwelling units, high-density areas and some planned higher end communities doing deeper fiber or fiber-to-the-home [is happening]. But as a wholesale [change] and going to node+0 kind of architecture, I don’t see that in the next two years.”

Are Symmetrical Speeds Important for Customers?

Verizon’s fiber to the home service FiOS uses symmetrical broadband speeds to its advantage in the marketplace.

Many fiber to the home networks offer customers identical upload and download speeds, but cable broadband was designed to favor downstream speeds over upstream. That decision was based on the premise the majority of users will receive much more traffic than they send. But as the internet evolves, some are wondering if cable broadband’s asymmetric design is now outdated and some competitors like Verizon’s FiOS fiber to the home service now use its symmetrical speed advantage as a selling point.

Cox Communications does not think most customers care, even though its network upgrades are laying the foundation to deliver symmetrical speeds.

“It’s a little but further out on the horizon,” said Hart. “The upstream growth rate is ticking up a couple of notches, but not to the tune that we would need significant additional capacity and/or a complementary need for symmetrical bandwidth. [A]t this stage, the symmetrical is a nice-to-have for residential and definitely will be a good option for our commercial customers.”

Rolls isn’t sure if symmetrical speeds are important to customers either and Charter has no specific plans to move towards upload speed upgrades.

“The world of applications and services continues to evolve, obviously, but so far we’ve been able to meet those needs with an asymmetrical topology,” Rolls said. “That said, things like real-time gaming, augmented and virtual reality, and the Internet of Things — some of those will likely drive more symmetry in the network. It remains to be seen.”

Most Netflix customers in the U.S. will be paying $1-2 more a month to the online streaming service starting in November.

Mashable reports Netflix is raising prices on its Standard plan (currently $9.99/mo) by $1 and those on its Premium plan (now $11.99) will pay $2 more a month. The basic $7.99 plan remains unchanged for now.

“From time to time, Netflix plans and pricing are adjusted as we add more exclusive TV shows and movies, introduce new product features and improve the overall Netflix experience to help members find something great to watch even faster,” Netflix said in a statement.

Most of the extra money will likely be spent on content creation and acquisition for subscribers. Netflix is expected to spend $7 billion on content in 2018.

Netflix plans are differentiated based on video quality and the number of concurrent streams. Here are the respective features of each plan. Customers and downgrade or upgrade at any time.

Prices reflected are prior to the impending rate increase.

The last Netflix rate increase was announced in 2014, but did not take full effect for all customers until 2016.

Alphabet, the parent company of Google Fiber, has lost interest in expanding its fiber to the home service and is showing signs of pulling the plug on its cable television alternative while it drags its feet on keeping promised rollout commitments.

The first sign of trouble for the upstart fiber network came as early as 2015, when without warning Google co-founder Larry Page suddenly unveiled Alphabet, a new holding company that would be at the heart of Google and its many ventures, including Google Fiber. The concept was tailor-made to please Wall Street and investors, because it would better expose which Google projects were earning money and which were hemorrhaging cash with no sign of profitability. But an equally important event occurred in May with the hiring of Ruth Porat, who would become Alphabet’s chief financial officer.

Known inside by some at Google as “Ruthless Ruth,” Porat is Wall Street’s definition of a proper executive that keeps shareholder interests first in mind. Porat lead Morgan Stanley’s technology banking division at the heart of the first dot.com boom in the late 1990s, served as an adviser to the Treasury Department on the taxpayer bailouts of Fannie Mae and Freddie Mac, and was chief financial officer at Morgan Stanley by 2010. Her mission at Google: put an end to expensive innovation for innovation’s sake. If a project did not show signs of making money for shareholders, it would face intense scrutiny under her watch.

“She’s a hatchet man,” a former senior Alphabet executive frankly told Bloomberg News.

Porat

Her key priorities are “discipline” and “focus,” something Google never had to be concerned with while earning truckloads of ad-click cash. Google’s reputation for cool innovation and free services earned the company a lot of goodwill with the public, but that left money on the table for investors who want the company to step up shareholder value. Google’s founders Sergey Brin and Larry Page had enjoyed a long run innovating and announcing new projects, including scanning every printed book on the planet, giving away e-mail and office apps, and laying fiber optic cables to deliver the kind of internet service big phone and cable companies were not delivering. The company also acquired other innovators, including Nest Labs — which made connected thermostats and Webpass, which provides wireless high-speed internet access.

But for all of its success, Google also had several high-profile failures that cost billions, setting the stage for future project accountability.

One of the biggest failures was its Google Glass wearable tech project. The first edition, dubbed Explorer, was a flop and received terrible reviews. But the device also clashed with a country increasingly preoccupied with personal privacy. Not everyone appreciated Google Glass’ always-watching camera pointing in their direction, and some wearing the device were derided as “glassholes.”

“I was a Google Glass Explorer, and the experience was horrible from the start. Google Glass now sits in my office museum of failed products,” said Tim Bajarin, President of Creative Strategies Inc. in this post at re/code. “The UI was terrible, the connection unreliable and the info it delivered had little use to me. It was the worst $1,500 I have ever spent in my life. On the other hand, as a researcher, it was a great tool to help me understand what not to do when creating a product for the consumer.”

Google Glass: a major misstep

Google’s other experiments weren’t exactly pulling in a lot of money either. The company’s vision of driver-less cars met the reality of real world driving conditions (some accidents were the result) and traffic planning and safety regulators were cautious about giving a green light to the concept on American streets and highways. A long-time favorite project of Brin and Page, Project Loon — sending 100,000 balloons, blimps, and/or drones into the sky to deliver internet access is still seen by conventional wisdom as weird. These and other experimental projects lost $3.6 billion of Google’s revenue in 2016, almost twice as much as they lost the company in 2014.

After “Ruthless Ruth” entered the picture, as Bloomberg News documented, it appeared the open door to the experiment lab was closed and an exodus of project leaders and engineers began:

Six months after Teller’s rousing speech, Loon’s Mike Cassidy stepped down as project leader. Around the same time, Urmson, the self-driving car engineer, left Alphabet, as did David Vos, the head of X’s drone effort, Project Wing. Vos’s top deputy, Sean Mullaney, left the company as well. Other recent departures: Craig Barratt, chief executive officer of Access, its telecom division; Bill Maris, the CEO of its venture capital arm, GV; and Tony Fadell, the CEO of smart-thermostat company Nest, who was also working on a reboot of Google Glass. That project, now called Aura, also lost its leads of user design and engineering.

Barratt: The former head of Google Access.

The bean counters also arrived at Google Access — the division responsible for Google Fiber — and by October 2016, Google simultaneously announced it was putting a hold on further expansion of Google Fiber and its CEO, Craig Barratt, was leaving the company. About 10% of employees in the division involuntarily left with him. Insufficiently satisfied with those cutbacks, additional measures were announced in April 2017 including the departure of Milo Medin, a vice president at Google Access and Dennis Kish, a wireless infrastructure veteran who was president of Google Fiber. Nearly 600 Google Access employees were also reassigned to other divisions. Medin was a Google Fiber evangelist in Washington, and often spoke about the impact Google’s fiber project would have on broadband competition and the digital economy.

Porat’s philosophy had a sweeping impact on Alphabet and its various divisions. The most visionary/experimental projects that were originally green-lit with no expectation of making money for a decade or more now required a plan to prove profitability in five years or less. Wall Street was delighted and Alphabet’s stock was up 35% since “Ruthless Ruth” arrived, winning praise for remaking Alphabet/Google into a conventional American corporation using familiar corporate principles.

But Alphabet’s transition seems to break a promise Google co-founders Brin and Page made when Google became a public company in 2004.

“We do not intend to become one.”

Both men promised Google would never focus on short-term profitability and would encourage employees to devote 20% of their working hours on exactly the kinds of innovative projects and product developments Porat was intent on cutting or killing. Porat even has a willing army of helpers — executives were paid bonuses to kill their projects before expenses got out of hand. This helped halt development of Tableau, a project to create enormous size TV screens originally championed by Brin.

Porat also had a major hand in slashing the budget at Google’s Nest Labs division. Google spent $3.2 billion acquiring the home thermostat and smoke alarm company in 2014. Nest CEO Tony Fadell came along as part of the deal and was initially considered a major asset, having been the former Apple engineer who built the original iPod prototype. But Fadell clashed with Google’s culture and reports surfaced he was a tyrannical boss comparable to Steve Jobs at his nastiest. Google executives expected more products out of the Nest division, and didn’t get them. Fadell blamed employees and ruthless budget cuts that broke Google’s commitment to allow Nest to lose up to $500 million annually for the first five years under Google’s ownership. Even when Nest managed to generated $340 million in revenue in 2015, Porat wasn’t pleased. The higher-ups expected more considering the amount of money Google spent buying Nest Labs.

Google Fiber was launched knowing it would take billions of dollars and years to pay off for Google. Laying fiber optic cable is expensive, time-consuming, and frequently bureaucratic. Google projects that still have support from Brin and Page are usually protected from Porat’s red pencil, but if either’s optimism waivers, Porat is likely to start cutting.

By the time Barratt tried to jump-start excitement for the slowly progressing fiber service by announcing a series of new launch cities, Page appeared to have lost interest. Former employees say Page became frustrated with Google Fiber’s lack of progress.

“Larry just thought it wasn’t game-changing enough,” says a former Page adviser. “There’s no flying-saucer shit in laying fiber.”

Charter Communications took out newspaper ads trumpeting Google’s abandonment of some of its potential fiber customers in Kansas City.

Left unprotected, Porat’s budget cutters invaded and further fiber expansion has been suspended, except in areas where Google was already committed to provide the service. But the cutbacks have been so significant, cities are now complaining Google is dragging its feet on its commitments.

In Kansas City — the first to get Google Fiber, the network remains incomplete. In March 2017, Google signaled it was likely to remain incomplete indefinitely after returning hundreds of $10 deposits — many paid years earlier — to residents who were informed Google Fiber would no longer expand into their neighborhoods. In the last two years, Google has become very conservative about the neighborhoods where it will expand service. In most cases, the company now targets multi-dwelling units like condos and apartments, which are cheaper to serve than single family homes.

In late September, Atlanta noticed Google Fiber was stalled in the city and nearby Sandy Springs and Brookhaven. A clear sign Google had effectively suspended construction was a sudden end to construction permit applications around six months ago. Google Fiber denies it is pulling out, but city officials notice work progress has slowed to a crawl.

“Google Fiber is currently available in over 100 residential buildings in the metro Atlanta area and in several neighborhoods in the center of the city. We’re working hard to connect as many people as possible, and encourage people to sign up for updates on our website,” a Google Fiber spokesperson said.

There have been similar problems with Google Fiber expansion in several Texas cities. Some neighborhood residents complained about shoddy installation work because of poor quality third-party contractors, and expansion has slowed down markedly in many areas.

Ironically, AT&T may have been responsible for helping kill Page’s enthusiasm for Google Fiber, serving as a regular obstacle to Google Fiber’s expansion in states like Tennessee where it has been delayed by bureaucratic pole attachment disputes, some resulting in legal action. For Ma Bell and its progeny, a five-year delay is nothing for a company that has been around since the early 1900s and took decades to build out its original telephone network.

Google Fiber Huts – Nashville, Tenn.

Utilities employ a small army of workers that do nothing but deal in a world of tariffs, permit applications, and various filings to regulatory bodies that still govern parts of their operations. For a dot.com company in a hurry, filing permit applications, negotiating pole attachment agreements, hiring subcontractors that can meet regulated specifications, and dealing with incomplete or inaccurate infrastructure maps could be hell on earth. But for phone and cable companies, it is just another day on the job, and their “concern trolling” over the danger of allowing a neophyte like Google to mess with existing electric, phone, and cable wiring to make room for fiber did give some local officials pause.

Page’s hurry to accomplish his fiber dreams were effectively dashed by AT&T’s very close relationship with local officials and its ability to generate a mountain of regulatory and legal paperwork. As a result, Google admitted with great frustration that in Nashville, after months of work, it had only upgraded 33 telephone poles out of 88,000 in the city. The delay also took its toll on a Nashville-based subcontractor helping to build out Google Fiber in the city. Phoenix of Tennessee declared Chapter 11 bankruptcy in September with liabilities between $1-10 million. It also laid off 70 employees. The reason? Google Fiber is stalled in the city.

One Alphabet employee mischaracterized the end effect of the dispute in comments to the Wall Street Journal last year, “Everyone who has done fiber to the home has given up because it costs way too much money and takes way too much time.”

Christopher Mitchell, director of the Community Broadband Project summarized the situation more succinctly for Gizmodo: “the new guy gets screwed.”

Yet it would be more accurate to say companies with short attention spans and an evolving commitment away from innovation and towards Wall Street and its fixation on short-term results will have more difficulty than other companies and communities that have successfully built fiber networks with a patient focus on the future.

Porat has been defending Alphabet’s increasingly conservative spending plans and pull-backs.

“As we reach for moonshots,” she told investors on a financial results conference call, “it’s inevitable that there will be course corrections along the way.” She called some of the shifting priorities and cutbacks “taking a pause” in some areas of business to “lay the foundation for a stronger future.”

For Google Fiber, that is coming in a number of different directions.

The company this week announced it is pulling back on offering cable television service in its new markets, including Louisville, Ky., and San Antonio, Tex., and is raising rates $20-30 a month for bundled customers in areas where television service is still being sold.

“The cost of providing TV programming continues to rise,” the company said in an email notifying customers of the rate increase. The price change will hit existing customers paying $130 a month for Fiber 1000Mbps service + TV. Current customers will pay $150 a month going forward. New customers will pay $10 more for the bundle – $160 a month.

“We’re not afraid to try new things as part of our normal way of doing business, focused on the end goal of getting superfast internet into people’s homes,” wrote head of sales and marketing for Google Access Cathy Fogler in a blog post.

Google Fiber has been a minor player in the cable television business, according to analysts, attracting around 54,000 customers nationwide as of December — only 24,000 more than it had in 2014.

As for the future, with Porat in charge of finances, it is likely Google will downscale expectations and rely on its acquisition of Webpass for future expansion, providing high-speed wireless internet to multi-dwelling apartments, condos, and businesses in dense urban areas. That eliminates costly fiber expansion to individual homes or businesses and is much less expensive to install and maintain.

Any plans for a major Google Fiber push in the future seems unlikely, considering Wall Street’s demands for Return On Investment are not easily tempered. That leaves independent local overbuilders with established ties to their communities the most likely to pick up where Google Fiber has left off. But even those are in short supply. Like any major project of this scope, the best option for getting fiber optics in your community, assuming the local cable or phone company isn’t doing it already, is to treat it as a public infrastructure project like water, sewer, roads or sidewalks.

Most cities were all too happy to compete for Google’s attention (and infrastructure investment). But now that is no longer likely, and many communities will have to decide for themselves which side of the digital divide they want to live in — the side without 21st century broadband or the side that has elected to control their own broadband future and not wait for someone else to get the job done.

U.S. cell phone providers are facing increasing criticism they are dragging their feet on restoring cell service in Puerto Rico while Mexican-owned Claro has now successfully restored service in 28 of the territory’s 78 municipalities.

Claro Puerto Rico, owned by Mexican billionaire Carlos Slim’s America Movil, has dramatically outpaced AT&T, T-Mobile, and Sprint in getting their damaged cell phone facilities back up and running. Claro is Puerto Rico’s second most popular cell company behind AT&T.

“Claro is the only one with service here,” Francisco Portales, 47, a customer of privately held Puerto Rico-based network provider Open Mobile told a Reuters reporter while waiting outside the Claro store in Fajardo hoping to buy a phone.

Looking for a signal.

The FCC’s latest update on Tuesday reported about 88% of Puerto Rico is still without cell service, but the agency does not break down network repairs by carrier, and American providers have declared their specific restoration plans to be confidential.

While AT&T complained the lack of commercial power remained its biggest problem, Claro said it had pre-positioned generators, diesel fuel, battery backups, and vehicles 72 hours before the hurricane hit, which appears to have made all the difference in restoring service.

Sprint said late last week its towers were still standing and “largely intact” although it gave no specific information on when service might be restored. T-Mobile was more frank, reporting “it’s going to be a long road to recovery.”

Claro is not taking advantage of its position as the island’s most reliable post-hurricane carrier, allowing customers of other providers to roam on its network where a signal is available. That may be all the good publicity Claro needs to win over new customers after the hurricane damage is repaired.

Claro’s repair trucks.

Mercedes Saldana, a 54-year-old school cafeteria worker and Sprint customer is just one of many now searching shops for a Claro prepaid phone.

“I don’t have any service, none,” she said. “We don’t know when Sprint’s going to be connected again.”

Customers unwilling to switch carriers and won’t roam may have long travel times ahead of them to find a signal. Luis Pacheco, 64, was planning to drive with his wife to Canovanas — 30 to 40 minutes west — in hopes of finding a cell signal to text his daughter in California. That is the nearest community where AT&T has a signal at the moment.

Before the storm, AT&T dominated Puerto Rico with a 34% market share, followed by Claro Puerto Rico with a 26% share. T-Mobile was third with 19%, Open Mobile has 11% and Sprint 10%. Verizon Wireless has no network facilities in Puerto Rico, but travelers with Verizon phones are granted roaming access on Claro’s network.

Less than a decade ago, AT&T was one of El Paso’s largest private employers, with 2,400 employees. Next month, it will be a shadow of its former self with fewer than 500 local workers after a series of layoffs and call center closures.

AT&T is planning to close its East El Paso office in November, giving 278 employees the option of leaving or relocating to San Antonio, Missouri, or Florida to remain employed by AT&T.

AT&T used to employ thousands of workers in its El Paso call centers and technical facilities. But much of that work is now being shifted to third-party contractors and offshore call centers overseas.

Since 2011, AT&T has eliminated 12,000 call center jobs in the United States, closing and downsizing call centers across the country, according to the Communications Workers of America.

In 2006, AT&T closed a major call center in Massachusetts, despite receiving generous tax benefits from the local and state government, and offered to relocate those employees to the same call centers in El Paso it is closing now.

In 2015, AT&T demanded El Paso and the state of Texas triple their $50 million annual tax break or else they would shift spending elsewhere. It appears tax abatements ultimately had little effect on AT&T’s spending decisions in the western Texas city.

The union reports the annual salaries for those jobs ranged from $32,000 to $65,000 per year, plus commissions and health and retirement benefits. Offshore customer care centers pay a fraction of those salaries and many third-party contractors do not pay benefits because they designate many employees as part-time workers.

AT&T disputes it is increasing its offshore customer service workforce at the cost of American workers.

“It’s important to note that there is a job for every employee who is willing to relocate to the facilities where the work is being consolidated,” and they will get a relocation allowance if they have to move, Marty Richter, a spokesman for AT&T, told the El Paso Times.

“We’re adding people in many areas of our business where we’re seeing increased customer demand for products and services,” and reducing jobs in areas where work volumes are decreasing, “in part because of changing technology,” Richter added.

Most of the remaining 350 AT&T employees in El Paso will be staffing five retail stores in the area or working as technicians or back-office workers.

Few are expected to take AT&T’s offer to relocate to San Antonio, if only because there are signs AT&T will continue to cut back on its domestic call center operations and shift that work online or overseas.