Sprint, the nation's third largest wireless operator, is largely finished with the mammoth Network Vision network modernization project it started more than three years ago. The effort involved improving Sprint's CDMA network, shuttering its iDEN network and--during the past two years--launching LTE on its 1900 MHz spectrum. The result, however, is an LTE service that only covers around 200 million people and is, by most measurements, the nation's slowest. Compare this to T-Mobile (NYSE:TMUS), which covered roughly the same number of people with LTE in half the time as Sprint with speeds that often rank at or near the top. And T-Mobile is enjoying significant momentum thanks to its "uncarrier" branding.
Nonetheless, Sprint executives are arguing that 2014 is "the year" for Sprint.
"We have a very bright future ahead of us," John Saw, Sprint's chief network officer, told me in an interview.
I think that remains to be seen. As 556 Ventures analyst William Ho said: "It comes down to execution."
Saw explained that Network Vision was a painful but necessary action that positions Sprint for the future: "I think we are the only carrier that has decided to take the band aid off and basically build brand new sites that are modular, that sets a better foundation for the future," Saw said. "A lot of the other carriers simply just added to what they have. Ultimately there will come a time when the customer demands for bandwidth, for capacity--the technology changes that are coming online for 4G and even beyond--will require carriers to start upgrading their platforms. Sprint basically took the brunt of it upfront knowing that we need a strong foundation for all the things that we need to do, and we did it with [Network] Vision."
Saw said Network Vision lays the groundwork for Sprint Spark, a technology announced last year that combines TDD-LTE network technology with FDD LTE network technology across Sprint's 2.5 GHz, 1900 MHz and 800 MHz spectrum bands. "Sprint Spark is really the future of our network here at Sprint," Saw said.
(Ironically, Spark runs on the same spectrum--2.5 GHz--that formed the basis for Xohm, the "4G" WiMAX service Sprint launched almost seven years ago. Xohm reportedly provided speeds of around 3 Mbps, roughly twice as fast as Verizon's CDMA EV-DO network at the time. Today, Sprint Spark supports peak speeds of roughly 50 Mbps--around twice as fast as the average speeds Verizon is seeing with its 20x20 MHz LTE network using 700 MHz and AWS spectrum.)
"Once they (Sprint) get this speed thing going… they're going to be in a strong position to take back market share," Ho said.
But since announcing Spark, Sprint has strangely continued to drum up the technology's potential. In December, Sprint said Spark could eventually provide real-world speeds of 150-180 Mbps. Then, earlier this month, SoftBank CEO Masayoshi Son said he hopes to deliver up to 200 Mbps, presumably using Spark. I don't think this is a good tack for Sprint to take. The carrier has been promoting Network Vision for close to three years now--and instead of focusing on where the carrier is today, as it completes Network Vision, Sprint is again promising incredible speeds and performance at some point in the future.
Sprint's network buildout "should go faster," acknowledged Ho.
It seems that Sprint's management agrees. The architect of Sprint's initial Network Vision plan, Steve Elfman, left the company abruptly earlier this month. Bob Azzi, Sprint's senior vice president of networks, also left at the same time. Sprint has largely declined to comment on the departures.
That leave Saw, Clearwire's former CTO who joined Sprint last year, and Stephen Bye, a former Cox Communications network executive who joined Sprint in 2011, as the people in charge of taking Spark nationwide. Bye, Sprint's CTO, is responsible for technology selection and spectrum planning. And Saw is heading the actual construction and implementation of Sprint's network.
They definitely have their work cut out for them. First, SoftBank's Masayoshi Son is reportedly a driven, aggressive and unrelenting manager who is pushing Sprint to become more competitive in the U.S. market.
Second, 2.5 GHz is notoriously difficult to build out due to the spectrum's propagation characteristics--lower-band spectrum like 700 MHz covers far more geography than higher-band spectrum like 2.5 GHz. For example, according to a study commissioned by T-Mobile, covering Arizona with a low-band 700 MHz LTE network would cost around $19 million, but it would cost roughly $58 million to cover the same area using mid-band 1900 MHz spectrum. Presumably, high-band spectrum like 2.5 GHz would cost even more.
On the other hand, Sprint's trove of 2.5 GHz spectrum will allow the carrier to provide faster speeds and more capacity.
In my interview with Saw, he said that Spark on Sprint's 2.5 GHz spectrum is today providing peak speeds of 50 Mbps and average speeds of 12-15 Mbps. He said Sprint plans to increase Spark's peak speeds to around 120 Mbps in a gradual rollout starting near the end of 2014, and then will increase those peak speeds again to 180 Mbps in a gradual rollout starting at the end of 2015.
Indeed, Strategy Analytics believes Sprint will become the "king of data speed." And TBR analyst Eric Costa wrote that "Spark can help the operator make a splash in the wireless market over the next two years by offering the industry's top network speeds."
Saw also said that SoftBank's Son is taking an active role in the buildout of Sprint's network. "The pace of the deployment has also picked up in terms of what we need to get done," Saw said of SoftBank. "We are pushing as hard as we can to hit those milestones."
"Masa is extremely passionate about the network," Saw added, explaining that SoftBank was one of the first network operators in the world to roll out a network using 2.5 GHz spectrum, and that Sprint is working closely with SoftBank engineers on Spark. "There is a tight collaboration. I speak to my SoftBank colleagues every day."
Here's where Sprint stands today: The carrier's 4G LTE network reaches over 200 million people. Sprint's 3G network reaches over 277 million people. With roaming, Sprint covers a total of 316 million people. As for Spark, Sprint offers Spark in parts of 18 markets today and plans to bring the technology to 100 million people by the end of this year and the top 100 U.S. markets within around three years.
Meantime, Verizon's LTE network currently covers 305 million POPs. AT&T currently covers 280 million POPs with LTE and aims to hit 300 million by mid-year. And T-Mobile's LTE network currently covers 210 million POPs, and T-Mobile is planning to expand that to 250 million by the end of this year. All of these carriers are working to add additional capacity to their LTE network through spectrum purchases and 3G refarming.
Obviously Sprint knows it needs to move quickly. And SoftBank's Son has hinted that Sprint could disrupt the industry not only with blazing fast speeds but also with rock-bottom prices. But time is ticking away and the competition isn't standing still. In 2016, when Sprint's Spark covers the top 100 markets, will that be enough? I think that remains to be seen --Mike | +MikeDano | @mikeddano