A recent independent study of LTE coverage within four U.S. cities showed Sprint's networks lagging far behind those of both Verizon Wireless and AT&T.
In comprehensive testing in Dallas, Fort Worth, Kansas City and Atlanta in mid-August and early September, research firm Advanced Frequency Engineering (AFE) found that Sprint's LTE service was "not present or not accessible in 75% to 90% of Sprint's advertised LTE coverage area in the cities tested at time of testing."
A Sprint spokeswoman called the results "surprising" and defended the carrier's initial LTE launches in a handful of cities over the summer. Sprint now offers LTE coverage in 19 cities, with about 100 more cities to be added to the list "in the coming months," she said.
"We're happy with the way launches are going," said the spokeswoman, Kelly Schlageter, today.
AFE found that both the Verizon and AT&T networks have 100% coverage in the areas tested.
The study involved nearly 1,000 miles of testing by drivers and from fixed locations in the networked cities.
Testers used both Samsung Galaxy S III LTE handsets and specialized spectrum analyzer equipment, AFE said.
During tests condusted in mid-August, AFE found that Sprint had just 10% coverage in Dallas, 20% in Fort Worth and 25% in the Kansas City area, which Sprint's headquarters in Overland Park, Kansas.
Early September tests found that Sprint had 15% coverage in Atlanta.
AFE looked at downloads speeds, performance with Web surfing and streaming video and LTE coverage, but only shared LTE availability data in a public summary of its findings.
The researcher also provided a map of Dallas showing that its tests found coverage far below what Sprint was showing on its Web site at the time.
Schlageter said the test results were a surprise to Sprint, as it carefully monitors its coverage as cell towers and cell sites on building rooftops come online with LTE.
"There's not a market where we don't know how many sites we have [on LTE] and how many are launched," she said. "That's why [the study is] surprising. It's a hard number."
Schlageter didn't attack the AFE results, but said it is possible that Sprint could have brought down a cell site serving an entire neighborhood in a large city for 24 hours while another new LTE cell site is integrated into the entire city's cluster of overall coverage. That could have caused at least some of the gaps AFE found, she said.
Sprint's philosophy with building LTE "is to get information on LTE out there as fast as possible," Schlageter said. While the company has announced its LTE launches in press releases, it hasn't yet publicized it via traditional advertising, she added.
When asked how an LTE map on Sprint's Web site could show an entire area is online but only a portion was activated as AFE found, Schlageter urged individual customers and potential customers to input an actual address instead of an entire city to see what their coverage will be.
Currently, for example, Sprint shows all of Dallas within LTE coverage at www.sprint.com/coverage.
Schlageter explained that Sprint's maps show on-street coverage based on the cell sites that are on-air, and that the information that is updated weekly.
Sites that augment coverage after a launch and are incremental will not be reflected on the map, but will improve a customer's performance, she said.
Asked how Sprint determines when to declare a city "launched" for LTE, she said that it uses an approach that is standard in the industry: "You launch when you have a bunch of [cell] sites ready. Our build has been pretty aggressive."
She said Sprint will hold off calling a city launched when a large majority of the cell sites there are activated, but really densely populated downtown area is not online. At the same time, Sprint may call a sparsely congested area activated when only a minority of the cell sites are activated, she explained.
Schlageter said that Sprint's rollout of LTE is progressing quickly and are now a month ahead of when AFE conducted its last tests in early September.
"Sites are coming on-air in launched and unlaunched markets all the time," she said.
"Sprint is touching almost every single cell site in our existing 3G network and swapping out older equipment for brand new 3G and 4G LTE equipment. By the end of 2013, Sprint expects to have the entire network build largely complete," Schlageter added.
After an LTE network is launched in a city, Sprint and the other carriers routinely return to the area to add more LTE cell sites. The follow-on work fills in areas in between existing sites based on demand from customers, she said.
If Sprint customers buy a new LTE phone, including the new iPhone 5, and are dissatisfied with the performance of the device, Sprint has a 14-day return policy, she said, which is detailed online.
While the AFE tests appear to be damaging for Sprint, some analysts were forgiving.
Jack Gold, an analyst at J. Gold Associates, said the test results are "not too concerning given the startup nature of Sprint's LTE network." But he warned that if Sprint's limited coverage continues beyond three to four months, "they have a real problem."
A typical cell site can cost $50,000 to upgrade, and Sprint and other carriers must add LTE to thousands of sites, he noted.
The Sprint findings reflect the nature of network upgrades, he added. "How quickly can you go out and touch each tower and at what cost? And how to do that without disrupting current customers and service? "
Still, Gold said it is "not a good idea" for Sprint to issue press releases and post maps promoting LTE coverage that it doesn't have. "Every carrier does that to some extent," he said. "But the maps should be accurate."
Matt Hamblen covers mobile and wireless, smartphones and other handhelds, and wireless networking for Computerworld. Follow Matt on Twitter at @matthamblen, or subscribe to Matt's RSS feed . His e-mail address is firstname.lastname@example.org.
Read more about mobile/wireless in Computerworld's Mobile/Wireless Topic Center.