Carrier coverage claims: What does covering “X-percentage” of Americans really mean?

When it comes to wireless service, there is a difference between coverage, and well, coverage. If there is any perceived weak point in the US wireless infrastructure, it could be argued that it is actual physical coverage, the ability to pick-up a carrier signal anywhere, anytime.

The United States is a large country, and believe it or not, the vast majority of the land mass is not developed. That makes providing any type of service that requires some type of infrastructure, wired or wireless, is a daunting task. That is why providers use carefully parsed marketing language like “service to over 95% of Americans,” versus claiming, “We cover 95% of the country.”

It is fairly self-explanatory that the physical gaps we see in a coverage map are areas of low, or zero, population density, even in the immediate areas outside some western cities. When it comes to AT&T and Verizon, a less relevant explanation for some gaps would be areas where those carriers do not have their own network. In those few instances, smaller competitive carriers are in operation.

The physical footprint is the next coverage battleground

Given the incumbent nature of AT&T’s and Verizon’s networks, going back to the original A-Side and B-Side 800MHz systems that were first built decades ago, it is easy to accept claims of having 97% or 99% of Americans covered by their networks. After all, some of that infrastructure pre-dates the first Motorola prototypes of the 1970s.

What about T-Mobile and Sprint? Everyone knows that these two carriers have younger and physically smaller networks. Yet T-Mobile, the United States’ third or fourth largest of big four carriers (depending on how you count), claims an overall network reach that covers 96% of the population. What gives? Well, T-Mobile is talking about its entire network, so that includes its LTE service all the way down to more rural areas that might still register as GPRS on a mobile device. T-Mobile is not lying about its statistic, but the reality is that percentages are rapidly becoming more meaningless in terms of wireless coverage in the United States. T-Mobile is also aggressively converting its legacy systems to LTE, and it has done a remarkable job in just two years. The conversion of its 700MHz licenses will give T-Mobile a substantially more robust physical footprint.

Sprint recently announced a physical build-out that may involve as many as 20,000 new cell sites. That too will dramatically increase its physical footprint, other network issues notwithstanding.

Project: Make your own carrier

Let’s pretend for a moment that you have some spectrum, and want to launch your own wireless carrier. As a new business however, you need to get the biggest bang for the buck. Your goal: create a network that within its first 6 months covers 50% of all Americans. What would that network look like? The answer may surprise you. As you can see from the adjacent map, you need only to devote the majority of your resources to California, Florida, and the northeast. Imagine that, with just a handful of build-outs in a few dozen counties around the US, you could claim a network that covers “50% of Americans” regardless of actual physical coverage.

This is where the coverage argument between carriers, and fans of carriers comes into play, and it also shows how mundane the points are becoming. We are rapidly approaching a point where claiming to cover a percentage of the population does not mean anything. Half the population of the United States lives in less than 150 counties out of more than 3,000.

It will take time to reduce the gaps

Confronting that reality, it is easy to see why the future will be less about who is covered, but rather where they are covered. This is where AT&T and Verizon currently have the advantage. The original 800MHz network licenses are still in use to this day, and have comparable coverage and building penetration propagation as the 700MHz block that was auctioned off in 2008.

T-Mobile’s and Sprint’s networks (including their respective predecessors) are anchored on what is called PCS spectrum, 1900MHz, auctioned in the mid-1990s, and rapidly built out in population centers and along transportation routes. That explains why the two carriers’ coverage maps look so similar at first glance, and trace the interstate routes all over the country. Underpinning that coverage, things have changed a little, both carriers have since acquired lower band spectrum through acquisitions and trades, but on the whole, the PCS spectrum of both carriers, and the AWS spectrum (1700MHz) that T-Mobile also uses, is less efficient at penetrating buildings or offering uniform coverage over a particular geographic area.

That is why the 600MHz spectrum auctions to be held next year are so important to carriers like T-Mobile and Sprint. It also explains why Sprint opted out of the recent AWS-3 spectrum auction, keeping its acorns for 2016. In theory, winners of these licenses will be able to cover a geographically wider space compared to the higher-frequencies (not accounting for power levels or other RF factors).

Even with the spectrum auction next year, it will be several years before any of the winning providers are able to build out commercial service. The same holds true for the recently completed AWS-3 auction. Existing license holders have to vacate the spectrum.

The maps will still look largely the same – for now

Even as more and more spectrum is allocated to mobile providers to enable connectivity on an even more massive scale, it is a safe bet that large swaths of a carrier’s coverage map will remain largely unchanged, at least for the foreseeable future.

However, as the communication standards advance, the demands of a completely connected world will increase. Soon everything from our homes, automated appliances and equipment, cars, and more, will need a signal. We may yet see a map that is completely covered, connecting everything, everywhere, but for now, it's all percentages.

references: Business Insider and Deutsche Telekom (PDF)