After T-Mobile’s announcements last week, I spent some time over the weekend evaluating T-Mobile’s work over the last two years. We all remember how T-Mobile was effectively paralyzed by the attempted acquisition by AT&T in 2011, and how it really messed up the company. Initially, I felt that T-Mobile’s announcement of the modernization of its network throughout 2012 was a great plan. I was hopeful that this would finally mean that T-Mobile will bring its awesome HSPA+ network to its entire native footprint this year.

Obviously, that hasn’t happened. Instead, the large metropolitan markets were once again upgraded first. I understand why that happened. The larger markets offer a quicker and larger return on investment than the smaller ones. Smaller markets upgraded often have a high percentage of T-Mobile subscribers, so the return on investment was justifiably large enough to do it. It probably also helped to convince Apple to sign an agreement with T-Mobile to allow T-Mobile to carry Apple products with minimal changes required.

But T-Mobile needs to focus on its entire footprint for 2013. The vast majority of the largest markets have already received upgrades throughout 2012, and will certainly be part of the LTE deployment for the first half of 2013. However, T-Mobile has a very large footprint of markets that do not have access to any mobile broadband services above EDGE. Or worse, it relies principally on roaming to serve the area. While these markets are typically considered “small” markets, they do make up a rather significant portion of the population of the country. It is a bad idea for T-Mobile to continue ignoring them.

The entire states of West Virginia, Montana, North Dakota, and South Dakota have zero native coverage. Ostensibly, these markets are covered by the UMTS roaming agreement with AT&T that went active this year. Alaska has zero native coverage, but it served with roaming agreements with local GSM operators in some areas. Geographically, large amounts of Washington, Oregon, Nevada, Idaho, Colorado, Utah, Iowa, Arkansas, Mississippi, Alabama, and the New England area are not covered by a native network either. T-Mobile has spectrum licenses to deploy in these areas. For one reason or another, it has chosen not to. But T-Mobile needs to do something soon about those states, because it has a lot of spectrum that it paid a lot of money to get. If it doesn’t do something with them, it could lose those licenses at the end of the decade.

In the areas where T-Mobile does have a native network, most of it is 2G-only. A bit of it is still GPRS, but nearly all of the 2G network has EDGE now. With the major markets all modernized, T-Mobile needs to push forward in markets that have been neglected over the years.

Here’s a pair of example markets: Oxford, MS and Starkville, MS.

Oxford, MS is an interesting choice because it is surrounded by native 3G/4G coverage, but the area itself has only 2G coverage. According to FCC data, T-Mobile has at least 10MHz (maybe 15MHz) of PCS (1.9GHz, band 2) spectrum and 10MHz of AWS (1.7/2.1GHz, band 4) spectrum. It isn’t likely to get refarmed unless the entire subscriber base uses newer phones that can run entirely off of the 3G/4G network. Despite having one of the largest universities in the state and a diverse customer base, T-Mobile has not yet deployed any AWS service at all.

On the other hand, we have Starkville, MS. Starkville has at least 30MHz of PCS spectrum and 20MHz of AWS spectrum. Up until mid-November, it was a 2G only market. Like Oxford, it is a college town (home to Mississippi State University). While the official coverage maps have not been updated yet, Sensorly shows that Starkville does have 3G coverage now (and I’ve verified it is indeed active). I’ve even personally seen one of the cell sites being upgraded to the new modernized equipment this past weekend!

I brought up these two markets because they represent spectral situations that many 2G-only markets have (I’ve also been able to personally verify coverage accuracy there). The western half of the country is PCS-deficient, so more aggressive switchover to 3G service is required before refarming can happen. The eastern half is more often than not like Starkville, where it is AWS-deficient. The various deals T-Mobile has entered into this year (including the one to merge with MetroPCS) will improve the AWS situation in many of those areas. Not to mention, many markets have really old equipment up there. The equipment in place in Starkville and Oxford was put up by Powertel at the turn of the century. That makes it over ten years old! Equipment that old and suffering through natural disasters does not perform that well, either. These markets badly need the upgrades, and I’m certain that other areas are just like these.

T-Mobile’s most important task should not be deploying LTE, though that is something that needs to happen. Its biggest issue is that it has not deployed UMTS across its 2G-only footprint to drastically reduce the load on the 2G network so that capacity can be scaled down for M2M and 3G service can be lit up on PCS in as many markets as possible. T-Mobile has national access to AWS spectrum to deploy UMTS service, so it can do it. Granted, there are some markets that it is totally unfeasible to deploy UMTS service on PCS. That shouldn’t mean that T-Mobile should not deploy service on AWS. In all markets with that situation, there’s enough AWS spectrum to go around. Many of them can have both HSPA+ and LTE on it.

T-Mobile’s biggest (and most important task) for 2013 should be the elimination of 2G-only markets. Either through direct deployment or setting up a program similar to Verizon Wireless’ 4G LTE in Rural America program to have other companies do it for them (which would establish reciprocal roaming and limited network sharing agreements), it needs to get done. While it is unfortunate that many rural carriers offer only CDMA or 2G GSM service, such a program by T-Mobile could encourage a switchover and a vast improvement in the UMTS coverage situation. It may even offer a path for more constrained carriers to deploy LTE who don’t want to deal with Verizon Wireless.

In any case, T-Mobile’s primary focus next year should be on upgrading its entire 2G-only footprint to 3G/4G service. That would give T-Mobile customers a better quality of service and allow for T-Mobile to reduce its overall costs by dramatically simplifying its physical network and using newer, more reliable equipment. T-Mobile and its subscribers would both be very happy when that happens.