I’ve been invited by David to post my thoughts on the acquisition of T-Mobile USA by AT&T from the perspective of a T-Mobile customer. I have slightly modified this article from the original I posted on Mobile Central to include a bit more information.
As a T-Mobile customer, I’m against the acquisition by AT&T. T-Mobile has a great reputation as a value leader and offering top notch customer service. In fact, they’ve been rated the highest in customer satisfaction by J.D. Power and Associates for several years now. T-Mobile also does an excellent job maintaining their towers and making them reliable. While they don’t have the 3G footprint of AT&T, they are getting there.
I’ve done my research regarding spectrum. I know that AT&T possesses quite a bit of spectrum on their own, much more than T-Mobile. They even own enough spectrum (somewhere close to 100 MHz, I believe) to build all over the nation. But, the problem is that their spectrum isn’t harmonized (meaning that they don’t have a single frequency band that is used all over their network). T-Mobile is fortunate enough to have at least 20-30 MHz of spectrum in every market in the AWS-1 (1700 MHz UL/2100 MHz DL) band range, with an additional 10-20 MHz in the PCS (1900 MHz) band range. AT&T’s spectrum is all over the place, some 10-20 MHz here and there in Cellular (850 MHz), some 15-20 MHz here and there in PCS, about 10 MHz in most areas in US Dig Div (U.S. Digital Dividend – 700 MHz), and about 10MHz in the western half of the country on AWS-1. This actually puts AT&T in a worse situation than T-Mobile, because their licenses are scattered all over the place. Without T-Mobile, they would require tri-band LTE devices right out of the gate, because they don’t actually have a single high quality spectrum band that covers all parts of the country. However, AT&T’s overall spectrum concentration in every single market that they both participate in (and several markets that T-Mobile doesn’t yet participate in), is much higher than T-Mobile’s.
Obviously, AT&T’s spectrum on PCS and Cellular is wide enough that they can run both 2G and 3G services on the same spectrum band. That means they have more than enough room to initially build out LTE on their current spectrum holdings. Provided, of course, that AT&T is willing to decommission their older networks in order to free up spectrum.
The spectrum crunch that AT&T tends to say that they have is complete and utter nonsense. For years, AT&T held some AWS-1 spectrum and they didn’t use it at all. Why? Because they didn’t want to force themselves to offer devices that would be compatible on their only competitor’s network (T-Mobile).
As far as band frequency support for LTE, I know for a fact that ST-Ericsson has several multi-band LTE chips that support most of the frequency bands used in the United States for 3G. Qualcomm’s Gobi chips that have LTE support also do support some of the frequency bands for LTE. Multi-band chips are not an issue, since the manufacturers were prepared this time. Nokia-Siemens Networks, Ericsson, Motorola Network Solutions, Alcatel-Lucent, and Huawei are also all ready for building out towers that broadcast LTE on any of the currently used frequency bands for 3G, and of course, they support US Dig Div spectrum. AT&T will not have a problem getting multi-band LTE chips for cheap, because they are massive enough that the economies of scale will kick in very quickly. They don’t need T-Mobile for any sort of spectrum crunch alleviation.
Another thing that bothered me about the AT&T announcement was the fact that they said that T-Mobile USA had no clear plans for 4G LTE. That is simply not true. They didn’t have definitive plans, but they did have several plans described for deploying 4G LTE. One plan was that they would re-farm their PCS (1900MHz) spectrum to use with LTE after they’ve built out their HSPA+ network sufficiently enough that they can begin taking down the 2G GSM network. That was to be enacted in 2012, when they began shutting down the 2G GSM network in areas that have HSPA+ built out completely. Gradually, they would have replaced 2G with 4G LTE. Since their 2G spectrum is pretty wide in most areas, the 4G speeds would have been pretty good. The expenses would have been the biggest issue, since it would have required several billion dollars. This plan was detailed during the January investors’ conference that T-Mobile USA and its parent, Deutsche Telekom, held.
There was also another option that was seriously discussed by T-Mobile USA executives for launching 4G LTE. That was doing another network-sharing agreement, similar to the one T-Mobile USA had with Cingular for their 2G GSM network before Cingular became “the new AT&T.” T-Mobile could partner with the many rural and regional carriers across the nation to combine their PCS spectrum and build out a nationwide 4G LTE network that would have a larger footprint than any other carrier in the country. They could also partner with Sprint and work together on building out a 4G LTE network on the PCS band. Considering the culture compatibility between T-Mobile USA and Sprint-Nextel, it could work out well for the both of them. They could have even partnered with AT&T, given the compatible technologies and complementary spectrum. Unfortunately, such a partnership probably wouldn’t work out so well because they have very different corporate cultures and values.
Well, if they had a plan of action, why did Deutsche Telekom announce the sale of T-Mobile USA to AT&T? Simply because they are reducing their scope from worldwide to just Europe, and probably later into just T-Mobile Deutschland. Their primary shareholder, the German government, probably wants them to focus more on the domestic market rather than international ones. That explains why Deutsche Telekom merged T-Mobile UK with France Télécom’s Orange U.K. last year to form the new holding company “Everything Everywhere” that manages the networks and brands of both T-Mobile UK and Orange U.K. Deutsche Telekom plans on exploring ways of minimizing their direct involvement with markets outside of Germany while still maintaining some sort of presence there. Expect to see Deutsche Telekom do similar actions for other T-Mobile branches across Europe that they wholly own.
Additionally, Deutsche Telekom gave up on T-Mobile USA in 2008, after the iPhone came the Cingular, which was transitioning to become “the new AT&T” at the time. I remember reading somewhere that Deutsche Telekom sent a memo in 2008 to T-Mobile USA executives saying that they missed their chance to become a significant player in the market, and the tone of the memo suggested that Deutsche Telekom was going to be more hands-off on managing T-Mobile USA. Later that year, T-Mobile was unable to acquire any US Dig Div spectrum because their parent chose not to participate in the auction. And of course, then T-Mobile brought out the G1, the very first commercially released Android phone, and it was a great success. The next year, Verizon brought out the Motorola DROID, which completely overshadowed T-Mobile’s myTouch line that launched that same year.
So, what does this mean? It means that the only way T-Mobile USA will get the needed cash infusion to build out a 4G LTE network is if the deal is blocked. The terms of the deal between AT&T and Deutsche Telekom state that if the deal is blocked or AT&T cancels the deal, T-Mobile USA will receive $3 billion and all the AWS-1 spectrum AT&T owns but isn’t using for 4G LTE at the moment of cancellation. The cash and spectrum infusion would allow them to build out their own 4G LTE network immediately.
I’ve noticed that very few people in the news media bring up the fact that AT&T’s acquisition of T-Mobile USA would effectively create a monopoly. Why would it do that when we have two other national carriers? It is because T-Mobile USA and AT&T Mobility are the only two super-regional carriers left in the USA that use the GSM family of technologies. The two competing GSM networks have allowed for a wide variety of devices to come into the American market. This is important because most people are too stupid to realize that their devices cost way more than they are actually paying up front, and that they demand it to be cheaper and don’t care that they are signing two year contracts. If the acquisition is approved, the number of GSM devices in the market will drop considerably. As of right now (March 30, 2011), there are nearly 300 different GSM devices currently available for sale directly from T-Mobile USA and AT&T. If the acquisition is approved, you can expect that number to cut in half, at least. Of course, it wouldn’t matter all that much because there’s only one carrier to use the devices on if AT&T acquires T-Mobile. You can also expect for prices on AT&T to get much higher because AT&T doesn’t really have a competitor that competes against them so evenly and on all fronts. Verizon and Sprint don’t really count because you can’t take your devices from AT&T and use them on Verizon and Sprint, since they use CDMA2000 with EV-DO Rev. A instead of GSM and UMTS with HSPA+.
AT&T is terrible at managing their network compared to T-Mobile. And the service and support has only really begun improving late last year. It will take quite a lot of time for AT&T to catch up in terms of overall quality to the levels expected by T-Mobile customers. And AT&T will crush T-Mobile’s open culture after the acquisition is complete, in favor of AT&T’s monopolistic conservative culture.
AT&T also doesn’t really get open networks and open solutions. Their Android devices are locked down, with sideloading blocked. They are the only carrier in the world that does that, by the way. AT&T prefers nickel-and-diming customers instead of making them happy enough to continue staying with AT&T, which is why they chose femtocells instead of UMA. UMA did exist before femtocells, by the way. UMA was built into the GSM standard with the EDGE and UMTS standards. In fact, femtocells rely on a variant of UMA technology to work. If AT&T acquires T-Mobile, they’ll probably kill off the G-series of Android devices, the ones that are pure Google Experience devices. UMA will probably disappear as well. And of course, no more Even More Plus plans and unlimited data plans.
There is some good news though. If the deal were to be approved, it would take at least three years after the acquisition is complete before they could take over the network and start repurposing the AWS-1 spectrum for 4G LTE. It is even likelier that it would take five years before they have everything in order to begin migrating the AWS-1 network to LTE. So our devices wouldn’t stop working for a very long time. Additionally, AT&T does practice grandfathering when it does acquisitions. Existing T-Mobile customers will never be forced off their current plans, even when they do phone upgrades into 4G LTE. But that is a small consolation prize compared to the larger problems AT&T would cause by acquiring T-Mobile.
Given all the real facts, it is definitely shown that AT&T should not be permitted to acquire T-Mobile USA.