RootMetrics second half 2015 US network report is out, T-Mobile shows data improvements


A couple of weeks after T-Mobile won the most awards in an OpenSignal network study, a new report has surfaced that puts T-Mo in the opposite position.

RootMetrics has published its Mobile Network Performance in the US study for the second half of 2015. Its results show T-Mobile finishing with an overall network performance score of 80.9 out of 100. To compare, Verizon won out with a score or 94.5, AT&T finished at 91.3, and Sprint earned a score of 86.0. Verizon ended up winning five of the six main categories — overall performance, network reliability, network speed, data performance, and call performance — and Verizon and AT&T tied in the text performance category.

Looking at T-Mobile specifically, Magenta won fewer Metro Area RootScore awards this time around, finishing with 209 outright and shared wins compared to the 221 it earned in the first half of 2015. T-Mo performed best in the data categories, with 16 outright wins for network speed and 14 wins for data. T-Mobile’s fastest median download speed was 44.9Mbps in Lansing, Mich., and its slowest was 2.2Mbps in McAllen, Texas.

T-Mo also improved in the Text RootScore Award category, growing by nearly 30 awards from Q1. It didn’t fare quite as well in Network Reliability and Overall RootScore, though, which RootMetrics says is “primarily because of additional blocked calls.” T-Mo’s Call RootScore Award count fell from 48 in the first half of 2015 to 19 in the second half of the year.


T-Mobile has taken issue with RootMetrics testing in the past, with CEO John Legere saying that the testers “can’t keep up” with how quickly T-Mo is improving its network. RootMetrics does note in its report that T-Mobile’s data speed improvements “correlate to the network’s slightly improved LTE coverage in this round of testing” and that “this improved LTE coverage in metro areas may be due to T-Mobile’s rollout of 700MHz spectrum.” RootMetrics does say that the devices it used for its second half 2015 testing were T-Mobile 700MHz compatible.

T-Mobile has also taken issue with the fact that RootMetrics doesn’t measure Voice over LTE calls in its study. John Legere revealed during today’s Q4 2015 earnings call that 50 percent of T-Mo’s voice calls are done over VoLTE, so that would have an impact on T-Mobile’s calling performance. The good news for T-Mobile is that RootMetrics will default to VoLTE for T-Mobile testing in the first half of 2016.

RootMetrics’ second half 2015 report was put together using 3.85 million tests conducted while driving 231,941 miles. The company says that it follows a scientific methodology with its testing, “unlike subjective surveys or summaries that rely exclusively on random crowdsourced data.” T-Mobile has done well in recent crowdsourced reports, winning the most awards in an OpenSignal network comparison and earning the title of “Fastest Mobile Network” from app maker Ookla. Considering all of that, I’d be willing to bet that John Legere and T-Mobile have something to say about RootMetrics’ latest report in the coming days.

Source: RootMetrics

Tags: , ,

  • Frankwhitess

    Wait… Isn’t RootMetric sponsored by Verizon ?? And to be honest, Verizon might be number one, but come on.. Let’s look at how much and how far t mobile has come in the last 3 years… Kudos for what they have done, and kudos for saving me money while giving me a awesome plan for the family… T mobile might be the discount carrier, but it does the job and more for me ☺….

    And last time I checked, not everyone is driving no 239k miles… Hell, most people don’t drive more than 100 miles a day.. So having service in North Dakota while I live and work in nyc is pointless to me… Folks will go with what works and bottom line, saves you money…

    • J.J.

      i agree its all about your usage, if you stay in a metro 95% of the time tmo is great. ive been with them for 10+ years and have definitely seen the heart ache of no service. those instances are far and between now but it still happens where att/vz will have service and i dont, but the features and cost far outweigh those few instances for me.

    • x

      I don’t know, but to me it looks like RootMetrics is designed to highlight Verizon’s network. Yes Verizon is the best, but not by very far as this results would like you to believe.

      6 awards for Verizon and 1 for AT&T doesn’t say the whole story, doesn’t help you understand and doesn’t help you decide… I think by design.

      • Mschmal

        Verizon is the best? My friend with a 2nd Gen Moto X on Verizon still can’t use data on a call. Verizon should be disqualified by this.

        • Acdc1a

          Verizon and Sprint are both complete rubbish in South Florida. If you want to have reliable service you have no choice but to use T-Mobile or AT&T.

        • SirStephenH

          That’s CDMA for ya…

    • AS118

      I agree with you, but back when I had Verizon, when I had to travel to the boonies for work-related reasons, I was glad to have coverage there. That said, I live and work in a local, fairly urban area now, so I have T-mobile. I don’t need coverage everywhere anymore.

      • Acdc1a

        I’ve traveled extensively with T-MOBILE over the years. Only one place I’ve ever been without at least roaming coverage, Ocracoke, NC.

  • Angel

    I still say RootMetrics ‘accurate test’ aren’t as accurate as it claim. Not only for their talent to used devices that lack something. But to be fair on EVERY carrier, they should use the top 3/5 devices to test coverage. Sometime one lost signal not because of the carrier but because of the phones. I can be standing in the same place as 3 of my friends on T-Mo, someone lose signal wile the others work great, or only one get signal. One of them have exactly the same device as I do. And he tells me of places where he have lost signal or his internet has gotten slow. And I would be on that same place. With the same phone but, with the only difference is that I neglect to update it because it’s working well. So I don’t want to risk it. And I would get signal no problem. So yeah, unless they use multiple devices at the same spot. I will always doubt that their shortcomings of some places could be the device and not the carrier. And that goes for all carrier.

    • Fabian Cortez

      It’s old data and they didn’t test VoLTE, which 50% of T-Mobile’s calls go over.

      RootMetrics is no longer credible as a technology testing firm.

      • Angel

        Yeah, I don’t see how they can be take seriously. They now say that the dvices it used for its second half of 2015 were 700MHz compatible. Funny not a month ago they said that they used the same device in the first half and second. and that bein a Galaxy S5 I don’t know how they magically got compatibility. They say its a scientific methodology BS. And now they say that they’ll default to VoLTE for T-Mo, I wouldn’t be surprise if they just judge using that, and take point when the call can go trough using VoLTE. They don’t have default to anything. They should just place the call. The network will know what to do. It’s 50% calls, so the others 50% its not. And even if the place does have VoLTE if you FORCE the device to use LTE for call, on a spotty place you might get a lower quality called than if you let it go down to WCDMA/UMTS/GSM whatever the hell it wanna use when its not using LTE. Their ‘scientific methodology’ sucks. Its like judging the performance of a device on how high it can get on a benchmark. Yes that help, but some device get super high numbers and still lag or stutter when using the device daily. Wile there are some others that don’t rule the benchmark and still work exceptionally well without stutters. They should really work on how they test the networks before they make any claims. I fail to see anything ‘scientific’ about it. PS, I did read that they suposedly change to a Note 4, which do support the 700, but they didn’t say that untill people started to question it. There had already been report that they use the same device for consistency. So yeah not sure what to believe.

    • dtam

      yeah, just use the latest iPhones and latest galaxy’s. every carrier carries those 2 phones.

  • Robert V

    Root metrics means nothing when you have to shell out 710 for 100gb a month for slower data, yes we all know Verizon has the most towers period. We have heard this story over and over again, however T-Mobile is growing faster and has become extremely reliable so unless Verizon comes into the light of reality with its pricing they will continue to lose customers to T-Mobile.

    • maximus1901

      their churn is less than 1%

    • Fabian Cortez

      T-Mobile actually has more towers. Though that doesn’t mean coverage is better.

    • Mike

      Verizon has a ton of roaming agreements hence why their coverage is better then everyone

  • mikeZo6

    Tmo speeds have went down big time end of 2015 cause all the new people added to tmo network, their network cant handle all these subscribers, tmo needs to focus on improving their network…

    • AS118

      I’m concerned that Binge-On and Music Freedom and other initiatives like this may cause more problems like this down the line. Having “unlimited” plans is already something that threatens their future network stability and saturation, but making more and more data hog services “free”, even throttled is something that may seriously backfire, imho.

      • EndlessIke

        I think I read that the net effect of Binge-On has actually been to lessen traffic, because people are using much lower bit-rates

        • patt

          pretty much that :D and Music-Freedom spotify even at 320kbps won’t stress that much network.

        • x

          I’m not sure about traffic. The P3 report was in part talking about overall consumption.

          If the report was accurate and if I’m remembering correctly, it said that people on the study were watching videos 3 times as before. So, maybe people watching for more hours could be congesting the network at peak hours.

        • x

          …where before they were not watching as much video at peak hours because they were watching their consumption. I’m just speculating.

        • SirStephenH

          I’ve read that people are just watching more video because of it and data consumption is actually going up.


      I completely agree with you

    • x

      When I turn off LTE sometimes I get faster speeds. I think B12 is saturated and without CA it really sucks. I’m gonna try my old non-B12 phone one of these days to see how better or worse is with B2 and B4 LTE.

      Does anyone know if they are already deploying LTE where there was HSPA/HSPA+ before at AWS?

      • Trevnerdio

        What phone do you have? You can usually manually switch your LTE band.

      • on certain android phones you can change the preference of LTE band. What phone do you have?

    • Jason Caprio

      Agreed. Whenever I am in a city or highly populated area, my speeds are abysmal. In Camden, NJ, Trenton, NJ, and in many parts of Philadelphia, PA, I am lucky to break 1 – 4mbps, however if I’m in a suburban/rural area with LTE coverage, I can get upwards of 80mbps. T-Mobile has some serious network capacity issues to deal with.

      • kgraham182

        This is what I’m experiencing too, here in south Florida. There’s pockets of greatness than there’s where I normally roam with averages speeds you mentioned. South Beach in between all the high rises, T-Mobile is a no go.

        • Jason Caprio

          I was at Universal Orlando during the week of Christmas. In the park, data speeds were so bad, the phone was useless. Speed tests showed speeds of 0.16Mbits/sec. Downtown Orlando the speeds were decent. A friend’s house in Clermont, FL LTE signal was weak, speeds were around 0.1 – 0.4MBits but switching off LTE and using HSPA speeds were a “usable” 3mbits.

        • Bradley Karas

          Improvements are coming to FL my friends…they are adding 350 new towers in the state by the end of Q1 according to my rep who had a big meeting with the techs in the area. These are known issues to them apparently

        • Trevnerdio

          You were able to complete a speed test on LTE at Universal?! Ha! I got “error” on either upload or download each time I did it.

    • SirStephenH

      My speeds have been going up due to the deployment of B12…

  • SEBA

    Everything went down. Speeds, calls and text messages. Advanced text messaging and VoLTE are so inconsistent causing more issues

  • gr8gatzby

    I have 19mbps 3G at my place. LTE is 26mbps..

  • John Brown

    I think anyone who has FULLY UNLIMITED data from T-Mobile should download Root Metrics from the play store and constantly test the service as they commute around their areas, much like we did with Sensorly. I seriously think the results for T-Mobile have been skewed by throttled tests as root metrics uses data for testing and a 3gb plan can be blazed through in 5 minutes using consistent testing.

    • SirStephenH

      RootMetrics doesn’t count against your data on ALL T-Mobile plans.

      • Where did you get this info from? Was this fairly new info? As far as I know speed tests run on the ookla speedtest app don’t count against data on T-Mobile plans. Sensorly and RootMetrics did.

    • Fabian Cortez

      RootMetrics doesn’t use crowd sourcing for their reports, unlike OpenSignal which uses crowd sourcing and their own testing.

  • SirStephenH

    RootMetrics is BS. First they outright ignored B12 in the first half of the year, claiming that that couldn’t find a phone provided by all four networks that supported it despite several choices being available. Then they continued to ignore VoLTE through the second half despite most calls being made over it. Seems like they’ve been trying to stack the deck against T-Mobile.