My biggest project of every year is out, and I'm really excited. For the first time in 12 years of testing, 5G matters, and for the first time, T-Mobile won PCMag's Fastest Mobile Networks study.
If you've been following this newsletter, that shouldn't be a big surprise; T-Mobile's breakneck pace of building mid-band 5G has pulled it out ahead of its two big competitors, who are waiting for their new C-band spectrum to become available at the end of this year.
That means next year will be wild, too. It's a lot of fun. It's certainly more fun than that stretch of 4G in the mid-2010s when Verizon won every year and the only real surprise was seeing how badly Sprint would do.
T-Mobile's success and the new competitive network situation can be attributed to the company's merger with Sprint, though I still feel uneasy about it. I'm not going to go into counterfactuals or could-have-beens, but I'm worried that after five years or so, when regulators start treating a three-carrier market as status quo rather than a new innovation, everyone will get lazy.
(And I still don't trust Dish at all, not one bit. It's already delayed its market launch from the end of 2020, to early 2021, to late 2021, and now to 2022.)
We Show Our Work
I started Fastest Mobile Networks in 2010, when there weren't a lot of other network awards. Since then, a whole bunch of consultancies have moved into the space, with seemingly unlimited budgets and bigger data science teams than ours, including GWS, RootMetrics, Umlaut, and even our own Ziff Davis family at Ookla.
So I have to ask, especially this year: Why do I keep doing this? What do I bring to the table? The answer has to be: I show my work. The most frustrating thing about all the other tests, for me, is how opaque they are. Nobody outside of RootMetrics knows how RootMetrics calculates its scores. I'd love to see exact maps of Umlaut's drive tests, but they aren't in its public reports.
My difference is that I'm here as your representative, so I want to show my work. I show you the locations we tested on every page, tell you what phones we used, and show you the components of our scores and our weighting. I already see some people disagreeing with my conclusions, based on my own component data. (It's all about reliability versus average speed weighting.) And you know what? I love that! The whole point of this study is to give you, the smart reader, the tools you need to make buying decisions for yourself.
Like what you're reading? You'll love it delivered to your inbox weekly. Sign up for the Race to 5G newsletterSign up for the Race to 5G newsletter.
My work is also not perfect. I have limited time and resources, which means I can't hit every city I would want to hit, and can hit only a limited number of spots in each city. But it's pretty good. I feel like I've had a good day at work when I've added some knowledge to the world. Today is a good day.
Fastest Mobile Networks is an extremely big story. We have our main page, full of charts and graphs, and then 36 pages devoted to our 30 cities and 6 rural regions. So to try to better expose some of the things we've been discovering, I've added some standalone stories this year:
- On Verizon and T-Mobile, It's Time to Turn 5G Back On (this year we found Verizon fixed its 'nationwide 5G' configuration problem)
- Don't Want To Give Up Your Old Phone? These Carriers Have The Fastest 4G (the gap between T-Mobile's 4G and 5G is so big that it leaves room for AT&T)
- Where Are the Dead Zones and Hotspots? (pulling some details out of the rural data)
- The Best and Worst Cities for Verizon 5G (this one is self-explanatory)
- T-Mobile Customers in These 10 Cities Need New iPhones (because the old ones don't have 5G)
Fresh off the US results, I'm rolling into Canada now. I'll write more about that next week, and with any luck I'll be talking to Bell, Telus, Rogers, and Freedom about how they've expanded their 5G networks and what's coming next now that they've bought 3.5GHz spectrum. Have any questions for me to ask the Canadian carriers? Tell me in the comments.
What Else Happened This Week?
- A Galaxy A21 phone exploded on a plane and people have been texting me nonstop about "Samsung's exploding phone problem." Samsung has not had an exploding phone problem other than with the Galaxy Note 7. Lithium-ion batteries catch on fire sometimes. There are a ton of stories about iPhones exploding. Because most of the phones sold in the US are Samsung or Apple, most of the phones catching on fire will be Samsung or Apple.
- Google is going with Samsung modems rather than Qualcomm in its Pixel 6 phones. The trick here isn't in the modems, it's in the antenna modules. So far the only two millimeter-wave antenna modules we've seen used in the US have been from Qualcomm and Apple. How will Samsung's perform? We'll see.
- Our friends at Wave7 say the chipset shortage is now affecting everyone but Apple, and it's having weird ripple effects on our market. For instance: OnePlus share is up at T-Mobile because there's a shortage of Samsung Galaxy A32 phones, but it's down at Metro because there's a shortage of OnePlus N200 phones. This is going to continue to be a messy story for the next six months at least.
- Fierce Wireless and Signals Research Group have an interesting story about voice-over-5G. It doesn't come to a lot of conclusions, but I like anything written about voice calling. I think we'll finally be able to test voice calling in our 2022 Fastest Mobile Networks drives!