AT&T and Verizon had big 5G-related announcements this week: AT&T published speed test results that seemingly validated its network (including its 5G E network) as the fastest around, and Verizon launched its 5G network in parts of Chicago and Minneapolis. But both of these announcements underscore just how much of a mess 5G is right now. AT&T’s results appear to be skewed in the company’s favor, and Verizon’s rollout seems slapdash, with poor coverage even in the areas that Verizon promised.
These are just the latest headaches for 5G, which has been marred by delayed rollouts, limited hardware tests, conflicting standards, political wrangling, and more. With telecom companies rushing to be first, odds are the mess of 5G will only get worse as the rollouts continue. If people’s first experiences with 5G are this shoddy, why should they trust — and pay extra — for the networks when they do actually arrive for real?
Take AT&T. The company seemingly scored a win this week by announcing that recent Ookla speed tests had found its network to be the fastest in the US, in part fueled by the company’s recent 5G E network (which again, is LTE, not 5G) — buoying the company’s message that the 5G E brand would help cement AT&T’s reputation ahead of its actual 5G launch later this year.
But those results aren’t as clear-cut as AT&T would have you believe: Ookla says that the spike in results for AT&T is due to an increase in speed tests from iPhone users after the release of iOS 12.2, specifically on the iPhone XR, XS Max, XS, X, 8, and 8 Plus, which are the same models that now display 5G E logos following that update.
AT&T’s network looks faster because more iPhone users — confused by the new logo — ran tests
Ookla believes that the recent AT&T speed increase is simply a result of iPhone users seeing the new 5G E logo and retesting their devices to satisfy their curiosity, thus skewing AT&T’s average speed results with an influx of tests from newer devices. In other words, AT&T’s network looks faster since it had more high-speed devices running speed tests to factor into its average than competitors.
It’s sort of a self-fulfilling prophecy: by making iPhone users question their network, AT&T was able to juice its average speeds by adding an influx of new data to its sample that its competitors didn’t have, thus making the lie that 5G E is somehow better than LTE an apparent mathematical reality.
Now, those speeds aren’t entirely misleading: Ookla says that, on the whole, speeds on 5G E devices really were faster than AT&T’s average speed — which makes sense since these are the devices that are designed to take advantage of the LTE-Advanced innovations that AT&T is using here. Ookla does note that AT&T was already on track to have the fastest network that quarter even before iOS 12, and that the 5G E tests helped solidify that lead.
But these latest tests don’t prove that AT&T’s speeds are faster on average, and other results — like OpenSignal’s recent study from before iOS 12.2 was released — shows that AT&T’s 5G E is actually slower than T-Mobile and Verizon. Verge readers on Twitter have also shared similarly poor results; it’s certainly not representative of all users, but it’s definitely not the kind of first impression AT&T wants attached to its 5G brand, either.
Things don’t necessarily improve when you get to actual 5G, either, as Verizon showed us this week during its real 5G launch. In his tests out in Chicago, my colleague Chris Welch learned that actual 5G will offer dramatically better speeds at rates between 400 and 600 Mbps for downloads. (Those are the kinds of numbers that AT&T’s 5G E network can only dream of.) But the network itself is extremely spotty. When Verizon says that only “select areas” will have 5G, it’s not kidding: service was reportedly inconsistent. Even where Verizon did offer 5G, it often appeared and vanished from one block to the next.
That’s a problem. Verizon may have technically launched its network first, but if it can’t offer widespread or consistent service, then it’s just that: a technical achievement without real practical application. Verizon’s rush to be first seems to be coming at the expense of a reliable 5G network, something that early adopters (who, remember, are paying $10 per month extra) will have to deal with while Verizon gets more substantial coverage.
Adding to the confusion: Verizon says that the 5G status icon will only pop up when you’re actively using 5G. That means you could be standing in a 5G spot and not even know it unless you’re actively using your phone. It’s the opposite of AT&T’s issue. Verizon seems content to hide the fact that you have actual 5G until you actually test the speed for yourself.
I can’t figure out why Verizon would do this, unless the goal is to keep the actual size of its 5G networks as vague as possible. That actually seems plausible, given the inconsistencies that Chris noted in his tests.
5G hype is outpacing the technical developments of the networks
All of this adds up to one of the biggest issues with 5G: the hype is outpacing the technical developments of the networks. That may not seem so bad, given that cellular companies love to promote their products and services, regardless of how they relate to actual real-world results.
But the stakes are higher with the launch of 5G, and if AT&T and Verizon (as well as T-Mobile and Sprint, which seem to have made the right call in delaying their launches to iron out details) can’t actually make good on things, they’re going to end up with disappointed customers once they find out that the reality doesn’t quite meet the promises they were given.
Update April 8th, 9:15am: Clarified Ookla speed test results to note that AT&T was already on track to have the fastest speeds even before the 5G E results from iPhone users came in.
Update April 11th, 4:25pm: Clarified that AT&T’s network claims aren’t exclusively based on 5G E speeds.