June 2021

AT&T has taken a measured approach to rolling out 5G. The only new spectrum that has been solely dedicated to the new technology has been millimeter-wave, only available in limited portions of certain cities. Nationwide 5G, meanwhile, uses a small portion of AT&T’s 850 MHz spectrum, with similar capacity and coverage characteristics as other low-band spectrum that has been used for LTE for years. 

In order to determine what benefit this 5G strategy is delivering for AT&T’s users, Tutela has measured the user experience of AT&T users with 5G-capable handsets in 10 urban areas with mature 5G deployments, and compared that experience to users with modern but 4G-only handsets in the same area. The results aim to answer the question of “If I upgrade to a 5G-capable device rather than a new 4G device, how different will my experience be?”

When looking strictly at user experience, as measured with Tutela’s Excellent Consistent Quality metric, there’s little difference between 5G and 4G-capable handsets. Excellent Consistent Quality is a measure of user experience, and it represents how often a connection is capable of demanding everyday use-cases, like streaming 1080p video, playing multiplayer online games, and performing a group HD video call. For those kinds of use-cases, there appears to be no noticeable difference for users who have access to the 5G network, compared to those who don’t. That doesn’t mean that there’s no performance difference — peak speeds may still be faster, so uses like downloading entire movies or apps may see a difference. But for the vast majority of applications that people use a cellular connection for, 5G does not move the needle on user experience. 

If we look specifically at tests taken when a device is connected to the 5G network, compared to 4G, there is a slightly more noticeable difference. Excellent Consistent Quality is nearly 5% better, while median download throughput increases by over 8%. While those numbers don’t represent an order of magnitude improvement over the already-excellent 4G network, it does indicate that the 5G network is capable of delivering a stellar user experience. 

The one surprise is the median latency, which is notably worse on 5G compared to 4G. Although 5G should eventually be able to support low-latency applications better than 4G, the current non-standalone 5G deployment that many operators use increases complexity, in some cases resulting in a slightly higher latency. 

Where 5G is offering a notably superior user experience is in congestion and peak-hour performance. Looking at the proportion of tests below 1.5 Mbps — a useful comparison for determining congestion — it is clear that more 4G tests are falling below that threshold, compared to 5G. 

In particular, that ‘minimum user experience’ is much more consistent on 5G than on 4G, when looking by time of day. While both network technologies have virtually no tests slower than 1.5 Mbps overnight — when the network is the least loaded — the 4G network slows perceptibly during busy hours. For 5G, that decrease is barely noticeable.

As it currently stands, some of the improvement in congestion is likely due simply to the lack of network traffic. 5G devices are still in the minority, and even those 5G devices only selectively connect to the 5G network. Over the course of this year, however, AT&T is expected to deploy some of its recently-acquired C-Band spectrum, drastically increasing network capacity — especially in urban areas, where it is needed the most. 

For now, users upgrading to a 5G device can expect to see minimal improvement in terms of overall user experience. For subscribers who live in areas that see problematic congestion, or for users who value a small improvement in download speed, the change may be more noticeable. Going forwards, however, we can expect the deployment of 5G in C-Band spectrum to bring a much greater improvement in peak performance as well as consistency. 

Get in touch today to schedule a demo, see Tutela Explorer in action and discover insights for your region.