Internet delivered through fixed wireless access (FWA) may account for a majority of traffic on the nation’s biggest mobile networks, but analysts say revenue from these subscriptions makes up a tiny percentage of revenues.
FWA makes up around 6% of revenues at T-Mobile and around 3% of revenues at Verizon, according to the new report from the analysts at MoffettNathanson Research.
Nevertheless, FWA may use more than half of all traffic on the Verizon network and more than two-thirds of traffic on the T-Mobile network (including traffic from mobile virtual network operators).
On average, each FWA customer uses about 635 GB of data per month, the analysts estimated. They used data from the networks themselves and from the data analytics firm OpenVault.
That puts FWA data usage on par with usage from cable internet customers.
Meanwhile, mobile users eat through about 17.2 GB per month, according to data from the Cellular Telecommunications Industry Association (CTIA).
The MoffettNathanson report questions whether FWA should still be understood as using “fallow capacity” and says there are still concerns about FWA network capacity that may limit its growth.
The new analysis aligns with OpenSignal research from 2024, which found that “Wi-Fi networks shoulder the brunt of our users’ phone data usage,” in the U.S., even outside the home. It suggested that around 80% of screen time was spent on Wi-Fi.
MoffettNathanson figures also align with predictions of FWA growth globally. An October 2025 analysis from ABI Research said FWA revenues were far from peaking. Researchers expect a compound annual growth rate of 18.8% between 2023 and 2030 and say the overall data market could reach $111 billion by 2030, with annual data usage of approximately 1 Zettabyte.
By 2030, according to ABI Research, about 233 million people will subscribe to FWA globally, and 81% of those people will be on 5G.
The MoffettNathanson analysis doesn’t update investment recommendations but concludes that FWA networks could “prove incompatible with a future where AI traffic loads grow significantly.”
