In findings that drop some rain on US broadband's ARPU parade, the Wall Street Journal issued results from a study this week indicating that many consumers are probably overpaying for the speeds they are getting.

The study, involving 53 WSJ journalists over several months across multiple major US cities conducted with researchers at Princeton University and the University of Chicago, show that typical US homes get only "marginal gains" by upgrading to faster speed tiers.

In the studied cases, the report found that most use a fraction of their available bandwidth to watch OTT services from the likes of Netflix and YouTube, and even in cases when they are viewing multiple OTT services at the same time. The study also found that improvements in picture quality and streaming startup times were negligible at higher speeds.

In response to the study, Comcast official told the paper that demand for speed and capacity continues to grow, adding that "Video is only a piece of the speed story." The study focused on "major streaming-video applications," and did not focus on other apps and devices, such as web cameras and online gaming, whose performance is typically aided by low-latencies rather than raw speed.

Of course, the speeds advertised by broadband service providers represent the maximum marketed speeds for a given tier, as it's rare that consumers actually require sustained speeds at the maxed-out levels they are paying for (a running joke is that Ookla's Speedtest is the killer app for 1-Gig broadband services). Still, the idea is to overprovision available capacity to ensure that there's always enough headroom and avoid data bog downs.

Still, the study argues that most household Internet data usage falls woefully below advertised maximums and that many consumers would be perfectly fine with slower, less expensive speed tiers than they are presently pay for.

"For many people they are not going to see huge differences between 50 Mbps, 100 Mbps and a gigabit per second," Nick Feamster, a University of Chicago network-performance expert and member of the study's research team, said in the WSJ piece.

Among examples highlighted by the study, a Comcast customer in Philadelphia named Peter Loftus's median usage over 35 viewing minutes was 6.9 Mbit/s, about 5% of the capacity he pays for.

The study's findings put a bit of a damper on the broadband business growth models for many cable operators and other ISPs and could rekindle the debate over usage-based data policies. While some operators have readily admitted that heavy usage doesn't necessarily have a big impact on network congestion, they have argued that data plans are necessary in the name of "fairness." Critics of those policies maintain that such policies gouge consumers and help ISPs profit from the ongoing consumer shift to OTT-delivered multichannel and subscription VoD services.

Even though subscriber bases are growing, at least among US cable operators that now view broadband/connectivity (and not pay-TV) as their core business, that rate of growth has been slowing. Meanwhile, cable ops and other ISPs have been pointing to the rise in data consumption and the shift in subs that are taking faster speed tiers or moving to unlimited data plans. Some recent examples:

Data consumption is rising
While the study does stir debate about what speeds most consumers actually need and should pay for, there's no denying that consumer Internet data usage continues to rise.

Cisco Systems's most recent annual Visual Networking Index found that the number of homes exceeding 1 TB in data per month more than doubled from 2016 to 2017, and that traffic in homes that have cut the video cord is 72% higher than the average. Also, in May, OpenVault found that "power users" (those that consume at least 1 TB per month) doubled in Q1 2019 to include about 4.2% of all broadband subscribers covered in the study.

Even as downstream growth trends appear to be slowing down a little, much of cable's current next-gen network activities, like Full Duplex DOCSIS and Extended Spectrum DOCSIS, take aim at cable's historic weak link -- the upstream. DOCSIS 4.0 will include both FDX and ESD and tie in new low-latency capabilities.

A sizable majority (80%) of US cable operators now offer up to 1-Gig in the downstream (thanks in large part to DOCSIS 3.0 and D3.1 network upgrades), according to NCTA estimates issued in January.

Even though most consumers don't need 1-Gig, cable's already thinking about a bigger and faster future. The industry's "10G" initiative, which envisions networks delivering symmetrical speeds of 10 Gbit/s, is underway amid expectations that new apps and services, such as holodeck-like light field displays, will continue to drive up capacity requirements in the years ahead.

Related posts:

— Jeff Baumgartner, Senior Editor, Light Reading