One of the most questionable facts circulating in the broadband industry is that a large percentage of homes in the country have multiple ISP options. I wrote a recent blog about a survey done by U.S. News and World Report. One of the facts noted in that report is that the FCC data shows that 94% of homes have a choice of three or more ISPs.
I’ve seen similar statistics elsewhere, and it’s not hard to see that this information comes from the latest FCC mapping data. I’m not surprised to find that the FCC maps show that 94% of homes in the country have three or more ISPs claiming the ability to provide service. But that is very different than saying that folks have three real choices of ISP.
Consider my home. I live in a city, and the FCC map shows that I have nine choices of ISP. But do I really? Some of the ISPs listed as serving my house are not viable options. For example, UScellular claims to offer a connection of 2 Mbps download and 0.064 Mbps upload. I live in a hilly and woody city, and I doubt that even that slow speed is actually available. Another local cellular company says I can get 25/3 Mbps, but I am skeptical in my hilly and wooded town. AT&T is listed with two DSL options at 75/20 Mbps and 18/2.5 Mbps. I have three satellite options listed, but my home is surrounded by mountains on all four sides and we have a limited view of the horizon. The fastest satellite option listed is Viasat at 100/2 Mbps – which is much faster than what speed tests indicate. In the previous version of the map, T-Mobile showed a speed of 0.2 Mbps download at my home but didn’t claim coverage in the latest maps.
I’ve looked at the similar details in the FCC maps at counties all around the country, and a large percentage of the ISPs available at most homes are not real options. If it wasn’t so expensive to do so, I’m tempted to try to buy broadband from every ISP listed at my home to find out if they can really serve here – and if they can deliver the speeds listed on the FCC map. My bet is that due to the geography in my area, that three or four of the options listed are not really available at my home and that most of the ones listed would badly underperform the speeds claimed in the FCC maps.
But that’s only half of the story. I think a large percentage of families these days like mine and want a broadband connection that is vigorous enough to support the way they use broadband. There are only a handful of broadband options for homes that want high-quality broadband.
- Fiber is the gold standard of quality broadband.
- The two biggest cable companies together have over half of the nationwide broadband subscribers. I’m on Charter, which claims I can buy download speeds up to a gigabit.
- Verizon 5G Home Plus is essentially fiber-to-the curb. This is available in limited markets, but I’ve seen recent speed tests for the technology with gigabit download speeds.
- Customers within a mile or so of towers equipped with FWA cellular wireless are seeing speeds from 100 – 300 Mbps download. But outside of that distance, the speeds drop quickly. The carriers also throttle broadband any time that cell phone usage gets heavy in a given neighborhood.
- Some fixed wireless WISPs are delivering fast speeds with the newest radios. Speed tests show that most WISPs have not yet upgraded, but faster speeds are on the horizon for WISPs that will make the upgrades.
- Starlink is a great option when there is no cable company or fiber available – but how many people in cities are willing to pay more for a slower option?
I work at home and am online a lot, and I would never take the chance of putting my businesses on an underperforming ISP – and I think a lot of people feel the same way.
I’ve lately been using the term Competition Gap to describe this situation. While many ISPs claim to have wide coverage, most households don’t feel like they have real options. In my city, it’s hard to picture somebody working at home who isn’t using Charter or the various pockets of AT&T fiber. For now, the FWA towers are too far outside of the city to be a viable option.
I think most folks in my city would tell you they only have either one or two real choices for quality broadband, and they would be laugh if you told them they had nine. ISP. There are certainly still people in my city here who use DSL. There might be a tiny handful using the satellite or cellular broadband. But most of the folks choosing these slower options are either casual broadband users or can’t afford the big ISP broadband prices.
The idea that 94% of U.S. homes have three or more real broadband options is ludicrous. I would guess that the percentage of homes in the country with even two quality choices is significantly lower than 50%.
The FCC hasn’t issued an annual broadband report since 2021, probably due to not having a fifth Commissioner. It will be interesting to see what the FCC claims about broadband coverage when it finally issues a report. I bet the FCC report say something along the same lines as mentioned in the U.S. News and Word Report survey – that the state of broadband is good and that almost all Americans have multiple choices for good broadband. But that would be political fantasy that ignores what most households know – most of us in cities have only one or two choices of quality broadband. Many folks in rural areas still don’t even have one.
Here’s the last post I did on this: https://mikeconlow.substack.com/p/the-us-is-covered-in-cable-broadband
5% – 3+ wired offerings
32% – 2 wired offerings
49% – 1 wired offering
14% – 0 wired offerings
Just back from Wispapalooza where I talked with a number of fiber providers who say they all work together in their areas. IF ISP1 builds out in these 5 blocks IPS2 doesn’t build there, they move on to the next 5 blocks. Nobody over building anybody because if you figure a 45% take rate, 2 ISP’s is just a no win situation. So why would we want more than one fiber ISP at a service location? If it’s just one the free market can afford it. If we demand 2 providers per location then the gov has to subsidize it. Which means we pay for it in taxes.
In the fixed wireless world I can see a lot more reason to have overlap because it’s per house on service or not, not per block like fiber.
remove the wired from the list.
Internet service is about deliverable speed, latency, reliability, and loss/jitter/bloat. The underlying tech doesn’t matter. Yes, it’s easier to get there with certain technologies, but there are a LOT of bad fISPs out there that are solidly beaten by a good wISP. The technology doesn’t matter, it’s the deliverables that matter.
I’m glad to hear there are places where WISPs are beating fiber providers, because that’s how competition is supposed to work – customers will choose the ISP they trust. I’d be interested in writing about a market where a WISP is winning more than half the customers competing head-to-head against fiber.
It’s worth noting that the vast majority of people live in cities where there is no good WISP option – none claim coverage where I live, and this is true for most of the urban areas where I dig deeply into the FCC maps.
But the point of the blog is that most folks don’t have three or more choices of quality ISP – certainly nothing like the 94% you would get by assuming that every ISP listed in the FCC maps is a valid option. In my own city, most people only have the choice of cable or DSL, with some pockets of fiber.
I’m sure in rural areas that WISPs tell prospective customers that they are superior to Viasat, DSL, and hotspots because of price, speed, or latency.
I know this is the debate that we’ve all had over and over, but…
I think the “advertised speed” on the FCC maps is not realistic, and particularly so for WISPs (not all WISPs, but lots*). And while I’d like to agree that it’s the deliverables that matter not the technology, the technology is the best view of deliverables right now, in my view.
I’d be interested if we could all get behind “expected speed” with measurement/verification and enforcement. If we can do that, I’d agree deliverables should start to be a better measure than technology.
*I admittedly don’t have data on this and I’m not aware of data that exists
Not disagreeing with that but I’d like to see expected latency as the primary number with speed on the side. We have focused too long on speed.
Mike, I’ll argue that the tech is a the best or even a good view of deliverables. If that were the case that Coax systems would be dominant because they sell higher class plans than most fISP and wISPs. I’m seeing lots of 100-400Mbps fiber plans out in the wild for $100+.
ONLY deliverables matters. The underlying tech has performance ranges but within those, it’s entirely up to the provider on how good the service is.
I don’t like the ‘advertised’ speeds on the broadband map for anything other than the pricing list. I would much rather see reported/measure speeds from devices. Not speedtest.net results but actual measure peak throughput with latencies and bloat attached.
I’m with Trendal here that latency/jitter/bloat needs elevated to be shown with equal emphasis and might argue that after you’ve met a threashold of say 50Mbps or so, latency is now 99% of what’s important.
Strong agree with all this… https://blog.cloudflare.com/making-home-internet-faster/
Still, given the available data, and for this particular question (how many broadband choices do Americans have broadly), I still think wired (fiber + cable) from the FCC is the best way to show it.