This almost sounds like a silly question when looking at the recent growth of FWA wireless, where the big cellular carriers are selling home broadband using cellular spectrum. But I find myself having to ask the question after discussing this with some of my peers. Universally, every person that I put the question to dismissed FWA wireless as a temporary technology with no real long-term legs.
The observations I heard about FWA included the following:
- The product is already oversubscribed, and customers are already getting tired of it.
- Cell towers weren’t designed to handle all-day broadband connections, and this is going to harm the carriers more than help them in the long run.
- The speeds aren’t fast enough, and people will tire of the broadband performance.
- The coverage is spotty and swings from fast to slow during the day.
Interestingly, the numbers seem to be telling a different story. Consider national quarterly broadband growth over the last two years, as shown in the following table. These numbers come from Leichtman Research Group, and they say the big companies represented in these numbers represent about 95% of all broadband subscribers in the country.
|
FWA |
Big Cable Companies | Big Telcos |
Total |
|
| 1Q22 |
532,000 |
482,830 | 50,350 |
1,065,180 |
| 2Q22 |
816,000 |
(60,289) | (87,837) |
667,874 |
| 3Q22 |
920,000 |
39,035 | (136,101) |
822,934 |
| 4Q22 |
913,000 |
55,527 | (95,204) |
873,323 |
| 1Q23 |
916,000 |
67,472 | (21,196) |
962,276 |
| 2Q23 |
903,000 |
9,566 | (61,833) |
850,733 |
| 3Q23 |
941,000 |
4,777 | 3,392 |
949,169 |
| Total |
5,941,000 |
598,918 | (348,429) |
6,191,489 |
Since the beginning of 2022, T-Mobile and Verizon FWA captured 96% of net new customer gains. While cable companies grew by 600,000 over that time, 80% of those gains were in the first quarter of 2022. Since then, the big cable companies have stagnated and collectively added less than 20,000 net new subscribers per quarter. Telcos have continued to bleed customers as they continue to lose DSL customers, but underneath those losses, the telcos are replacing DSL customers with fiber customers.
Some of the observations of FWA I heard ring true. Cell sites weren’t designed for the kind of connections people expect for home broadband. But after seeing this unexpected success, I’m sure that the carriers are working to bolster cell sites. I know that there are a lot of requests floating around the industry asking for faster broadband connections to cell sites. In the long run, accommodating FWA permanently means changing the way that cell sites operate.
It’s also true that the bandwidth is not guaranteed. Both carriers warn customers that FWA broadband might be curtailed any time there is a lot of demand for cellphone traffic. As much as the carriers love this new business, it’s hard to think they will endanger their cellular business. However, a few of the folks I talked to think that’s exactly what they are doing.
The one observation that nobody made to me was that FWA cellular costs less – particularly less than big cable broadband. The standalone list price for T-Mobile FWA is $65 and is $60 for Verizon. Both companies offer a discount for autopay, and Verizon offers an additional $15 discount for existing Verizon cellular customers. Both are clearly trying to grab market share. I saw an advertisement this past weekend for Verizon FWA with a price of $30 guaranteed for four years for existing Verizon cellular customers. In my experience, there are a lot of customers for whom price is the predominant factor.
While my peer group thinks FWA is a flash-in-the-pan, T-Mobile and Verizon predict they will collectively reach 15 million customers, about twice where they sit today. This is clearly the most interesting dynamic in the broadband industry, but I don’t think anybody knows for sure where these carriers are going to top out. But it doesn’t feel to me like they’ve already hit the top of the market.
I’m a Network Engineer at a WISP, and I think FWA still has lots of life left in it! Lots of great gear coming out from the likes of Ubiquiti and Cambium! We can offer speeds of 100/20 pretty easily these days without being overloaded! Ubiquiti Wave we can offer speeds of like 300/300! The higher end Cambium we could also do symmetrical speeds. (Although it seems like people don’t really NEED symmetrical speeds, it just sounds cool)
I think it’s a mistake to dismiss FWA completely, there are a LOT of areas that Fiber just doesn’t make sense from a financial point of view! I live in an area that would cost tens of thousands of dollars per person! Is fiber the best? Yes, hands down no question… but the question that needs to be asked, is “Is this the best tool for this job” In some cases Fiber is the answer, in some cases FWA is the answer, and in some cases even satellite is the answer. All three technologies need to be looked at from a complete un-emotional, objective view and judged as such.
I think another aspect that isn’t looked at much, is a lot of these cellular carriers are using protocols like LTE, 5G, etc that wasn’t designed for FWA. Ubiquiti, Cambium, Tarana, etc all make gear that is specifically designed for the FWA industry, and that makes a big difference, using the right tool, for the right job.
Cellular should not be considered ‘FWA’. WISP FWA and Cellular services are dramatically different structurally.
This discussion needs to consider time scale. Over the short term, FWA is certainly an option. But even as the equipment advances, radio spectrum is a limited resource and thus offers constrained capacity to accommodate future bandwidth demand needed for the long term. Current subsidization policies that favor FTTP are therefore the correct use of public funds for advanced telecom infrastructure to provide the best long term value for taxpayers.
Also, the United States adopted telecom policy that brought copper voice telephone infrastructure to nearly every American doorstep. There’s no reason why the same can’t be done for fiber.
time scale does matter. Fiber is multiple orders of magnitude more expensive (in inflation adjusted numbers) today than running copper was half a century ago. I’m not saying fiber can’t be done, but it’s really not comparable based on cost.
Fiber is slow to build, it’s expensive to build, and it creates monopolies in local environments based on the US governments ‘tendancy’ to fund a single provider in an area. Fiber should certainly be built as part of core infrastructure and over time built to more and more places, but this should really be on a demand basis. I have some ideas on the models to do this and promote competition but that’s outside the scope of this post. There’s a whole lot of government funded fiber runs that simply aren’t in use, often because they’re not accessible. Fiber checks most of the boxes for ‘superior technology’ but not all of them and political battles are a big weakness in fiber. Most underserved areas in the US have fiber run right through them or very nearby that they can’t access. basically the term ‘fiber’ is more than a tech, it’s also bogged down in politics and that can’t be ignored.
Meanwhile modern mmwave wireless technologies have very near unlimited spectrum. They are in their infancies, with 16QAM modulation while WiFi6 is 1024QAM and wiFi7 is 4096QAM capable, ie lots of room to improve there, and with mmwave’s characteristics it can be re-used in 1Gbps rates for about 1:25 the cost of fiber. We have another tech that can do 2Gbps in a larger range (or smaller, there are choices) and is about 1:50 the cost of fiber and also offers up >1Gbps rates.
Building internet service for the people in the future is great, but the need is today. modern wireless networks can deliver effectively instantly, they can be upgraded with primary fiber runs as demand grows, and once the demand to get 10, 40, or 100G into an area is there.. well the fiber has been pulled to the neighborhood so the fttx build is a fraction of the cost.
So I argue that a direct fttp build for much of the US is actually a really poor investment. It’s spending money for future use. Building (or otherwise making access to) fiber to the town/city and then a competative marketplace for any tech and provider that gets high quality service as soon as possible is a far better investment. the ‘ROI’ on this is extremely fast because it’s actually available to almost everyone almost instantly AND having brought that fiber into the area for people that are paying for it right now means it’s economically viable and capitalist mechanisms start working right away.
We’ve build mmwave sites in just a couple of hours and hooked up the first customers same-day on 300M-1Gbps services.
We are in agreement FWA provides a good short term option. As for millimeter wave tech, can carry a lot of bandwidth but tradeoff is weak propagation. Then bringing fiber to the prem makes better sense.
As for fiber “creates monopolies” as I mentioned, FTTP like other utility infrastructure functions as a natural terminating monopoly. Same reason why we don’t see multiple electrical and water lines to a home. Competition makes little sense when costs of entry and first mover advantage taken into account.
“As for fiber “creates monopolies” as I mentioned, FTTP like other utility infrastructure functions as a natural terminating monopoly. Same reason why we don’t see multiple electrical and water lines to a home. Competition makes little sense when costs of entry and first mover advantage taken into account.”
Take one look at the massive problems with electrical and water services in the US. Makes an incredibly strong case that this is a bad model.
One has to look at their structures. Many are investor owned. That model is inherently conflicted between the interests of shareholders and end users. The same conflict we see with investor owned advanced telecom infrastructure with people wanting affordable access that’s averse to the interests of shareholders who want to minimize capex and maximize profit, share price and shareholder dividends.
right, so all the bad parts of monopolies are showing through. I think you’re doing most of the work to make my point here.
The government should build highways and make onramps through contractors and then let it be. Trade highway and onramp to suite the subject matter. This is where government selection works well but at any smaller level this breaks down rapidly. When the government picks the local monopoly it’s a downhill slide from the day of completion on. Imagine hiring a single construction company to work on all the roads in town with a 20+ year bulletproof contract. That’s the pitch for this first-wins model for fttx. The fact that there’s fiber money being offer to overbuild coax services is proof that concept fits ISPs, those coax companies got that government selection, first to the pole wins advantage and here we are.