Several people have asked me where the cellular carriers are finding the millions of customers they are adding to FWA cellular broadband. T-Mobile added 550,000 customers in the third quarter to reach 4.6 million total subscribers. Verizon added 380,000 in the third quarter to reach 2.7 million customers. AT&T is late to the game but says it is now adding 2,000 a week and plans to step this up significantly. Other carriers like UScellular are poised to enter the market.

The new FWA customers have to be coming from somewhere – and there are a lot of possibilities. In urban areas, the customers would have to be coming from cable companies, telco DSL, or fiber. In rural areas, customers would be coming from DSL, fixed wireless, satellite, or cellular hot spots. I’ve not seen any discussion or announcements from T-Mobile or Verizon about where they are gaining customers.

T-Mobile and Verizon together now have almost 6% of all broadband customers using the new wireless product. I’ve seen those companies speculating about growing to a 15% market share, which would mean adding another 10 million customers. That would make FWA collectively the third largest ISP behind Comcast and Charter.

I’ve been doing some digging, but my research is far from scientific. I’ve looked in detail at half a dozen counties using detailed Ookla speed tests to see where the FWA customers live. I looked at a few rural counties with only one or two sizable towns, and I looked at two suburban counties where cable companies serve the large majority of the geography. The first thing I noticed is that speeds are only good for around two miles from a cell tower – that’s not far in a rural area, but the carriers are now passing a lot of customers.

I also noticed in this tiny sample is that there are a lot more FWA speed tests in rural areas than in towns. That makes sense. In many places the FWA product delivers speeds of between 100-300 Mbps download – and for rural customers who have had no decent broadband option, this is a spectacular upgrade option. Upgrading to FWA from a creaky 10/1 Mbps DSL probably feels like the change we remember from going from dialup to DSL. I’ve noticed that there are still a lot of rural towers that have not yet been upgraded to the product, and perhaps those upgrades will provide a lot of the path to adding another 10 million customers.

There are FWA customers in cities and dense suburbs, but the relative penetration of FWA in densely populated areas looks to be relatively miniscule. The FWA sales proposition in urban areas is price. It’s still not unusual to find a 15% market penetration of DSL in cities, and it should be easy to sell faster speeds and better reliability at the same price as DSL. CenturyLink recently said that its rapid drop in DSL is due mostly to FWA.

One place that the cellular carriers are getting the customers is from their own embedded customer base. In many rural counties, T-Mobile and Verizon already have a 2 to 4% market penetration of cellular hotspot customers. These older broadband products delivered whatever cellular speed was available on the rural 4G network – and were exceedingly expensive due to severe data caps. It’s a pretty easy sell for a cellular company to tell customers they are sending them a new receiver that will get faster speeds and that will provide unlimited broadband – but with the small print caveat that bandwidth can be throttled any time that cell phone traffic is heavy.

I know that AT&T is using the product to convert its rural DSL customers. I happened to be in an AT&T store recently and heard them tell a customer that their copper DSL and voice line was soon going to be cut dead and replaced with FWA. The sales pitch to compete against satellite is also pretty easy. The FWA product is faster and costs less than any of the satellite options. Competing against WISPs using fixed wireless is more of a local situation since the quality of WISP broadband ranges from miserable to great.

The big question everybody wants answered is the impact of FWA on cable companies. In the last quarter, the big cable companies collectively added only 8,000 customers, and only Charter grew a little. The big cable companies have made themselves vulnerable due to high prices and in many markets due to reliability. Converting to FWA can almost cut a broadband bill in half. In cities, almost everybody lives within a decent distance of a cell tower – but the big concern for urban FWA is the capacity to support the product. While the cellular carriers love the new FWA product, they still view this as a footnote product in their annual reports. They are not about to jeopardize cellphone quality by oversubscribing fixed broadband customers on networks that weren’t designed to provide continuous broadband connections.

I still have no idea of the extent to which cable companies are losing customers to FWA or fiber. They are certainly losing to both – but the cable companies and their competitors are staying mum on the topic.