This week I read a number that should probably scare me: roughly three out of every four new home broadband connections Jio added this year came over the air, not over fiber. Their 5G fixed wireless product, AirFiber, crossed 8.5 million users. Airtel's version is past 3.7 million. In March alone, Jio added almost a quarter of a million homes — without digging a single trench.
I run a company whose entire premise is digging trenches. So I've spent the week thinking about whether AirFiber is the thing that makes ViberNet obsolete, and I want to write down where I landed.
What AirFiber actually is
Despite the name, there is no fiber in AirFiber — at least not to your house. It's 5G fixed wireless access: a tower near you has fiber running to it, and your home gets a receiver that pulls a broadband connection off that tower over the air. The fiber stops at the tower. The last kilometre — the part that costs companies like mine the most money and time — is replaced by a radio signal.
That's the whole trick, and it's a good one. The last mile is the expensive mile. Every ISP founder knows this. Permissions, trenching, poles, cable, the guy whose shop front you have to dig past, the road department that resurfaces over your duct three weeks after you laid it. When Jio skips all of that, they can connect a home in a day that would take me weeks.
So the growth makes complete sense. If I could light up a whole mohalla from one tower instead of pulling cable down every lane, I'd be tempted too.
Why it grows where fiber doesn't
Look at where FWA is actually growing fastest and the pattern is clear: it's strongest in places with limited fiber. That's not an accident — it's the strategy. The telcos are using wireless to reach homes that no one has wired up yet, because reaching them by air costs a fraction of reaching them by cable.
I recognize those places, because I work in them. Half of Kashmir is "areas with limited fiber." The economics that make a village hard for me to justify trenching into are exactly the economics that make it attractive for a tower-based product. In those places, AirFiber isn't competing with ViberNet. It's competing with nothing — with the 4G hotspot someone props in a window.
And honestly: a family getting 100 Mbps over the air in a village I couldn't reach for two more years is a good outcome. I'd rather lose a customer I never had than watch them stay disconnected.
The physics still favours the cable
Here's the part the plans and the ads don't dwell on. A tower is a shared resource. Every home on it is splitting the same spectrum, and spectrum doesn't scale the way glass does. When I need more capacity on a fiber line, I change the equipment at the ends. When a tower sector fills up, every customer on it feels the evening slowdown at the same time.
This is the oldest story in broadband: wireless wins on deployment speed, wired wins on physics. A strand of fiber into your home is yours. The speed is symmetrical, the latency is flat and boring, and it behaves the same at 9 PM as it does at 9 AM. The 8 PM test matters more than the speed test — and 8 PM is when a loaded tower shows you what you actually bought.
Weather matters here too, in ways that are easy to forget if you're writing strategy in Mumbai. Kashmiri winters are hard on radio links and harder on power. A buried fiber route doesn't care about snowfall. Our biggest winter problem is electricity, not signal — and that problem hits a wireless receiver in your window exactly the same way.
What this means for ViberNet
Let me be straight about the threat level, because I wrote about Starlink earlier this week and I want to rank these honestly. Starlink, at Rs 8,600 a month, is not coming for my urban customers. AirFiber, at a few hundred rupees' difference from my plans, absolutely is. Of everything that's happened in Indian broadband this year, 5G FWA is the most direct competition a local fiber ISP faces. It would be dishonest to pretend otherwise.
But the same three things keep showing up when I look at what we actually have. First, in the dense areas where we operate, fiber delivers a better product at the hour people use it most — and customers figure that out within a month of switching. Second, when something breaks, my technician is a local person who shows up the same day, not a national call centre reading a script. In a market built on trust, that's not a soft advantage; it's the reason a third of our new connections come from referrals.
Third — and this is the long game — every tower that makes AirFiber work needs fiber running to it. The more wireless grows, the more glass the country needs in the ground. The asset I'm building doesn't get less valuable in a wireless world. It becomes the thing the wireless world stands on.
Three competitors, one lesson
This is the third post I've written in a week about something coming for the last mile: Starlink from orbit, the government with subsidized backbone, and now the telcos from their towers. Reading them back, the lesson is the same each time. Everyone with capital is trying to solve the last mile from far away — from space, from a ministry, from a tower. The last mile keeps being won up close.
My plan doesn't change: keep trenching the dense areas where fiber's advantages are felt daily, let wireless serve the places I can't reach yet, and treat every customer interaction as the moat. If AirFiber forces every small ISP in India to be better at service and honest about speed, good. The ones who were coasting on being the only option deserve to lose those customers. I intend not to be one of them.
If you're building infrastructure in a market the big players just noticed, I'd genuinely like to compare notes: me@mehranshahmiri.com
Related: I Run a Fiber ISP. Here's What I Actually Think About Starlink. · The Government Is Laying Fiber Too. Here's What That Means for ViberNet.