India Has 969 Million Internet Users and Almost No Broadband

This week TRAI put out a number that sounds like good news and is actually the whole problem. India now has more than 969 million internet subscribers. We are one of the most connected countries on earth. And yet fixed broadband, the wire or fiber that runs into your home, sits at around 63 million connections. That's roughly 20% of households, against 80 to 90% in developed countries.

I run a fiber ISP in Kashmir, so I don't read that gap as a statistic. I read it as the thousand conversations I've had at people's doors. Almost everyone already has internet. Almost nobody thinks they need broadband. Those two facts are the entire business I'm in, and they explain why this number has barely moved.

Internet is not the same thing as broadband

When TRAI says 969 million, it mostly means phones. Cheap data, a SIM in every pocket, reels playing on the bus. That revolution already happened, and Jio mostly won it. What it did not do is put a reliable connection into the house, the kind a family of five shares for work, school, calls, and a TV all at once.

So when I knock on a door and say "fiber to your home," the honest reaction I get is: I already have internet. Why would I pay again for the same thing? On the phone, it looks like the same thing. The job of an ISP in India in 2026 isn't to connect the unconnected. It's to convince the already-connected that there's a different, better category they haven't tried.

The math that keeps penetration at 20%

Here's the trap. A mobile data plan in India can cost a couple of hundred rupees a month. A home fiber plan costs more, plus an install, plus a router sitting on the wall as a visible monthly bill. For a household that already feels connected, that's a hard sell on price alone.

And the upgrade only makes sense once the home actually demands it: someone working remote, kids in online classes, a small business run off the kitchen table, four people streaming at 9 PM. In a lot of Indian homes that demand is still partial. The phone is "good enough" for the way the house currently uses the internet, so the fixed line feels like a luxury, not a necessity. Penetration stays at 20% not because the wire isn't available, but because the need hasn't crossed the line where people will pay for it.

That's the real competitor a local ISP fights. It isn't Starlink, it isn't Jio AirFiber, it isn't the guy down the road. It's the mobile plan the customer already pays for and the belief that it's enough.

What the "hybrid model" actually admits

At DigiCom this month the TRAI chairman flagged the same low penetration and pushed what he called a hybrid deployment model: fiber where it's dense and worth it, fixed wireless where laying cable is too slow or expensive, satellite for the far edges. The government also decided satcom spectrum will be allocated rather than auctioned, clearing a path for Starlink, Jio-SES, and OneWeb.

I think the hybrid framing is correct, and I've argued the same thing from the ground in earlier posts on Starlink and 5G AirFiber. No single technology connects India. Fiber is the best pipe where the density justifies the digging, and most of the country's density doesn't. But notice what the hybrid model quietly admits: the hard part was never the technology. We can reach almost any home three different ways now. The hard part is that reaching a home and getting that home to subscribe are two completely different problems, and only the first one is an engineering problem.

What actually moves a household

On the ground, the thing that converts a "no" into a "yes" is almost never a spec sheet. Nobody at the door cares about symmetric speeds or jitter. What moves them is a concrete reason the phone failed them: the video call that froze during an interview, the kid who couldn't submit an exam, the shop's payment machine that kept dropping. The need has to become visible before the price stops looking like a luxury.

So the real work of growing fixed broadband in a place like Kashmir is unglamorous. It's showing up, being the ISP whose line actually holds at 9 PM, fixing a fault the same day so the household never goes back to "the phone was fine anyway," and letting one convinced neighbour sell the next three. The 20% number won't move because a faster technology arrived. It moves one household at a time, when the internet they already have stops being enough and someone reliable is standing there when it does.

The opportunity hiding in a bad number

Here's why I'm not discouraged by 20%. In most markets, a number that low means the product failed. In India it means the product has barely been offered to four out of five homes in a way that landed. The demand is being manufactured in real time, by remote work, online school, UPI, and streaming, all of which quietly raise the bar on what "enough internet" means. Every month, more households cross the line where the phone stops being enough.

That's the whole bet of a local ISP. Not that we'll out-engineer Jio or out-fund Starlink, but that when a household finally needs real broadband, we'll be the name they already trust on their street. The gap between 20% and 80% isn't a problem to lament. It's the next ten years of work, sitting there in plain sight.

If you're building an ISP, a local infrastructure business, or anything that depends on convincing people to upgrade from "good enough," I'd like to compare notes: me@mehranshahmiri.com