Starlink is now licensed in India. Hardware kit: roughly Rs 34,000. Monthly plan: Rs 8,600. Speeds between 25 and 225 Mbps, from a dish on your roof, anywhere in the country. The question I get most often these days is some version of: doesn't that kill what you're building?
I run ViberNet, a fiber ISP laying last-mile infrastructure across Kashmir. I've spent the last year getting permissions, pulling cable, installing OLTs, and onboarding customers street by street. Here's what I actually think.
What Starlink actually is
Starlink is low Earth orbit satellite internet. Unlike the older geostationary satellites that sat 35,000 km above Earth and gave you 600ms of latency and 15 Mbps on a good day, Starlink's constellation orbits at around 550 km. The result: latency closer to 20–40ms and real download speeds. It works. I have no interest in pretending otherwise.
But "works" doesn't mean "same as fiber." LEO satellite internet is shared bandwidth. When more people in your area subscribe, speeds per user drop. It's weather-affected — not catastrophically, but a heavy snowfall in the Kashmir valley can degrade service. The hardware is a significant upfront cost. And at Rs 8,600 a month, it's priced for a customer with a specific profile: someone who needs connectivity badly enough to pay for it and either can't get fiber or doesn't have access to a better option.
That customer exists. There are a lot of them. Starlink is solving a real problem for them. That customer is just not primarily the customer I'm building for.
Who Starlink is actually competing with
Starlink's real competition in the Kashmir valley is not ViberNet. It's DSL. It's the 2–5 Mbps copper-line connections that most households here have been tolerating for years. It's mobile data that drops when the tower is congested. Those are the products Starlink is meaningfully better than, and there are a lot of them to displace.
A dedicated fiber connection from ViberNet — individual wavelength, symmetric speeds, not shared bandwidth — is a different product at a different price point. When I put the two side by side for a residential customer in Srinagar, the math isn't complicated: local fiber is cheaper per month, has lower latency, has no weather dependency, and when something breaks, a local engineer can be there the same day rather than the customer troubleshooting with a support line that doesn't know what Srinagar looks like.
The honest competitive threat Starlink poses to ViberNet is not in the dense urban areas we're already wiring. It's in the places we haven't reached yet — remote villages, high-altitude areas, communities near the Line of Control where laying fiber is either technically difficult or logistically not viable within our current operating radius. For those areas, Starlink may solve the problem before we can.
I find that useful, not threatening.
Where Starlink wins, no argument
There are deployment scenarios where Starlink is simply the right tool and no local ISP is a real substitute. Mobile setups that need connectivity at a new location every week. Disaster response, where existing infrastructure has failed. Temporary construction sites. Remote agricultural operations. Emergency backup links for businesses whose primary connection is critical.
I'd actually recommend Starlink for some of these cases. Not because I'm being generous — because it's accurate. There's no version of ViberNet that makes sense as a solution for a camp at 4,000 meters. The physics of laying fiber doesn't work there. If you need reliable internet at that altitude, Starlink is the answer and I'm not the alternative.
A technology that does something well that you can't do is not a threat. It's a complement. The categories don't fully overlap.
What this changes for how I build
Starlink's arrival in India sharpens one thing clearly: there is no longer a version of local ISP work that survives on scarcity alone. If you were a regional ISP coasting on being the only option, that's over. Customers now have a credible alternative they can order online and deploy themselves.
The ISPs that survive this are the ones that are genuinely better on the metrics that matter for their specific market — price per megabit, latency for real-time applications, reliability during winter, support response time when something goes wrong.
For ViberNet, that means the work I was already doing matters more, not less. Dedicated fiber beats shared satellite on price and latency for dense urban customers. Local support beats a remote helpdesk for customers in the valley. The competitive advantage was never "we're the only option." It was always "we're the better option for this market." Starlink just made it more urgent to be right about that.
There's also something Starlink can't do: employ local engineers, create a local support operation, build the institutional knowledge of what infrastructure actually looks like in each neighborhood. That kind of local depth is an asset that compounds. A satellite company run from the US has the technology and the capital. It doesn't have the street-level knowledge that determines whether an installation works for this building in this part of the city in this season.
The honest answer to the question
Starlink doesn't kill local fiber ISPs that are actually doing the work. It kills local ISPs that existed because customers had no choice. Those are different businesses.
The version of this industry that will survive the next five years is the version that earns its customers rather than inheriting them by default. That means better infrastructure, real support, honest pricing, and a genuine understanding of the specific market you're serving. A satellite constellation 550 km above Earth cannot substitute for that. It can compete with the parts of the local ISP business that weren't doing those things well.
I'm building for a market that has been underserved by infrastructure for a long time. Starlink is also building for it, from a different angle, with different tools. That's probably fine. The demand is large enough that multiple solutions can exist. And competition from a credible product is a better forcing function than I could design myself — it makes you better, or it removes you. I'd rather be better.
If you're thinking through connectivity options in Kashmir — for a home, a business, or a remote operation — I'm happy to give you an honest comparison based on what we're seeing on the ground. me@mehranshahmiri.com
Related: The Internet Problem in Kashmir · Why I'm Not Raising Money