Do you think starting an ISP is still a possible venture? I've been looking into it for the last few months but everyone around me tells me it isn't possible/worth while. How much capital did you have up front?
It's going to be extremely location-dependent. Your best bet is to fill a niche that nobody else has gone after because it doesn't fit their national standards.
Remote location, or otherwise protected against a national moving in?
No current competitors, or only low-quality competition?
But with enough latent demand?
If all three of those are yes, you might be able to make it work.
You'll need to think about fiber or wireless, and if you're doing wireless, how to get fiber to your distribution points. Licenses: right-of-way for fiber, radio operation for wireless.
IPv4 space? Good luck.
IPv6 space? Much easier, but people still won't find it compelling all by itself. (On the other hand, starting an ISP today means you can probably do all your internal work on IPv6. Do so!
Can you get two independent high-bandwidth connections? If not, can your customers stand the inevitable downtime?
What level of reliability are you aiming for, anyway?
What services are you going to provide besides bandwidth, IP addressing, routing, NTP and DNS?
Are you prepared for customer service? Business or residential? Either way, you will eventually have to listen to someone on the phone telling you that they deleted the internet.
I have a friend that runs as a WISP on the side in Michigan as there are no other options. He only has a /26 given to him by a single provider, so the 5-10 customers he has are behind a NAT.
