9 ms·

We have an ISP in San Francisco and it really is a great business. Lots of long lasting recurring revenue and very low recurring costs--bandwidth is actually on

We have an ISP in San Francisco and it really is a great business. Lots of long lasting recurring revenue and very low recurring costs--bandwidth is actually one of our smallest expenses. The biggest hurdle is really tapping in to the tribal knowledge on how to set up and run an ISP. You've got networking, provisioning, how to buy bandwidth, even what type of power plugs you need to have in your rack (spoiler: there are multiple and they are not all compatible). Besides the cost of initial deployment and setting up new customers, there's not a whole bunch you need to spend money on. We really like the business, and more people really should start their own. The numbers make sense in a bunch of different markets and scales. If you deploy with fixed wireless, it brings the costs down to the 10's of thousands of dollars range. Laying fiber is obviously a lot more expensive, but not really necessary in most circumstances. We're working on helping more people start ISPs-launching here on HN next week, but TechCrunch wrote about the project just now: https://techcrunch.com/2018/05/10/necto-looks-to-help-individuals-get-their-own-local-isp-businesses-off-the-ground/ https://techcrunch.com/2018/05/10/necto-looks-to-help-indivi...

belltyler 8y ago
Potentially relevant, I remember seeing this a while back: https://startyourownisp.com/ https://startyourownisp.com/ HN Discussion: https://news.ycombinator.com/item?id=16160394 https://news.ycombinator.com/item?id=16160394
- montasaurus 8y ago
  This is a great guide--I highly recommend reading it if you're interested in the space. We met Graham (the author) and he really knows his stuff.
- grahamburger 8y ago
  Oh Hi! Author of https://startyourownisp.com/ https://startyourownisp.com/ here. Happy to answer questions.
  - rrcaptain 8y ago
    You mentioned problems scaling in the article. Can you elaborate on that? Will you update that later on?
    - grahamburger 8y ago
      Yeah, good question. I don't think it's anything more cryptic than what it takes to scale any business, I was just trying to point out that I made a deliberate decision to focus on what's needed to get from 0 customers to 'some' customers instead of what's needed to get from 'some' customers to many customers. I do hope to eventually go in to more depth about what it takes to scale larger. Briefly, here are some of the challenges WISPs run in to as they scale: -Getting contracts with relay sites / towers / places to put their wireless infrastructure. This is one of the most time consuming parts of building a WISP, and it's primarily just about waiting and red tape and making lots of phone calls. -Geographic limitations - IE we have our little valley covered and we've got as many customers here as we're gonna get, now what? It would cost $XXX to go over to the next town/city/county and we don't really have that or don't have the resources to feel like we could handle that logistically, so we're just gonna stay where we are. -Network upgrades. IE 10 years ago a 5Mbps connection was pretty good! So we built a network that would deliver that. Now customers want 50Mbps, and the technology is there, but we don't have the capital or the know-how to go do a full network and technology upgrade.
      - firmgently 8y ago
        For a moment there I almost thought I was reading about a business that was content making some money and employing some people, rather than presuming that mindless, infinite growth is the only possible path. I was probably imagining it or projecting my perverted ideals but what a beautiful, refreshing moment it was :)
        grahamburger 8y ago
        I get what you're doing and at the risk of taking the bait I want to point out that this is actually the way that a lot of WISPs that I know of operate. Build a network up to a certain number of customers that can be supported by 1-5 people and just leave it there and run it as a lifestyle business, or eventually sell it. I personally know maybe 10 or 12 people who are now doing or have done that in the past. There are of course also businesses in this market that want to continue growing and I don't think there's anything wrong with that.
Zarath 8y ago
One thing I've never understood is how computers using one ISP connect with computers using another ISP. Is there some sort of law that requires that one ISP's cable must be able to interact with another ISP's cable? What incentive does someone, like say Comcast, have in allowing their users to interact with people using a smaller ISP like yours?
- andrei_says_ 8y ago
  Is this a trolling attempt? If not, ISPs do not connect to “each other”. They “connect you to the internet.” Which all the ISPs connect to.
  - wyldfire 8y ago
    Is this the ol' double troll? :) In fact, ISPs do connect to one another via peering [1]. [1] https://en.wikipedia.org/wiki/Peering https://en.wikipedia.org/wiki/Peering
    - andrei_says_ 8y ago
      I apologize, this is my ignorance. Thank you for the link, I know more now.
  - thescriptkiddie 8y ago
    Are you trolling? The internet isn't some kind of giant server in the sky. It's made up entirely of ISPs who connect to each other and route each other's traffic for mutual benefit. ISPs are the internet. One of the hardest things about starting your own ISP is convincing the big guys to connect with you.
    - naveen99 8y ago
      Lower tier isp’s usually pay the big guys for peering.
      - 8y ago
jtchang 8y ago
I love to learn more about what you guys do! Is there a good way to get in touch with you guys or help out?
- montasaurus 8y ago
  Definitely! I'll shoot you an email.
icedchai 8y ago
I helped start a few ISPs in the late 90's, early 2000's... Is it really a good business these days? I figured it's low margin and hard to compete with the Comcasts and Verizons of the world. There's one fixed wireless ISP I know of regionally that is a public company, and they're bleeding money.
- MrLeap 8y ago
  It's gotta be. My brother and I rent commercial space as a workshop on a main stretch of road in a pretty big midwest city. I asked a few nearby businesses who they use for wifi, they all said AT&T. So I go onto AT&T's website and the cheapest option they had was 2mb service for 44$ a month AND a year commitment. I checked 4 or 5 more ISPs and they all had the same or worse deals. I ended up buying a hotspot from T-Mobile. Now that I'm reading about this though.. If I can create a wireless ISP that can do better than 20$ a MB for just a mile or two of the business boulevard I'm on and it costs about ~10k I'm going to do it. Time to research I guess.
  - montasaurus 8y ago
    We service some "DSL dead zones" like that...it's crazy to see the prices people are forced into paying. We had one customer that was paying over $100/mo for low single digit Mbps before they switched.
- montasaurus 8y ago
  Thankfully it's gotten a lot better since then. The RF tech has improved dramatically and bulk bandwidth costs are much more reasonable. Is it Proxim?
  - icedchai 8y ago
    I'm right outside of Boston. Towerstream is the ISP I was speaking of... they always seem to be losing money.
- TheSmiddy 8y ago
  I've worked at a few ISPs in Australia. There are still huge efficiency gains to be made, especially on the customer service side of the business which accounts for 70+% of the labour cost.

belltyler 8y ago

Potentially relevant, I remember seeing this a while back: https://startyourownisp.com/ https://startyourownisp.com/ HN Discussion: https://news.ycombinator.com/item?id=16160394 https://news.ycombinator.com/item?id=16160394

montasaurus 8y ago
This is a great guide--I highly recommend reading it if you're interested in the space. We met Graham (the author) and he really knows his stuff.
grahamburger 8y ago
Oh Hi! Author of https://startyourownisp.com/ https://startyourownisp.com/ here. Happy to answer questions.
- rrcaptain 8y ago
  You mentioned problems scaling in the article. Can you elaborate on that? Will you update that later on?
  - grahamburger 8y ago
    Yeah, good question. I don't think it's anything more cryptic than what it takes to scale any business, I was just trying to point out that I made a deliberate decision to focus on what's needed to get from 0 customers to 'some' customers instead of what's needed to get from 'some' customers to many customers. I do hope to eventually go in to more depth about what it takes to scale larger. Briefly, here are some of the challenges WISPs run in to as they scale: -Getting contracts with relay sites / towers / places to put their wireless infrastructure. This is one of the most time consuming parts of building a WISP, and it's primarily just about waiting and red tape and making lots of phone calls. -Geographic limitations - IE we have our little valley covered and we've got as many customers here as we're gonna get, now what? It would cost $XXX to go over to the next town/city/county and we don't really have that or don't have the resources to feel like we could handle that logistically, so we're just gonna stay where we are. -Network upgrades. IE 10 years ago a 5Mbps connection was pretty good! So we built a network that would deliver that. Now customers want 50Mbps, and the technology is there, but we don't have the capital or the know-how to go do a full network and technology upgrade.
    - firmgently 8y ago
      For a moment there I almost thought I was reading about a business that was content making some money and employing some people, rather than presuming that mindless, infinite growth is the only possible path. I was probably imagining it or projecting my perverted ideals but what a beautiful, refreshing moment it was :)
      - grahamburger 8y ago
        I get what you're doing and at the risk of taking the bait I want to point out that this is actually the way that a lot of WISPs that I know of operate. Build a network up to a certain number of customers that can be supported by 1-5 people and just leave it there and run it as a lifestyle business, or eventually sell it. I personally know maybe 10 or 12 people who are now doing or have done that in the past. There are of course also businesses in this market that want to continue growing and I don't think there's anything wrong with that.

Zarath 8y ago

One thing I've never understood is how computers using one ISP connect with computers using another ISP. Is there some sort of law that requires that one ISP's cable must be able to interact with another ISP's cable? What incentive does someone, like say Comcast, have in allowing their users to interact with people using a smaller ISP like yours?

andrei_says_ 8y ago
Is this a trolling attempt? If not, ISPs do not connect to “each other”. They “connect you to the internet.” Which all the ISPs connect to.
- wyldfire 8y ago
  Is this the ol' double troll? :) In fact, ISPs do connect to one another via peering [1]. [1] https://en.wikipedia.org/wiki/Peering https://en.wikipedia.org/wiki/Peering
  - andrei_says_ 8y ago
    I apologize, this is my ignorance. Thank you for the link, I know more now.
- thescriptkiddie 8y ago
  Are you trolling? The internet isn't some kind of giant server in the sky. It's made up entirely of ISPs who connect to each other and route each other's traffic for mutual benefit. ISPs are the internet. One of the hardest things about starting your own ISP is convincing the big guys to connect with you.
  - naveen99 8y ago
    Lower tier isp’s usually pay the big guys for peering.
    - 8y ago

jtchang 8y ago

I love to learn more about what you guys do! Is there a good way to get in touch with you guys or help out?

montasaurus 8y ago
Definitely! I'll shoot you an email.

icedchai 8y ago

I helped start a few ISPs in the late 90's, early 2000's... Is it really a good business these days? I figured it's low margin and hard to compete with the Comcasts and Verizons of the world. There's one fixed wireless ISP I know of regionally that is a public company, and they're bleeding money.

MrLeap 8y ago
It's gotta be. My brother and I rent commercial space as a workshop on a main stretch of road in a pretty big midwest city. I asked a few nearby businesses who they use for wifi, they all said AT&T. So I go onto AT&T's website and the cheapest option they had was 2mb service for 44$ a month AND a year commitment. I checked 4 or 5 more ISPs and they all had the same or worse deals. I ended up buying a hotspot from T-Mobile. Now that I'm reading about this though.. If I can create a wireless ISP that can do better than 20$ a MB for just a mile or two of the business boulevard I'm on and it costs about ~10k I'm going to do it. Time to research I guess.
- montasaurus 8y ago
  We service some "DSL dead zones" like that...it's crazy to see the prices people are forced into paying. We had one customer that was paying over $100/mo for low single digit Mbps before they switched.
montasaurus 8y ago
Thankfully it's gotten a lot better since then. The RF tech has improved dramatically and bulk bandwidth costs are much more reasonable. Is it Proxim?
- icedchai 8y ago
  I'm right outside of Boston. Towerstream is the ISP I was speaking of... they always seem to be losing money.
TheSmiddy 8y ago
I've worked at a few ISPs in Australia. There are still huge efficiency gains to be made, especially on the customer service side of the business which accounts for 70+% of the labour cost.