On Starting a Successful ISP?

Tigris666 asks: "I would like some handy hints from all /.'ers on what is needed in order to start up an ISP. I'm asking you guys in hope that someone out there has started one before, and i think that's a fair assumption. There are obvious things like mail/web servers, dial-in modems/systems, tech-support employees. But what i'm looking for is an idea on costs (we're talking Australia here), hardware required, and basically an idea on how hard it might be? Things like setting up the internet link, organising the phone lines to be put in, how many servers are needed, and the big thing! Is it all worth it? Does it pay off in the long run?" With larger fish finally jumping into the waters of Internet connectivity, is there still room for smaller companies? I would think that any new ISP would not be able to survive by solely providing dial-up service, and would need to look into the possibility of providing DSL or cable connectivity. However, providing broadband connectivity is a significant and expensive venture, made even more difficult considering the current economic conditions. What suggestions do you have for anyone who thinks themselves up to it?

"Basically the idea came up because the area I am from is in the country. There is only 1 service provider out there, and they are really bad with disconenctions, among other things, and everyone i know absolutely hates them. I think starting an ISP would be a good oppurtunity. I have recently moved to the city to get a real job, however I much prefer living in the country, so this will certainly be a big step."

Comments

Here are complete instructions

[emil]

I published this a few years ago for Unixworld. You will want to use the PAM (RH5) configuration.

A single Pentium 150 handled 32 lines with no problem.

http://rhadmin.org/uw/015.html

The best advice: don't do it.

[defile]

Starting an ISP in the USA nowadays is most likely a mistake. Here's why:

• The dialup market is almost effectively gone. Most ISPs now use their dialups to either add value to their existing services or because they service a tiny area with extremely dedicated customers. We service the NYC area and get only 1 or 2 dialup customers a month now (and lose 5 or 6), when they used to come in at about 10/day and we couldn't even meet the demand.

• Dialups are practically sold as a commodity by big corporations who just want to sweep it up. You'll not only be competing with AOL, but AT&T, MSN, IBM, Earthlink, but also a bunch of ISPs that will give it away for free (like Altavista, of all people).

• If you want to compete using other technologies, such as DSL, you'll have to deal with the phone company. Your main business will be selling the same technology that the phone company sells, but at a higher rate [because the telco prices it that way, since you're a competitor]. And when the telco downs your service for whatever reason, customers call you, not them. Your core business will center around escalating trouble tickets on behalf of your clients for DSL. There's a real reason as to why Northpoint, Red, and Covad have folded or are in deep shit. If you want to make money at this, you'd better have a lot that you plan to put down up front and won't mind throwing away based on the telco's whim. You will also compete directly with cable, which will almost always be less error-prone

• The market is getting -more- saturated, not less. This is contrary to how most people thought it would end up (3 or 4 big ISPs dominating the continental US)

• You will also deal with commodity web hosting providers who will host sites for practically nothing. You cannot effectively compete with them unless you either offer fantastic tech support or plan to provide custom development.

My advice for people starting new ISPs is that you shouldn't. It's a terribly bad idea. We're a small, established, conservatively run ISP (in terms of how we spend) and it's very hard to survive. We will survive, and we're finally out of the red, but it took us 5 years to get here, and the market back when we started is nowhere near as hostile as it is now.

If you're still interested: Forget dialups unless you really expect to target a strong niche market that AOL and the rest miss. I'd concentrate more on custom development for web sites (where you also host their sites), or finding some possibly untapped technology (like satellite internet, which could work very well in some areas).

If you live in a particularly metropolitan area, you could also concentrate on dropping T1's to huge office buildings and running ethernet to each client from there. That is a much better deal than the cable/DSL that they're probably stuck with. There's more sales involved than anything, though, with this approach. (Word of mouth doesn't seem to work as well for this since corporations don't generally get along as great friends within office buildings, least in my experience)

Good luck.

Re:Best Advice

[matth]

A friend of mine and I have started an ISP for very few start-up costs. We accomplished this by renting lines from UU.NET and then we provide the mail/radius, etc servers. They are in a data center. So it's been very inexpensive for us, and the dial-up line quality seems to be very good frmo UU.NET

Swift-Networks - Nation Wide ISP!

Australia? Who knows...in the U.S., now...

[RavenDarkholme]

Not knowing anything about Australia and telco costs and so forth, I can't really comment about that. (So why are you posting??)

In the U.S., though, I've worked for an ISP who goes into a lot of little towns where there is no (or little) ISP service. They usually start out with 12 phone lines, which in US dollars is about $300 a month, and a frame-relay 56K connection back to the main ISP which is another couple hundred dollars a month, an Ascend Max 4000 or Portmaster III which you can get on eBay for not too much.

Webserver/mail server/DNS server? Heck, get a couple of lower end Celerons with, say, 128 megs of ram, a couple of 20-gig hard drives, throw Linux with Apache, Sendmail (or Qmail), Radius of some sort (I rather like FreeRadius) and BIND on them, and Ka-boom: instant servers.

Generally, what they do is say to the town: something like, you guarantee us X number of users, and we will bring Internet service to this town. Many times, the people will sign up (and pay!) for service before the ISP even gets out there, thus making it more or less a sure thing for the ISP (and for the users, since if they don't get enough people, the ISP gives the money back).

The big ISP's pretty much ignore smaller communities, so there is still a very large untapped market (at least in the US) for Internet service to small towns or rural areas. You can actually get quite a lot of users online before you have to get more phone lines and higher bandwidth, as well.

So, to sum up: minimum needed to be an ISP in a smaller town:

• Internet connection, at least 56K Frame relay, or higher.
• At least 12 phone lines.
• Dialup server (e.g. Ascend Max 4000/Portmaster III, or linux box with multi-line modem cards)
• Web/DNS/Mail/Radius authentication server Celeron 400, 128 meg Ram, 20 GB drive to start out. You can make these separate servers, but I've seen people run up to 500 virtual apache domains and about 10,000 email boxes on the same machine.
• Ability to remain calm under all customer calls.

That's my 2 cents. :-)

Best Advice

[Flounder]

There is an ISP in Hawaii that provides very inexpensive dial-up service, with a catch. No tech support. Period. They send you a sheet with your dial-up settings. And that's it. For experienced users only.

Tech support for newbies is, by far, the biggest pain in the a** for an ISP. Eliminate them, the job of running an ISP becomes almost enjoyable.

Re:Start up an ISP?

[jguthrie]

Indeed, and very much so.

With all due respect to that person of questionable intelligence who posted that rant about asking Slashdot about this as opposed to doing "research", well, I am the "pop" half of a "mom-and-pop" ISP in Houston, so I can probably give some sort of useful advice. In fact, in my opinion, and I've been doing this for most of the last decade, this sort of question asked in this venue is likely to produce more useful information than going to some library and trying to find a book that describes how to run an ISP. I really pity someone who tries to figure out how to start an ISP by reading back issues of "Boardwatch".

The first thing that I recommend to people who want to start their own ISP is professional psychiatric help. (In fact, I have the names of several very good psychiatrists and even a psychologist or two in and around Houston, TX, US.) If that doesn't convince them that it's more fun to go broke on a trip to Vegas than to accomplish the same task by starting an Internet business, then I can get down to brass tacks.

So, the advice (some of it contradictory) observations, and opinions in no particular order:

• The first thing you need to know about the ISP business is that it is not primarily a "technology" business. What I mean by that is that there is essentially no technical risk. What I mean by that is that you can start an ISP with equipment and software purchased off-the-shelf. That fact is why the margins are so low in the ISP business, and why it's so difficult to stay in business.
• I don't know what it costs in Australia, but it takes most people somewhere between 50,000 USD and 100,000 USD to start an ISP around here. The more technical expertise you have, the less you'll need to spend, but expect to spend at least as much on advertising as you do on ongoing service.
• Unfortunately, people who have money to invest typically want to invest vastly more than that in hopes of getting vastly more return. Around here, this means that you aren't likely to be able to get any venture capital unless you can figure out how to write a business plan that calls for spending maybe 20,000,000 USD per year and breaks even in 48 months. (Lately, I've been toying with the idea of putting together such a plan, getting an investor's money, then continuing to operate on the cheap. I could then break even in 48 hours.)
• If you do go for VC funding, try to talk only to people with money to invest. (Most of the people I've managed to talk to over the years have nothing but some fantasy about brokering a deal with some VC person for a share of the money. This works about as well as jet-propelled pigs.)
• Customer retention is the key to long-term success.
• The biggest barrier to customer retention is persistent connection difficulties. Most of the connection problems your customers will have will be due to the telephone company. However, your customers will be mad at you about it and demand that you fix it despite the fact that you can do nothing to help or hinder the process.
• If you can't balance a checkbook, hire someone who can and then watch them to make sure they don't have sticky fingers.
• If you can't set up a router, RADIUS server, and access equipment, hire someone who can and pay them a salary. Make sure their bonuses are related to uptime rather than "face" time. ("I don't care if the sysadmin's not here as long as the network is running," should be your motto.)
• Everyone at the ISP should resign themselves to the fact that, at a startup ISP, everybody does sales and everybody does tech support.
• Put together service packages and every time someone wants you to bid on a special project, either make it out of those service packages (after the fashion of a "Chinese menu") or don't bid. Putting together bids is a major time sink if you let it become one.
• Try to not lose money on anything you sell.
• Billing systems suck. Some of them suck in different ways and some of them suck expensively, but they all suck.
• Talk to a lawyer and an accountant about the form that the business should take. Limit your liability as much as you can. You won't be able to get out of all of it, because some of the creditors will insist on personal guarantees of payment, but get out of as much as you can.
• Explore and develop ancillary sources of revenue. I know an ISP here in Texas that does for-pay computer training three or four times a year.
• Learn how to work the telephone system to get what you want. I don't know what that means in context of an Australian ISP, because it's different from working the SBC system, but you'll need to do it.
• Use access concentrators and digital phone lines wherever possible. I've used Ascend (now Lucent) Max equipment and they work. Others have used Cisco equipment with similar results.
• It may be possible to find someone who will lease you access to their modem banks. If you can do that, it might be worth your while. However, it adds an additional layer for a customer's problem report to go through, so it may not be worth the hassle.
• Keep backups of all customer data.
• If a vendor neglects to bill you, put the money you would be paying them into an interest-bearing account and leave it there until they notice that you haven't paid. Just because you didn't get the bill doesn't mean that you don't have to pay the bill.
• Try to make yourself superfluous as quickly as possible. Essential people don't get vacations.
• The most necessary tech support training isn't technical. The most important skill a telephone tech support person has is the ability to control a call. The user should be responding to you, not the other way around.

I'm sure there's more, but that's enough for now.

The myth of the failing mom & pop ISP

[PacketMaster]

I've done extensive contracting work for "local" ISPs and I can tell you that they are in no way on their way out. Most people out there are happy with dial-up and aren't interested in the prohibitative prices of broadband. I recently completed some contracting work for an ISP in Western Pennsylvania and they went from 0 users to 1500+ users in about 3 months. There are four major keys to having a successful ISP:

1) You have to be financially committed to grow. When your dial-in lines are full during peak times consistently you need to add more. Nothing will cost you users faster than an ISP that rings busy for 20 minutes before a user can connect. A good rule of thumb is to have enough lines to support 25%-30% of your userbase being connected at any one time. Keep good logs of connect times and if you need more lines, buy them!

2) Provide good service. People will stick with the ISP that provides good service to user's problems, even if it's slightly more than the ISP down the road. Get a good ISP management tool that makes handling your radius/dual-up authentication, e-mail and other services easy and hire a couple of minimum wage people with half a brain to field "1st Level" calls -- high school students would be perfect in this area. That'll take care of 90% of your problems with users who most likely can't type their password or fiddled with their settings. Turnover in these jobs is high, so make sure you have a dummy-proof system that makes training a new hire easy. There are many freeware FAQ/Knowledge Base applications out there to automate this. The one application you DO NOT want to use is ISP Power no matter what their salesman says.

3) Have a solid person or persons behind the technology side of things. Either do it yourself if you have the knowledge, hire someone knowledgable or contract out the work (what I do part-time). Corporate IT is a lot different than ISP IT. Hire someone who knows routers, Radius, etc.. They need to be articulate becuase you'll have an uphill fight with the local teleco for both your frame connections and your Dial-In BRIs. Remeber that local Telecos push their own ISP service and you will not get good support from them if you're an ISP. You need to have someone prepared for a long drawn-out battle who can provide sound answers and be able to monitor and gather data on bandwidth and performance with which to bombard the teleco's tech support. The first words out of their mouth will be "Do you have your router configured properly" and will hammer this at you until you prove conclusively that it's not your router. You need to pick a platform and stay committed to it. Pick an e-mail server that is EASY to configure and maintain. MDaemon for NT/2000 and Qmail for Linux/Unix/BSD are good choices. Pick a hardware vendor you can have a good relationship with. 3Com is an excellent choice for ISP type hardware. Very few ISPs needs the power of Cisco equipment.

4) Take security seriously. Your 31337 Skriptors love to find ISPs with little thought to security or else security that has gone lax. Enforce a password policy, keep good logs and have monitoring systems up and running. Have a zero tolerance policy for spammers and other crackers. Invest in at least a minimal firewall setup for your servers. Spend the time to learn the Unix tools for firewalling or look at a good NT package such as BlackIce (again depending on chosen platform).

You can still be very successful with dial-up ISPs. Broadband will eventually either become cheaper allowing local ISPs to compete in that area or the government will eventually crack those markets open. It's just a matter of time.

One last thing, offer Front Page extension support! I can hear the booing from the /. community on this point but that is what people want, especially from their local ISP. They don't want to mess with FTP regardless of how good the directions are. They want to use their nice shiny pre-packaged Microsoft Web Publication Wizard.

Building an ISP, caveats ...

[JoeGee]

From personal experience these are the best tips I can give you:

1. Speak to your local telephone company at length. Ask them questions like "how old is your wiring", "do you support channelized E-1 lines for 56k dialup", "do you use load coils to extend your reach", and "what kind of distance charge do you have for bringing in a data connection"?
2. You must be a very, very patient person. Be prepared to work long hours with the telephone company and your data connection provider. Be prepared to have them give you conflicting information. Be prepared for your local telephone company to pass the buck for connection difficulties to you. Be prepared to call your local telephone company on the carpet.
3. Ask your data provider and your telco to provide you with their tech support hours. Ask them for direct phone numbers and contact names. If they limit their hours, ask them if they have extended support options.
4. Get to know all of your service representatives on a first-name basis. Send them Christmas cards. Be nice to them. You may need to call in favors some time at 3:30 in the morning when your data circuit dies and you have to call their tech support.
5. Expect to spend more money initially than you bring in. Do not expect your business to pay for itself in under twelve months, meaning have at least a twelve months' supply of operating capital available (the more, the better.)
6. Suscribe to your competitor's service. You have to know how they perform, they set the standard which you must at least meet, or exceed.
7. Give referral credits. Give a $5.00 discount on a month's service to anyone who refers a friend to your service.
8. Pinch your pennies (or your five cent pieces ;).) You do not want to be a dot bomb. Have a three year business plan in place when you start up, stick to it.
9. You are a utility, not a service. In the rural area where I live I advised the small startup ISP to sell themselves as a utility, meaning they are more like a cable company or a satellite TV provider than a service. In my opinion this helps to foster a "must have" idea in customer's minds.
10. No one ever brings in a television to their cable company office saying "my TV is not working, what is wrong with your service", but they'll do it to you with their computers. Be prepared to answer all sorts of ridiculous questions, face all sorts of ridiculous situations, and have at least five percent of your neighbors actively hating you. :)
11. Get an unlisted telephone number. People you do not know will be calling you at home at 4 AM, screaming in your ear, MY SERVICE IS NOT WORKING.
12. If you are starting with yourself as the only employee, be prepared to forgo a social life. Search "monasteries, coping skills, celibacy" on Yahoo. Implement their suggestions.
13. Know your equipment. Be prepared to study, study, study. Neither your telephone company nor your data provider are likely to be experts on the equipment you purchase, so be prepared to be on your own in configuring equipment. Before you buy, see if your telco has any prefered RAS equipment. As a suggestion, see if the equipment provider will work with your telco in configuring equipment.
14. Find a support group. Get in touch with a few other ISP's not in your area. Use newgroups. Keep on good terms with knowledgable friends.
15. Know your local laws regarding Internet and telecommunications. Get a user agreement, have it gone over by an attorney, and enforce it rigidly.
16. Finally, never turn down a prescription for Prozac. Prozac is your friend. Prozac makes it better. Prozac may keep you from strangling the next guy who walks in and asks "I just bought a Commodore 64 at a rummage sale, and I cannot wait for you to connect it to the Internet." :)
17. HEY, which reminds me of one more suggestion -- know your limits. Are you going to try to connect any DOS machines? What about Macintosh? How slow of a modem is acceptable? :)

Think it over, carefully. It's not a living, it's a lifestyle. Go with God. :)

Two Important Words: Think First!

[Joel Rowbottom]

There will probably be a lot of people say this, but the first answer which springs to mind is don't do it. It's very costly, and certainly no longer something you can do with a single Linux box and a DSL connection.

That said, if you do want to do it, first thing you'll need to do is make sure your business plan will be profitable. I know it's tedious, but sit down in front of a spreadsheet program and work it all out: hard questions include:

• Is it feasible? You did say you lived in the country - there's probably a reason there's only one ISP.
• Will it make money? If not, why are you doing it - you're going to have to work out how to pay the rent some other way if it's not going to pay.
• How will you support your customers? When they phone you in the middle of the night because their printer's stopped working, have you got the patience to help them or tactfully tell them that it's not your problem?
• What if your upstream provider goes bust? Once upon a time nobody thought this would ever happen, but after several major providers filed for Chapter 11... :(
• Do you have contracts? Seriously, in this increasingly litigious world you can save lots of hassle, stress, lost sleep, and ultimately cash, by hiring yourself a good lawyer from day 1 who will make sure that all the limited liability blurb is in your contracts. There are a lot of bedroom ISPs out there who have fallen over through someone down the road taking legal action, and them not having the money to fight it. Hence the case doesn't even get to court and the ISP is bust. Be serious: get your contracts sorted and stick to them.
• How will you buy all the kit? Dialup stacks and routing hardware don't grow on trees you know - however I've seen quite a few good deals on Cisco dialup kit on eBay, and if you're doing partial BGP to peer with your upstream providers a secondhand Cisco 3640 will quite happily do the job.
• How will you manage the subscriber base? There are several prebuilt and homecooked packages out there, but you'll probably find that you don't really know what you need until you're up and running. Remember it's a rental system you're running, so you'll need to decide how to work the finances and contracts if someone cancels mid-term - do you refund them an unused portion of a month? What defines a "month", is it the number of days in that month or do you just split a year into 12 of them?

Now assuming you've asked yourself the difficult questions and got satisfactory answers, go out and find yourself a good accountant or at least someone who'll take care of the day-to-day finances for you. If you're a scatterbrained geek like I, then you'll have to reel in some favours perhaps. I use my wife for that sort of thing - it works quite well ;)

Then, and only then, do you start to work out your network map, and do all the fun stuff. Don't be a Dot Com ;)

Note: I've been brainstorming while writing this so there will be a lot I've missed out. I've rescued and set up ISPs and businesses before, some of which have succeeded and some of which have failed. I speak from experience of 1991 through to the present, so don't take this as a base course in setting up a business ;) - usual #include <disclaimer.h> I'm afraid ;)))

Joel.

Or Better Still...

[Patoski]

There are several companies to whom you can farm out your dial up service to. They even take care of all the tech support for you which is by far the biggest pain in the ass. You get charged about $10/per customer to use their phone lines and tech support droids. That way you only have to worry about a pipe for your servers and high speed bandwidth (DSL/Cable) start up costs [which can be substantial]. Technically you could run your whole operation off of one box (I used to know of a small shop that did so years ago) although I wouldn't recommend it. When it comes to your servers think redundancy, redundancy redundancy! It's going to be tough for the little guy just starting out but there are areas that are more rural areas that are under serviced or have no service at all by the big boys. This would be your best bet I think. Signing up companies for domain and web site hosting should be your cash cow. The dial up folks won't make you rich but companies you get signed up should help alot. Have fun!!!

I've done it, and it wasn't very pretty.

[mikehoskins]

I had major problems with the equipment, users, telco, and software. I'd steer clear, IMHO. I lost major bucks on the deal, and I'm no newbie. You need to have major cash flow, have great business sense (far more important than your technical skills), need to really run it like a business, have an awesome accountant, market, have great customer service, oh and have at least semi-good tech skills.

My point is that being an ISP is a BUSINESS, not a tech job. Trained monkeys can (almost) do the tech work, especially now that things are SOOOO easy. Hopefully, you can get funding and business people on board to run it. Business concerns are primary, tech concerns are either secondary (or even tertiary). And, hopefully your local telco or other local ISP's aren't better at running an ISP business than you are, or they'll kill you off....

Nope, as one who has lost major money, it isn't worth it....

Re:Start up an ISP?

[slyall]

I work with a larger Australian ISP so I talk with a lot of customers who run smaller Australian ISPs. Many of them are one or two empolyees, a couple of hundred customers and only around 128k/b of bandwidth. Most of them are never going to make money. Points to watch for:

• Work out how much income you need. Normal dialup customers are only going to make you a couple of dollars per month each ( if you are lucky) so you need a couple of thousand just to pay wages for one or two people.
• From the start decide how you are going to handle helpdesk. Even a small number of calls can quickly tie you up. If you are paying yourself $20 per hour and only making $2/month from each dialup customer then a 30 minute phonecall will use up 5 months profit from that customer. Others have suggested no-helpdesk ISPs.
• Talk to other ISP people in your area about what costs and problems they are having with Telstra etc. You say the existing ISP in the area is bad for disconnections. Is this because of their equipment or because of the bad lines in your area.
• Watch out for freeloaders if you have limited resources and cheap accounts. Think about what will happen if 5 of 10 percent of your customers stay logged in 24x7 running Napster type programs. This is a realistic number right now and only likely to get higher. Can your provisioning and revenue model support this? If you then you may need to structure accounts to avoid these customers or make them pay extra.
• Make sure you provision realisticly. Don't think you can get away with a 128K/b upstream circuit and 20 dialup lines. If people can go to one of the bigger National providers and get decent download speeds then they won't touch you.
• Plan for growth. Make sure you have aliases for all you servers (smtp , pop etc). Get your own AS number and own networks early. Make sure customers use server-assigned dns servers and dynamic IPs.
• Don't even think of using modems. Get Ascend or Cisco NAS boxes or similar. A cisco router might be overkill at the start so use a dedicated PC running zebra and you upstream circuit until it is worth buying it. Get a switch for your LAN, hubs are not worth the trouble.
• Shop around the different providers (telstra, optus, connect.com etc) for the best deal. Don't forget to check out the satellite providers. Look at getting a link into a local Ausbone node and perhaps buying bandwidth off someone there.
• Run a transparant proxy server from the start. It'll say you huge amounts of bandwidth.
• Don't overdo things. Buy 3 or 4 PCs to handle everything. Run indentical Linux installs on each ( Debian is good for this and easy to automate updates) and then divide your services between them. If you keep growing you can buy a new PCs every few months and split off services. When possible have multiple or backup smtp/mx/proxy/news servers. Make sure you backup regularly, with debian servers I just tar up /etc , grab the list of packages and copy them to another machine. Once a week I backup those to CDROM.
• Take care with you advertising. It's easy to spend thousands of dollars on radio or Newspaper ads and having nothing to show for it. Target the people you are after, encourage customers to sign up others. Make sure that when people login to irc, chatgroups or usenet your domain shows up on their reverse.
• Be careful with monthly bills. Don't give people credit. If they are not paying you then lock their account. Try and setup your system so people are paying in advance.
• Keep your accounts simple. You shouldn't have more than 3 or 4 different types of home accounts (1 or 2 is better). Make sure each type of account is going to make money. Consider an offpeak account that only lets people login during the night or when usage is low.
• MAke sure you terms and conditions allow you to close accounts for spamming, hacking and IMPORTANTLY for no reason at all.
• If a customer is giving you grief and taking up huge amounts of your time then just close their account and refund their money. You should NEVER spend hours per week on someone that pays you $20-$30 per month. To make money you have to spend well under 1 hour per YEAR per customer.
• Make sure you monitor your network. Something like bigbrother hooked up to a pager is a good start. Get the after hours numbers for your suppliers, get the cellphone number for your account manager etc. Look at going with multiple providers for backup. A flat rate main circuit and a $-per-megabyte backup is a good combination. Run you own BGP and you can make fallover automatic.
• Have a backup computer or two ready. If one of your machines dies completely at 10pm on a Friday you want to be back up quickly. Even if you don't have multiple machines for everything a single machine you can convert to anything should be enough.
• Keep your servers tidy. Get a rack and consider rack mounted PCs. Don't buy 1 or 2 RU machines unless you are paying for the space but full size rackmount cases are tidy and easy to maintain.
• Make sure you have money to survive. You will need at least $100k to get yourself setup and surviving for the first year. There is no way you will be profitable during that time and more than minimal wages are unlikely.
• Don't waste money. You don't need a fancy office, high end laptops etc.

Wireless?

[yasth]

Hmm a limited (downtown only???) wireless ISP would rock, and would probably be your best bet for leapfroging the established players, as even in Cable/DSL there are some very large players, but wireless is still untapped(for the most part)

Services, support, smiles

[HadronPie]

My suggestions (as a user of multiple ISPs) would be to offer services such as:

• shell-only accounts
• static IPs
• metered toll-free access
• "free" email accounts
• maybe email virus scanning at the server

Charge a nominal monthly rate for the first three. Lump the other two into your monthly cost. Of course you must have automated credit card billing.

Offer 24/7 toll-free tech support. Keep the call center well-staffed and don't punish/reward the support staff based on how long it takes to complete a call. Integrate the web and email support services. Setup a support evaluation survey that gives people points towards freebies if they complete the survey. Track this info like all hell. That costs a lot, obviously, but it'll be a very good selling point.

Your advantage over the Big Guys will come from offering services they don't offer at comparable rates with friendly, effective technical support. There's not much else to an ISP.

Oh yeah, game servers, irc servers, news servers.

A Time Machine

[Tech187]

To start a successful ISP you'll need a time machine that can transport you back 4-6 years. And going six years back might not even be enough. The biz is locked up now and the consolidation is in process. It's completely the wrong time to try to start one up. Unless you've got a source for cheap wholesale bandwidth and are way, way out in the hinterland somewhere that AOL, MSN, and Juno don't have a local number.