Newbie's Guide to Starting an ISP
You get an adrenaline rush every time you think about starting your own ISP. Although the cost of entering this game drops every day, competition is heating up, so this is a big decision.
by Christopher M. Knight
[August 9, 1999]
[It says it right in the title, but I want to re-emphasize that this article is aimed at ISP newbies; it's not meant to be comprehensive. It also assumes you're not operating a business of any kind.]
Brief initial reality check:
This not an easy business to navigate. You will work very long hours, you'll have to read tens of thousands of emails to educate yourself, and you can count on having it cost at least twice as much as you expected. Many critical items?such as your telco?will be out of your control, whereas your subscribers always expect you to be in control of everything. Finally, believe it or not, it's near suicide to start this business without having a clear ISP exit strategy.
Another note: If you're familiar with the differences between facilities based and non-facilities based ISPs (terms I reference later in this article), read on. Make Your ISP Facilities-Based or Virtual?
Okay, now let's take a deep breath and look into the basics of what you must have in order to get started and open for business.
Top 7 Items You Will Need To Start Your ISP
1. Generic business underpinnings
This includes any licenses your state requires, your federal and state employer ID numbers (for ISPs in the U.S.), your domain name registration and trademark application, your business checking account and banking relationship?along with a possible preapproved credit line?and a merchant account so that you can accept credit cards and direct checking account drafts. You will also need some other business basics, such as a phone system, security service, CPA & Attorney.
2. ISP billing and accounting software
This can make or break your profitability and your ability to maintain your business as you grow. A sizable industry has grown up developing and offering customizable off-the-shelf ISP invoicing systems?many for under $2,000.
Whatever billing program you buy should integrate with a general ledger, and be able to track accounts receivable, accounts payable, invoices, customers, usage, overages, and your checking register. Some packages even provide instant profit/loss balance sheets?along with reports that help you make better business decisions. You will also need a collection agency to help you collect on bad debts; I don't know a single ISP that doesn't have bad debts collections issues.
3. Internet feed and local loop
In order for you to provide Internet access, you have to get it from a regional or national backbone provider. The connection between your provider and your physical location is called the local loop. The cost for this can range from a few thousand dollars one-time charge?if you are co-located within an ISP building and own your local loop?to between $125 and $5,000 per month or more, depending on the size of your pipe and how many miles you are from your provider and your local telco central office or CO.
Internet feed will cost you $800 to $2,000 per T1 worth of Internet access per month. Assuming an 8:1 user-to-modem ratio on your dialup ports, a single T1 will be enough to feed 200 concurrent connections or about your first 1,600 customers. (This means for each modem port you have eight customers signed up. They won't all be trying to use it at the same time, you hope.)
While some small ISPs use POTS (Plain Old Telephone Service), adding a line when the old ones become crowded, this is not something you should even consider. Rather, start up with a 24 channel T1 or PRI (Primary Rate Interface) circuit for your first access switch. Shop around, because prices for these services vary a great deal. Unless you have no alternative, do not negotiate directly with the incumbent major telco. Many of the best deals or promotions are available through their resellers and distributors.
Many backbone providers include Usenet feed via NNRP (Network News Relay Protocol) as part of their package. If your does, take it and run. You do not want to have to run your own Usenet servers if you're just starting out. You can do a DNS trick which will fake it and make it appear that you have your own Usenet server, when you don't.
4. Your network of servers
This includes your PC based servers that will handle your primary functions, which include DNS, mail, Web (both regular: httpd and secure: https), athentication, and news (if you don't outsource it). Theoretically, you can do all this on just one PC, but in my opinion, anyone who starts an ISP with fewer than four separate dedicated PC's is asking for trouble.
If you're going to be a facilities-based ISP, you will also need access servers (a relatively recent innovation that combines terminal server and modems). Good ones range from $7,000 on up. Some of the popular names of access servers include US Robotics/3Com, Ascend, Livingston, Cisco, and Nortel.
5. Marketing/Sales plan & budget
Most new ISPs skip this completely, and then wonder why they run out of money so quickly and have to sell out or go bankrupt. This is really the foundation for all of your ISP decisions. Here are some hard questions that require concrete answers before you begin:
What will make your ISP unique?
Why should someone buy access from you?
How are you going to acquire your first thousand customers?
How will you automate the order process?
How much should you pay to acquire new subscribers?
What are you going to do to retain your clients?
What will be your minimum standards for the level of service you wish to deliver to your subscribers?
What is your exit strategy?
While word of mouth can be a powerful sales agent for you, make sure you start out with a marketing budget; it will cost thousands of dollars to get the word out about your service. Moreover, it's foolish to believe that you can do it all (build, manage, support, and grow your ISP, no matter how small it is). You will need sales assistance?or at the very least, order taking?in the beginning.
6. Dialup kits
This is what the customer receives to get on your service. The easier it is to use and install, the higher your chance of converting your new trial customers into long term subscribers. You don't have to design or create your dialup kit; there are many companies that can provide this. For example, both Microsoft and Netscape have custom kits you can implement.
Your entire growth and signup strategy is wrapped around how well your dialup kit gets your new users up and running on your service, so give it serious attention. Keep in mind that it can?should?serve as a marketing tool as well. For example, you may want to include a users guide for your ISP, some tips and tricks to help your new subscribers get the most out of their new service, and of course a toll free or local phone number they can call for help.
7. Technical expertise (tech labor)
I'm not sure whether it's better to be starting an ISP as a techie without much business knowledge, or a business person without the technical chops. Either way, your up-time, reputation, and revenue all ride on your company's technical competence. Skimp, and you'll have hundreds of subscribers keeping you up at night, wondering why they can't get on the Internet, and eventually wanting to cancel.
There are many different levels of techies, and they range from front-line customer service techs to your Chief Technology Officer. Your front-line tech answers incoming calls, and if s/he can't answer them, refers them up to the Sys Admin. If the Sys admin can't handle the call, s/he refers it up to the Sr. Sys Admin on duty. If the Sr. Sys Admin can't answer the question, it isn't possible.
Sys Admins don't interact as much with clients as much as they are the secret behind keeping the back office running properly. Your Chief Technology Officer is your smartest tech person in the office, who also understands the business implications of budgets, uptimes, metrics, customer service, and he or she probably has years of experience in the real world. A typical ISP under 2,000 subscribers should budget at least $150K-$300K+ for first the year's tech labor expenses.
If there's an 8th thing you should have, it would be Cash, and lots of it. If you've read my past articles, you'd know that I'm a big friend of outsourcing dialup ports, and funneling your all-too finite war chest into marketing and sales activities to get the highest ROI (Return on Investment). If you outsource your ports, you not only save on the capital startup costs, but you may not need the same level of tech labor on staff to manage your business. There are real disadvantages to outsourcing your Internet access ports, such as not having ultimate control over your network and being at the mercy of someone else when your network is down.
Tip: A good rule of thumb to follow is that your ISP should make you more money than you could earn by day trading ISP stocks. (Many ISP friends have reported 6 and 7 digit gains on their ISP stock market plays, vs. losses their ISP businesses.) I'm not suggesting that you play the ISP stocks instead of starting an ISP business, just pointing out that for the business to be financially sane, its goal must be to out-earn what you could make in the stock market with your finite investment capital.
Renewed Interest In Dial-up Surprises Old-Timers
Around the world, naive computer enthusiasts want to build new ISPs, much to the surprise of veterans who left the business. Here's what these college graduates will need?and how they may fail.
by Max Smetannikov
[June 21, 2002]
The first dial-up ISPs were startups ten to fifteen years ago. Now, a new wave of dial-up enthusiasts seems to be knocking on the established service providers' doors.
The world of the Internet has changed dramatically?but some simple truths that governed the last round of ISP buildouts remain.
First, the startups will learn to hate the local phone company (David Robertson, vice president of the Texas ISP Association (TISPA) describes the relationship between a local phone company and its ISP customer as similar to that between a car's windshield and a bug).
They'll need some cheap software. They'll need some servers. They should invest in a router. They are most definitely going to shop a lot for used networking gear.
"It's been almost ten years since I started my ISP and three since it went static on the ISP (dial-up) side," says Avi Freedman, chief technologist at Akamai and one of the very first commercial Internet entrepreneurs (he recently sold his ISP business to FASTNET). "After we finished the conversion to PRIs and to a commercial billing package (Optigold) four years ago, very little changed on the dial infrastructure side."
Much to the surprise of individuals like Freedman, there is a revival of sorts as entrepreneurs hope to launch small dial-up and dedicated access ISPs. These people are all over the world, and are looking for help and clues as to how to start an ISP.
"We, two brothers, freshly out of college, both graduates in Computer Engineering & applications, are very much willing to start a company providing Internet connectivity," writes Sourish Chandra to ISP-Planet from a Hotmail account.
Chandra is not alone. Bret Mingo, president of Annapolis, Md.-based CoreTel, says calls from small operators shopping for wholesale access have picked up considerably this year.
"I am not sure why," says Mingo, who aims to sell to mid-size service providers who compete with the regional phone companies.
Newbies like Chandra often come to entrenched providers like CoreTel with na?e questions about what they need to buy, both services and equipment, to get into the business. The good news is that the basics have hardly changed over the last decade.
Any ISP starts with a back end. There, start with a RADIUS server, which matches user names with passwords and e-mail addresses so that only paying subscribers can use your network. The necessary software is available either commercially or as an open source application (i.e. for free). It's a key piece of infrastructure. If it goes down, users can't log in. Most operators run it on two servers so that if one server fails, the other can take over. RADIUS software is typically optimized to run on Unix or Linux machines.
RADIUS has a role beyond security; this is also the machine that feeds the billing application. Keeping billing simple is important. It's a major reason why most ISPs went to flat rate plans. Flat rate plans (charging per month instead of per minute or per kilobyte) make it much cheaper to track users.
Even with a simple price plan, an ISP needs a good billing system. There's a lot of billing software on the market.
While every ISP needs to have good billing software, no ISP can skimp on customer service and tech support. The current industry formula says that an ISP needs at least one tech support person for every 2,000 to 3,000 users (and more for smaller ISPs), and, for every 100 to 300 new users signing up each month, one or two technicians handling problems and complaints.
The second piece of must have ISP software is a DNS server. This is an application that matches Internet IP addresses with site names. Without this software, users will not be able to access Web pages. The most popular software for DNS is BIND, and it's free. But the application has to run on a server as well, and the hardware is not free.
While the list of applications an ISP can sell online seems endless, it has become customary to offer e-mail with access. To support that, an ISP needs an e-mail program. Again, there are free software options like SendMail and a good array of commercial e-mail servers like Netscape. To process e-mail, an ISP will need to enable one or more of several e-mail transfer protocols, like Simple Mail Transfer Protocol (SMTP), POP3, or Internet Message Access Protocol (IMAP).
While the ISP is in startup mode, all of this software can live on one server, with a back up for redundancy. Linux is the operating system of choice for many Internet professionals because of its UNIX-like stability and Microsoft-like (don't tell anyone we said this) simplicity. It used to be that UNIX servers, like those made by Sun, were prohibitively expensive for startups. Now Sun's small machines, classified as 1U (one rack unit, the amount of space used on a server rack) go for about $1,000.
Once the back end is ready, the ISP is one step closer to getting the business off the ground. The next key element is a termination device to receive modem calls?or terminate dedicated lines if leased line access is part of the business plan. A brand legendary in the ISP world in that regard is Ascend, which was bought out by Lucent several years ago. Ascend's Max TNT was the most popular multiplexer in the late 1990s, and is still one of the favorites. Besides Ascend, Nortel and Alcatel as well as several lesser known manufacturers now make similar boxes.
This is how the ISP works: a subscriber makes a modem call into a phone number provided by an ISP. The phone call is terminated into a multiplexer, which is connected to the ISP's back end. The user is authenticated by the RADIUS server, and the call is sent on to the Internet. While servers can support certain routing functions, no self-respecting ISP would be in business without a router, typically a 2500 or 7500 series Cisco machine. A good size ISP needs a router that can handle Border Gateway Protocol (BGP), the switching protocol used at the core of today's networks.
Many startups buy most of their equipment used, and nowadays lots of used equipment is available on eBay and other Internet auction and ecommerce sites. Street prices for servers hover between $1,000 and $2,500, and a server must have dual power supply and Ethernet ports for easy and reliable connectivity. While a new Ascend TNT box costs close to $200,000, a used one can be bought for about $10,000.
The router is connected to the Internet with a dedicated line of some sort?T1, frame relay, Ethernet, anything the ISP can buy. The telco and backbone options will be determined by the businesslocation.
Dealing with the local phone company is a source of hardship and frustration for any ISP owner, especially at the startup stage. More often than not, a telco sees an ISP as just another customer, or even as a competitor, and is oblivious of the fact that service disruptions can destroy a service provider's business. Some ISP owners feel that phone companies even try to destroy their business. Says Freedman, "I still remember the time they needed UPS to deliver our phone bill."
Because the phone company can be difficult to work with, many ISPs choose to buy Internet access from larger ISPs. In the early days of the ISP business, when alternatives did not exist, a service provider would basically ask the telco for lots and lots of phone lines, and manage them through a multiplexer. Dial-up users would literally call the ISP's office location. As ISPs grew and became more geographically dispersed, many ISPs started to have telephone companies aggregate these phone numbers for them at a remote location (a stack of modems with no humans) and just send the calls to the ISP via a PRI (ISDN) line. So end users would call the switches of the telephone company, but get connected to the Internet via ISP's facilities.
In many parts of the globe, there is still no alternative to the above solution, which relies entirely on the phone company.
"People don't realize it, but it's actually harder to provide robust dial service than to provide quality leased line or colocation service," says Freedman.
In the United States, startups can now rent infrastructure?phone lines, RADIUS servers, PRIs?from wholesalers and become a "virtual ISP" (ISPs that build infrastructure are called "facilities based").
Take MegaPop, a StarNet subsidiary, originally a provider of wholesale dial-up exclusively. The company has a network covering the U.S. and Canada, and its latest program allows future ISP owners to set up their entire business through a Web portal.
"The cost for people who want to be ISPs these days is pretty much nothing except for marketing," says Michael Alonzo, StarNet manager of dial-up sales. "Large wholesalers like us enables end users to private label phone numbers and then resell them."
While large ISPs like MSN are among MegaPop customers, the company can go very low on price for smaller players. Its secret is in developing a low cost vehicle for ISP owners to add customers to the MegaPop network without letting end-users know they're on the MegaPop network.
Basically, an entrepreneur logs on to MegaPop's portal and assigns a name and an e-mail account to an end user. Users then log on to MegaPop's network under the name of their local ISP. MegaPop does require a $500 monthly minimum from its ISP customers, which its executives estimate means having 61 dial-up users on board.
"We feel like asking ISPs to have at least 61 users is pretty fair," says Alonzo.
http://www.tispa.org/
Renewed Interest In Dial-up Surprises Old-Timers
Around the world, naive computer enthusiasts want to build new ISPs, much to the surprise of veterans who left the business. Here's what these college graduates will need?and how they may fail.
by Max Smetannikov
[June 21, 2002]
The first dial-up ISPs were startups ten to fifteen years ago. Now, a new wave of dial-up enthusiasts seems to be knocking on the established service providers' doors.
The world of the Internet has changed dramatically?but some simple truths that governed the last round of ISP buildouts remain.
First, the startups will learn to hate the local phone company (David Robertson, vice president of the Texas ISP Association (TISPA) describes the relationship between a local phone company and its ISP customer as similar to that between a car's windshield and a bug).
They'll need some cheap software. They'll need some servers. They should invest in a router. They are most definitely going to shop a lot for used networking gear.
"It's been almost ten years since I started my ISP and three since it went static on the ISP (dial-up) side," says Avi Freedman, chief technologist at Akamai and one of the very first commercial Internet entrepreneurs (he recently sold his ISP business to FASTNET). "After we finished the conversion to PRIs and to a commercial billing package (Optigold) four years ago, very little changed on the dial infrastructure side."
Much to the surprise of individuals like Freedman, there is a revival of sorts as entrepreneurs hope to launch small dial-up and dedicated access ISPs. These people are all over the world, and are looking for help and clues as to how to start an ISP.
"We, two brothers, freshly out of college, both graduates in Computer Engineering & applications, are very much willing to start a company providing Internet connectivity," writes Sourish Chandra to ISP-Planet from a Hotmail account.
Chandra is not alone. Bret Mingo, president of Annapolis, Md.-based CoreTel, says calls from small operators shopping for wholesale access have picked up considerably this year.
"I am not sure why," says Mingo, who aims to sell to mid-size service providers who compete with the regional phone companies.
Newbies like Chandra often come to entrenched providers like CoreTel with na?e questions about what they need to buy, both services and equipment, to get into the business. The good news is that the basics have hardly changed over the last decade.
Any ISP starts with a back end. There, start with a RADIUS server, which matches user names with passwords and e-mail addresses so that only paying subscribers can use your network. The necessary software is available either commercially or as an open source application (i.e. for free). It's a key piece of infrastructure. If it goes down, users can't log in. Most operators run it on two servers so that if one server fails, the other can take over. RADIUS software is typically optimized to run on Unix or Linux machines.
RADIUS has a role beyond security; this is also the machine that feeds the billing application. Keeping billing simple is important. It's a major reason why most ISPs went to flat rate plans. Flat rate plans (charging per month instead of per minute or per kilobyte) make it much cheaper to track users.
Even with a simple price plan, an ISP needs a good billing system. There's a lot of billing software on the market.
While every ISP needs to have good billing software, no ISP can skimp on customer service and tech support. The current industry formula says that an ISP needs at least one tech support person for every 2,000 to 3,000 users (and more for smaller ISPs), and, for every 100 to 300 new users signing up each month, one or two technicians handling problems and complaints.
The second piece of must have ISP software is a DNS server. This is an application that matches Internet IP addresses with site names. Without this software, users will not be able to access Web pages. The most popular software for DNS is BIND, and it's free. But the application has to run on a server as well, and the hardware is not free.
While the list of applications an ISP can sell online seems endless, it has become customary to offer e-mail with access. To support that, an ISP needs an e-mail program. Again, there are free software options like SendMail and a good array of commercial e-mail servers like Netscape. To process e-mail, an ISP will need to enable one or more of several e-mail transfer protocols, like Simple Mail Transfer Protocol (SMTP), POP3, or Internet Message Access Protocol (IMAP).
While the ISP is in startup mode, all of this software can live on one server, with a back up for redundancy. Linux is the operating system of choice for many Internet professionals because of its UNIX-like stability and Microsoft-like (don't tell anyone we said this) simplicity. It used to be that UNIX servers, like those made by Sun, were prohibitively expensive for startups. Now Sun's small machines, classified as 1U (one rack unit, the amount of space used on a server rack) go for about $1,000.
Once the back end is ready, the ISP is one step closer to getting the business off the ground. The next key element is a termination device to receive modem calls?or terminate dedicated lines if leased line access is part of the business plan. A brand legendary in the ISP world in that regard is Ascend, which was bought out by Lucent several years ago. Ascend's Max TNT was the most popular multiplexer in the late 1990s, and is still one of the favorites. Besides Ascend, Nortel and Alcatel as well as several lesser known manufacturers now make similar boxes.
This is how the ISP works: a subscriber makes a modem call into a phone number provided by an ISP. The phone call is terminated into a multiplexer, which is connected to the ISP's back end. The user is authenticated by the RADIUS server, and the call is sent on to the Internet. While servers can support certain routing functions, no self-respecting ISP would be in business without a router, typically a 2500 or 7500 series Cisco machine. A good size ISP needs a router that can handle Border Gateway Protocol (BGP), the switching protocol used at the core of today's networks.
Many startups buy most of their equipment used, and nowadays lots of used equipment is available on eBay and other Internet auction and ecommerce sites. Street prices for servers hover between $1,000 and $2,500, and a server must have dual power supply and Ethernet ports for easy and reliable connectivity. While a new Ascend TNT box costs close to $200,000, a used one can be bought for about $10,000.
The router is connected to the Internet with a dedicated line of some sort?T1, frame relay, Ethernet, anything the ISP can buy. The telco and backbone options will be determined by the businesslocation.
Dealing with the local phone company is a source of hardship and frustration for any ISP owner, especially at the startup stage. More often than not, a telco sees an ISP as just another customer, or even as a competitor, and is oblivious of the fact that service disruptions can destroy a service provider's business. Some ISP owners feel that phone companies even try to destroy their business. Says Freedman, "I still remember the time they needed UPS to deliver our phone bill."
Because the phone company can be difficult to work with, many ISPs choose to buy Internet access from larger ISPs. In the early days of the ISP business, when alternatives did not exist, a service provider would basically ask the telco for lots and lots of phone lines, and manage them through a multiplexer. Dial-up users would literally call the ISP's office location. As ISPs grew and became more geographically dispersed, many ISPs started to have telephone companies aggregate these phone numbers for them at a remote location (a stack of modems with no humans) and just send the calls to the ISP via a PRI (ISDN) line. So end users would call the switches of the telephone company, but get connected to the Internet via ISP's facilities.
In many parts of the globe, there is still no alternative to the above solution, which relies entirely on the phone company.
"People don't realize it, but it's actually harder to provide robust dial service than to provide quality leased line or colocation service," says Freedman.
In the United States, startups can now rent infrastructure?phone lines, RADIUS servers, PRIs?from wholesalers and become a "virtual ISP" (ISPs that build infrastructure are called "facilities based").
Take MegaPop, a StarNet subsidiary, originally a provider of wholesale dial-up exclusively. The company has a network covering the U.S. and Canada, and its latest program allows future ISP owners to set up their entire business through a Web portal.
"The cost for people who want to be ISPs these days is pretty much nothing except for marketing," says Michael Alonzo, StarNet manager of dial-up sales. "Large wholesalers like us enables end users to private label phone numbers and then resell them."
While large ISPs like MSN are among MegaPop customers, the company can go very low on price for smaller players. Its secret is in developing a low cost vehicle for ISP owners to add customers to the MegaPop network without letting end-users know they're on the MegaPop network.
Basically, an entrepreneur logs on to MegaPop's portal and assigns a name and an e-mail account to an end user. Users then log on to MegaPop's network under the name of their local ISP. MegaPop does require a $500 monthly minimum from its ISP customers, which its executives estimate means having 61 dial-up users on board.
"We feel like asking ISPs to have at least 61 users is pretty fair," says Alonzo.
Viability Analysis
The desire to be in business is an important ingredient for success, but the ISP industry is maturing rapidly and, these days, startup plans need a stiff dose of reality checking. Prescription: Business viability study.
by Christopher Knight
[September 19, 2000]
So, you want to be an ISP? Before you rush out and start ordering equipment and talking to upstream providers, you need to do some major homework.
Back in 1994, no one entering the ISP space did viability studies. Hundreds of Internet pioneers simply charged into the Internet access service business because of the fun, challenge, culture?and the sheer market demand for it.
Today you'd be a fool to enter this business without a solid business plan, plenty of capital, as many 'unfair' competitive advantages as you can possibly muster, and an aggressive sales/marketing strategy as part of your core being.
In case no one has mentioned this to you, dialup Internet access has become a commodity in much of the world. In developing countries, if it isn't already a commodity, it will be soon.
If you take a look at the many publicly traded ISPs who have lost hundreds of $millions off their market capitalization over the past year, you'll begin to realize that the market has decided that no one is going to get rich in the ISP space . . . at least as of right now.
If you've read my columns, you know that I believe the "market is always right." On the other hand, I believe you still can get rich in the ISP space, but it will require you to adjust your thinking, going beyond what may look like a typical ISP today to what a typical "profitable" ISP will look like tomorrow.
Two Components of Viability
Component 1, Business Viability, is largely subjective, but can be gauged by the following:
Proven management/leadership team. Any prior success in this field increase your business viability rating and improve your odds of succeeding.
Capitalization: Do you have enough capital? How long can you go on before negative cash flow shuts you down? Or to put a more positive spin on the question, Can you find a way to stay alive until achieving profitability or a successful exit?
Flexibility: Are you willing to do whatever it takes?including abandoning everything you thought about the way 'things' should be done?in the name of achieving business viability? Your current thinking that has gotten you this far is the same thinking that will not be able to solve your new problems. Business viability requires the top leadership to evolve in order for the company to become successful in the long term.
Component 2, Market Viability, is more concrete and quantifiable. It can be determined by:
Significant market demand or market receptivity to your services within the geographic region you wish to serve.
Significant market size.
Lack of significant aggressive competitors in the region.
Getting past the critical mass needed to break even and sustain growth.
Company strengths that are in line with the market's needs.
Running the Numbers
Here is a simple formula for calculating ISP market viability, as it relates to Internet access services:
Market size = (Number of computers x Internet access penetration) / Total number of competitors.
Here is a highly simplified example based on this formula:
In the city of Vulcan, there are 200,000 people and 10,000 businesses. We know that 55% of the residents have computers and 85% of the businesses have computers. There is about 60% market penetration. We know there are 35 competitors in the region, including regional and national.
Here's the calculation:
200,000 x 55% = 110,000 individuals have computers that potentially could use the Internet.
10,000 x 85% = 8,500 businesses have computers that potentially could use the Internet.
110,000 + 8,500 = 118,500 computers (or networks) in this population.
118,500 computers x 60% Internet access penetration = 71,100 people and businesses that need Internet access services.
71,100 / 35 competitors = 2,031 subscribers per competitor?if every competitor were equal, but experience tells us this is a 'survival of the fittest' game and that every competitor will not be equal.
What if there were only five competitors?
71,100 / 5 = 14,220 per competitor, which means there is a serious chance for you to come in and achieve success in this market.
Details, details
Missing from this oversimplified formula are yearly growth rates and the effects of increased penetration of more computers and more folks wanting Internet access in your region.
You must also assess your competitor's strengths and weaknesses. If, in the above example, 30 of the competitors were national ISPs and 5 were local, then perhaps your local approach and aggressiveness might be able to displace the distant, almost non-caring national ISP player.
Where to Get the Market Data
To do a competitive analysis, you need good data?not data that tells you what you want to hear. Don't be afraid to pay for market data. Most of the best ISP market information from in-depth studies is available for sale from these ISP Industry Research Analysts:
internet.com's AllNetResearch: The Superstore for Internet Research
The Burton Group: Focuses on analysis of emerging network-computing technologies
Cahners In-Stat Group: Helping you make better business decisions through real-time research
internet.com's CyberAtlas: Internet Statistics and Market Research for Web Marketers
Forrester Research: Helping business thrive on technology changes
Gartner Group: Access to extensive collection of IT research and analysis
International Data Corporation: The industry's most comprehensive resource on worldwide IT markets
META Group: IT research and unlimited analyst consultation
NUA Internet Surveys: Often covering the online/ISP access industry
Patricia Seybold Group: Strategic technology and business solutions
StatMarket: Accurate Internet statistics and user trends in real time.
Yankee Group: Strategic planning, technology forecasting, and market research
Zona Research: Information and advice for the Internet industry
Which Niche to Conquer?
"If you can't be number one in the niche that you've chosen (because of competition or other external influences), you should move into?or create?a niche where you can dominate or be number one."
?Al Ries and Jack Trout, The 22 Immutable Laws of Marketing
by Christopher Knight
[October 4, 2000]
If Jack of all trades is a master of none, then why does Jack keep trying to be everything to everybody? Answer: He just doesn't know any better.
In today's competitive ISP space, the choice of attempting to be a master of all ISP-related services is no longer a realistic option. Nowadays, you must choose a narrow but profitable path?a product and service mix that the market will support?lest you lose the game altogether.
I have found that there are two primary business modes that ISPs operate in:
Profit Maximization Mode
Market share takes a back seat to profits?across the board. Rather than investing aggressively in acquiring new sales (except inasmuch as that leads to increased profit), you strive to maximize the profit you are making on your installed base. This is common in most mature ISPs.
Growth Maximization Mode
Short-term profit does not matter; you strive to achieve a market dominance position at all costs. (A down-side of this mode is that it's easy to run out of cash, resulting in bankruptcy or scrambling to find additional unplanned financing to sustain operations in the process.) This is most common in startup ISPs.
Since the title of this article is "Which Niche to Conquer," I'm going to assume that your ISP is operating in Growth Maximization mode, because that is the approach you'll need if you're going to conquer your chosen niche. Generally speaking, you cannot conquer a niche without being willing to shave off a few margin points to acquire the clients that come with that niche.
Choosing a primary niche
Even though you can segment the markets you plan to serve, if you hope to be successful in the long term, you must still choose one primary niche that you will call your own.
Here is a list of factors you should consider when choosing your ISP niche:
Market size and the kind of service or product you want to offer
Market demand for the service or products you want to offer
Your geographic coverage area
Your current or expected future business size
Your available capital and credit
Your talent and staff expertise
Your network capacity and strength or lack thereof
Your vendor relationships and alliances with strategic partners
To stimulate your thinking, here is a list of possible ISP niches that you could enter. There are many more; this is just to get your wheels spinning:
DSL only, or primary focus
Wireless/microwave specialists
Corporate WAN management (security/firewall/support)
ISDN/dedicated leased line
Wholesaler to other ISPs
Website hosting, promotion, design services
E-mail hosting or outsourcing services
Application or Business Process Services/Support
Outsourced expert technical support services
In-home or On-site setup and installation services
Identifying who you do not wish to serve
If you're having a hard time choosing which niche you wish to serve, start by deciding whom specifically you are not going to serve. For many ISPs, it may seem like 90 percent of the problem clients are from 10 percent of your client base, and therefore you should fire your problem clients - those who do not conform to your ideal customer profile.
Right now, stop reading, and make a list of undesirable customer traits that you and your team dislike serving. Once you have this list, it can help guide you closer to which niche specifically you wish to serve as your primary target.
Make Your ISP Facilities-Based or Non?
Part 1
Planning to build a new ISP? Here's an important set of choices you'll face.
by Christopher M. Knight
The decision to go facilities-based or non-facilities-based isn't a simple one. In this two-part series, we'll raise a number of issues that you might want to chew on while considering this question. But first, we'll nail down some definitions before we launch into the subject.
Facilities-Based ISP means that you OWN your own dialup access servers or switches.
Non-Facilities-Based ISP means that you do NOT own ANY dialup access servers or switches.
Hybrid ISP means that you own some of your POPs (Points of Presence) but your customers can connect to your network through you have other POPs that you do NOT own, but someone elses POPs.
Many ISPs are implementing a combination of facilities based ISP and non-facilities based wholesaling with any one of a number of dialup access switch aggregators, such as UUnet, MegaPOP, GTE, PSInet, etc. Just about every major tier one backbone provider is providing wholesale access to their POPs.
We'll take a look at some of the advantages and disadvantages of each setup. Today, we'll focus on facilities-based businesses; tomorrow, we'll examine at the non-facilities-based alternative.
Advantages of Facilities-Based ISPs:
You control 100 percent of what is going on with your dialup switches
Possible lower cost if you hit sufficient economies of scale
Makes it easier to wholesale your network to other ISPs?whereas it may be impossible to cost effectively wholesale a non-facilities-based ISP where you are buying wholesale and selling wholesale.
You may be able to receive additional termination revenues if you have telco ties or CLEC status.
Possibly faster to the market with the latest dialup enhancements since you control the speed of rolling out new technology. Most non-facilities based ISPs have to wait till their wholesaler catches up with the market when new technology rolls into town.
Disadvantages of Facilities Based ISPs:
It can be a negative cash flow pit. You'll spend a lot of capital on which you may not get a very high return. Flexibility is extremely limited because you may have signed leases and commitments to your telco that carries high penalties for early termination.
Your hardware will depreciate faster than the government will allow for tax purposes, so your depreciation and write-off expenses after you dispose of old access switches may be high.
Must have highly trained experts on staff to manage your network around the clock. It can get very expensive to dispatch technicians all over the country to manage the network if your POPs are very spread out.
Advantages of Non-Facilities Based ISPs
Make Your ISP Facilities-Based or Virtual?
?????Part 2
Yesterday, we examined the Pros and Cons of running Facilities-Based ISPs. Today, we turn the tables and consider the advantages and disadvantages of NON-facilities based ISPs:
Advantages of Non-Facilities Based ISPs
You can spend your available capital on marketing & sales expenses, which improves your chances of generating new subscribers, rather than sinking it into hardware, which does not increase sales.
Flexibility is high. You can start an ISP overnight and private-label, it ALL through any dialup-access-switch wholesaler. This is great for folks who want to do the new vertical ISP concept we're seeing being introduced this year.
You don't need high-end technical support for dial-up access switches. This can save a lot of tech labor when things go wrong: you just pick up the phone and call your wholesale provider.
Disadvantages of Non-Facilities Based ISPs
You may have to wait for the newest enhancements?for as much as 6 to 12 months after the rest of the market gets them-as wholesalers usually follow fast, but don't lead new technology.
You are totally reliant on someone else to deliver the service you are selling, and your reputation may suffer if they do not care as much about up-time and quality as you do.
Which is right for your ISP?
There is no easy answer to answer this question, as a series of issues comes into play when deciding how your ISP will be setup. These include?
your available capital
how far you want your geographic reach to be
the strategic future direction for your ISP
You may also think about your ISP exit strategy in the course of making the above decision. If you choose to be a non-facilities based ISP, it's much easier to sell to another non-facilities based ISP, who can just take over the Accounts Receivable and Accounts Payable without much disruption for your subscribers.
If you're a facilities-based ISP, and you plan to sell to a deep pocketed local telco who wants to get into the game, they're likely to give you a higher valuation than if they had to set up a new relationship prematurely-or before they were ready to go national (say, if you were a non-facilities based ISP with 200 POPs).
Finding Financing Online
Now you can put your ISPs financial plan together without ever leaving the office. Of course, it pays to do some research.
by Mark E. Battersby
[December 21, 2000]
A number of ISP owners have discovered that the Web has done something comparable to what e-Bay has done for auctions?it has created a global market that gives buyers leverage they lacked in the past.
There are number of Web sites that allow a borrower to complete one loan application that is then reviewed by hundreds of potential lenders, all eager for your business. The Internet offers every ISP owner both advice and solutions.
An ISP seeking to compare the amount of time other credit card processing companies take to credit payments to their accounts from those utilizing their services merely turn to the Internet.
Unhappy with your present bank and the limited number of services it offers? Internet banks are competing for your business and the services that they offer can easily be compared?along with their costs.
? BusinessFinance.com
The "Largest Network of Business Capital Search Engines." Visitors are offered the opportunity to search for capital, download a workbook, or send loan information, etc.
America's Small Business Center
Offers credit checks, insurance, office products, news, business links and, of course, taxes and financing online. Visitors to this site can apply for loans, leases, credit lines, mortgages, etc.
Shop around
The number of financing options available to the average ISP owner on the Internet is growing every day. For those who merely know that they and their ISP business require financing, but aren't certain what type would best suit their unique situation, the Internet offers both tutorials, comparisons between various options and much more.
AllBusiness
Free membership site that offers visitors both the benefits and drawbacks to a number of financing options; and links to sources for that capital.
anyloan.com
Div. of New Century Mortgage Corp., also offer business loans in addition to consumer and debt services.
?
? For those ISP owners who require financing and have a pretty good idea of what type they need, the amount required, the length of time it will be needed, and how it will be repaid?the Internet offers sites that can refer you to specific lenders or investors.
Or, you can turn to sites such as Capital.com, which specializes in arranging loans in the $250,000 to $100 million range for small- to mid-sized businesses.
It is not a lender nor is it a bank. Rather, Capital.com runs a virtual bazaar where small businesses and banks can get together to talk about what it takes to operate a successful business.
Mark Opel, Capital.com chief operating officer, said "we're offering a level of services to small- and mid-sized companies that they usually don't get from banking and finance firms."
The site can also help an owner place a value on his or her business as well as help with buying or selling that business. What's more, Capital.com's clients aren't required to pay a fee?the site makes its money from a fee paid by lending institutions.
? Bankrate.com
Hodgepodge of information on both financing and taxes, timely reminders and, of course, a source of rates for online finance, Internet banking deals and more.
Business.com
Contains an extensive section on commercial finance. Not only does it provide information and links to industry sources, direct links to a number of lenders and financial institutions is provided.
Another site, fairly new, has recently entered the financing market. It's primary purpose? To bring together "angels" with ventures that have unique funding needs. Aptly dubbed Angel.com , naturally this type of funding carries greater risks.
At many Web sites, visitors can fill out a basic application, then use it to apply for a range of loan types. Many sites, especially industry-related sites, have loan criteria from lenders who are associated with the site, so applications for financing can be pre-qualified in seconds.
? GE Business Solutions
General Electric financial solutions site, provides news, case studies and, of course, access to GE's extensive line of financial products and business services.
As already mentioned, many sites only automate part of the process. Loan applications may be submitted and sometimes accepted online, but often require conventional means of transmitting information to potential lenders. So remember, specific arrangements, frequently off-line ones, are normally made to formalize the agreement and transfer funds.
Reliable sources
Every ISP owner should also be concerned about who gets a look at his or her financial information. Virtually all finance-related sites have a "privacy policy."
Remember, the current attrition rate among dot-com companies in general is very high. If a dot-com financing firm fails, all financial statements and applications within their system may become just another "asset" placed on the auction block. The next owner of that "asset" may not be subject to the same privacy policies as the previous operator.
The actual application process, made far safer today by developments in Internet security, may nonetheless, be a step that you would feel safer accomplishing via the mail, Fax or in some cases in person. The experts claim that the entire financing transaction can be safely accomplished online but most systems are flexible enough to offer options for the cautious.
Today's tool
Any ISP owner who ignores the Internet for his or her financing needs is overlooking a tremendous resource. Today, an owner can determine what type of financing best fits his or her needs, do the research necessary to compare rates and terms from a number of sources and apply directly online.
How you take advantage of the tremendous amount of information that is available, the many financing options or the services that continue to grow on the Web is up to you. But to ignore those services or those volumes of information can mean drastically curtailing your opportunity.
Final word to the wise
Because of the number of sites that are available and the relatively short period of time many of them have been in business, it is virtually impossible to recommend any one site. Even those sites established and run by major financial institutions can change radically from one month to another and, of course, may disappear completely after a merger.
Advantages of Non-Facilities Based ISPs
Make Your ISP Facilities-Based or Virtual?
?????Part 2
Yesterday, we examined the Pros and Cons of running Facilities-Based ISPs. Today, we turn the tables and consider the advantages and disadvantages of NON-facilities based ISPs:
Advantages of Non-Facilities Based ISPs
You can spend your available capital on marketing & sales expenses, which improves your chances of generating new subscribers, rather than sinking it into hardware, which does not increase sales.
Flexibility is high. You can start an ISP overnight and private-label, it ALL through any dialup-access-switch wholesaler. This is great for folks who want to do the new vertical ISP concept we're seeing being introduced this year.
You don't need high-end technical support for dial-up access switches. This can save a lot of tech labor when things go wrong: you just pick up the phone and call your wholesale provider.
Disadvantages of Non-Facilities Based ISPs
You may have to wait for the newest enhancements?for as much as 6 to 12 months after the rest of the market gets them-as wholesalers usually follow fast, but don't lead new technology.
You are totally reliant on someone else to deliver the service you are selling, and your reputation may suffer if they do not care as much about up-time and quality as you do.
Which is right for your ISP?
There is no easy answer to answer this question, as a series of issues comes into play when deciding how your ISP will be setup. These include?
your available capital
how far you want your geographic reach to be
the strategic future direction for your ISP
You may also think about your ISP exit strategy in the course of making the above decision. If you choose to be a non-facilities based ISP, it's much easier to sell to another non-facilities based ISP, who can just take over the Accounts Receivable and Accounts Payable without much disruption for your subscribers.
If you're a facilities-based ISP, and you plan to sell to a deep pocketed local telco who wants to get into the game, they're likely to give you a higher valuation than if they had to set up a new relationship prematurely-or before they were ready to go national (say, if you were a non-facilities based ISP with 200 POPs).
No comments:
Post a Comment