PegaSoft Canada White Paper
Draft - 2003
By Ken O. Burtch
So you want to start your own computer company? You've chosen a very rewarding but very difficult task. You've chosen changing the world over making money. You've chosen self-discipline over corporate overlords.
Grand visions are easy to describe but hard to build into reality.
Consider the small town I was born in. My grandfather moved to Canada from Pennsylvania in the late 1800s. Together with the local farmers, he created the "Jordan Fruit" company, which later became the largest fruit exporter in North America. I call his generation the "What can we do?" generation.
My father wanted to best my grandfather. Using grandfather's connections, he became one of the innovators in the fruit industry. He was the first man to use a mechanical grape harvester, the man who invented the boom sprayer, the man who pioneered concentrate spraying when it wasn't cool. I call his generation the "What can I do?" generation.
But in my town, the current generation is the "For whom will I work?" generation. No longer interested in making a name for their country or themselves, they are content to rest on the laurels of their forefathers, coasting on the accomplishments of others. Instead, they want waste their lives in front of a TV, buy lottery tickets and allow the accomplishments of previous generations to corrode from neglect.
Like the story of the grasshopper and the ants, the world doesn't owe you a living. To start your own business and see it succeed, you need “What can we do?” or “What can I do?” people. If the people who want to join your startup are busy watching television at night, they will be too busy to create a future success. You need people hungry to see the dream come true and who are willing to hang onto that dream when times are tough.
The “For whom will I work?” people are not necessarily useless, but they are not innovators or revolutionaries. Cringely suggests there are different kinds of programmers, all of which are necessary for a company to succeed. Smart, beachhead programmers aren't going to be happy in a job doing software maintenance. Maintenance programmers aren't happy when they are developing innovative software with tight deadlines. Some programmers do not have the personality to work in a startup.
Goals are targets that let you know how close you are to accomplishing something. Do you want to form a Linux startup? Do you want to get rich? These aren't goals.
A startup isn't something that works by magic. You first need a concrete goal and then develop a plan to make the goal into reality. An author doesn't decide to write a best-selling book. A game designer doesn't decide to write a best-selling game. You start with a concept you want to express, a problem you want to solve, and you create a plan to turn it into a successful venture. Success will follow on its own.
Goals are targets, not methods nor strategies. You want to get rich to accomplish something else. You want to form a Linux startup to accomplish something else. You may want a million dollars to impress your scoffing relatives (who won't be impressed anyway)--this may not be a good goal, but it is a target. Earning two million dollars before you are 65 years old so you won't have to work in your old age is also a target.
Second, goals like "I want to be rich" are too vague to let you know if you are meeting your objectives. How rich is rich? When is your Linux startup fully-formed? Getting together with your friends once a week to talk about Linux and to design a web site for the end of the year are specific goals. You can measure how well you are meeting these goals and whether or not your expectations are reasonable.
If you set a goal of being "the next Microsoft", you're not dealing with reality. Set goals you can achieve. For a startup, meeting together and working on interesting projects may be realistic and reachable goals.
Plan for the next step when your current goal is achieved. At PegaSoft, one guy said he needed no plans, that he could deal with things when they happened—and when things did happen, he was overwhelmed and didn't know where he wanted to go. Have basic plans to move yourself to the next step. Try an exercise with me. Stand up and reach out with your hands. Can you reach each of the walls in your room? Even for such a simple task, your influence only reaches two or three feet. This is the reality of who you are, a single individual who can't affect the room you are in without walking around. You must have a plan to take you from where you are in to someone who can is standing at the goal. It can be achieved but you must know how you are going to accomplish it.
Planning for things too far down the road will only mean that you're spending a lot of energy trying to plan for things that probably won't happen. This is a sign of worrying too much. Worry about the consequences of the things you are doing today: you can only change today so only think about the future in terms of how it will influence your decisions today. Setting up detailed plans for when you are a 2,000 employee company is not helpful if you're trying to keep 6 people happy.
At the same time, think far enough ahead so that the “next step” hijack your company when it appears at your door. Henry Ford said, "Before everything else, getting ready is the secret of success." Many who start their own companies hope that they will "get rich quick", stumbling into success and hope that everything else will take care of itself. They do not plan for success, sometimes because they don't want to be laughed at by friends, or don't want to feel like it was all a wasted effort if the attempt should fail. Like a football coach that doesn't train his team in plays, and when his quarterback finally gets the ball, he doesn't know what to do with it. If you want to make the most of you chance when the ball comes your way, you've got to have a clear plans of what to do when comes.
For example, be prepared to fill orders. When I released my DrawTools package and it was promoted by Micol Systems, I had 100 copies ready. I sold over half, but they went out the door promptly. Delays in distribution are a headache for everyone.
My father said, "Be true to yourself." If you are dishonest with yourself, you can make no plans or set no goals. To achieve great goals, you must overcome great hurdles.
When starting any business, good people are hard to find. This is certainly true when talking to the people who will lay the foundation of your startup. You can't always pick and chose your partners, but be aware of the risks are.
Some kinds of personalities don't make good founders. As you might expect from the 127 rule, founders should be a "1" personality. Heroes take risks for the greater good. Heroes learn from there mistakes. Heroes are reliable when things go wrong.
You might say, "Yes, Ken, but if I have a contract I'll have many people lining up to join my company." These aren't the people you want for founders. "7" personalities are only interested in themselves: if they don't see a steady advancement, they get discouraged and quit. Founders have to work long hours, forgo pay in exchange for a stake in your company. Some people don't want that kind of risk. When the chips are down, it's no good having the best CEO in the business if he's only going to abandon his partners to their fate. If you have people who are only interested in easy, immediate reward, you cannot form a startup.
An example is the "get rich quick" person. Although two-thirds of university students believe that they will become millionaires, only 1 in 300 actually make it. Many try to start their own company because they want to make money easily. They are not prepared for the dedication and work involved to become successful. At first, they are dedicated, pushing everyone, encouraging everyone...but as their excitement wears out, so does their commitment. That leaves you, the original founder, holding all the problems and debts.
"2" personalities are, of course, a disaster waiting to happen. They see you little startup as something to exploit for their personal benefit, and if the rest of you get broken contracts and massive debts, well, that's too bad for you. They will tell you how much you need them, and how indispesible they are while they look for free office space and make personal calls that the other founders have to pay for. Don't be fooled by destroying angels. Avoid hiring 2's as founders at all costs.
There are more fools than real business men. There are more people who say, "I could have rich" than those who really do become rich. As my grandfather used to say, "close only counts in horseshoes and hand grenades."
The primary difference between the two is a business man knows when to gamble and how much. He knows the difference between potential and reality, and he must be shrewd enough to recognize the hurdles that must be overcome to see a project through to financial success. Shrugging off problems or being reluctant to find someone to test a project under real-world conditions is foolish because that is being afraid to face the faults of the project.
Good people are hard to find. You'll never have a perfect team when you start a company, and the imperfections of your partners will give you headaches. But you must do the best with what you've got--they are your company.
As long as they are loyal and are sincerely interested in success, bear their failings because you know that most people wouldn't show up or contribute for free.
When forming a startup, chances are you'll have some of your friends involved. A "friend" is someone who shares your interests but who you can talk to freely because they are not directly involved with your problems. On the surface, friends seem safe as business partners, at least, safer than strangers. However, when a friend becomes a partner, they become involved in the business and you can no longer go to them freely with your problems.
For example, the same friend you used to complain about unreasonable deadlines with may become someone who thinks the deadlines you impose are unreasonable. If you didn't have a close relationship, this kind of issue would not seem as serious, damaging or threatening.
When making friends partners, there can be a lot of stress on the friendship. Like the movie “Treasure of the Sierra Madre”, a team of friendly people can soon be at each others throats due to greed, deadlines, distrust, gossip, jealousy, envy, backstabbing or other actions motivated by fear. This isn't about friendship: it's about human nature, faults and frailty. Your friends will make mistakes, and mistakes by friends always hurt more than mistakes by strangers.
There was a friend of mine who joined PegaSoft and never did any work and never showed up to any meetings. We talked about it and gave him an ultimatum. He ignored it, and we told him he was no longer a partner. In this case, we gave him a chance to prove himself and it didn't take it. To be a partner, you must shoulder the responsibility of being a partner.
Be prepared to fight, endure, to be patient with, and to forgive the friends you work with if you want your friendship to survive in such a difficult environment.
A startup needs people. The purpose of a company is to gather like-minded individuals and pool your resources to accomplish a common goal that none of you could have accomplished on your own. This simple statement is surprisingly hard to swallow for many in computers who take it as a personal affront if you suggest there's anything that they can't do on their own. A team is always more than one person.
You can't start a part-time company without dedicated staff to carry the weight. Somebody has to tell the customers that you exist. Somebody has to be arranging promotions with other companies. Somebody has to be overseeing the operations and taking care of the finances.
Before you chose a project, first find good people. As Cringely suggests, "avoid stupid and unlucky people. If you are stupid or have bad luck, don't start a high-tech business." If you don't have good people, then you will not be able to accomplish your project, no matter what it is. If you are a poor judge of people, find someone skilled to assist you in finding the best people.
No matter how good the people are, they must have enough time to commit to the project. You need enough manpower to make the company succeed. When all your people are part-time, use a spreadsheet and estimate your resources. Are your part-timers willing to give up television, video games, hobbies and family time? If you're able to spend a half hour a day on running your own company, that means that you are only about 1/8 as productive as you would be if you were working full-time. A short six-month project would end up taking four years.
Be reasonable about what your group can accomplish. Without enough people, you'll end up with angry customers. Computer companies have a reputation for not caring and that's what customers will believe if you're short staffed and provide poor service. People are distrustful of businesses.
Once you have good, dedicated people, and you know what they are willing to commit, you will be in a better position to design your first project.
Destroying angels—people who look honest, valuable or important but who are really a destructive influence on a company—are a threat to any endeavor, including startups.
At one startup I worked at, there was a salesman who insisted that he was a part owner of the company—even though he wasn't. When I asked him why he claimed to be part owner, he said, “Perception is reality.” He believed that to be an owner and control the company wasn't a matter of paperwork or agreement...if he kept claiming to be an owner, over and over, eventually the team would give up and treat him as an owner. If he repeated a lie long enough, the lie would be accepted as truth. If the lie was not accepted, he would pretend to be a friend. If friendship wasn't accepted, he would threaten the team members to do what he said. He was determined to take control of the company from the owner.
The owner was convinced that the salesman was valuable. The salesman promised large amounts of money and important contracts, but the money and the contracts never occurred. I asked the owner why the salesman, if he was such a hot shot, was living in a poor house and working at a startup. There was inconsistency in the salesman's life and his behaviour. But the owner believed the salesman even though the salesman cost him money and dragged the morale through the mud.
Perception is not reality. See reality because a startup is real, a project is real and people are real. Destroying angels work by denying reality. Friendship, threats and empty promises don't mean anything if reality is the person is a leech trying to prove something at the expense of the group. Recognize when you are being lied to by people who don't care about your success. No matter how small or large a company is, a destructive person can cause great amounts of damage.
It's called many names. Back room politics. Someone casually asks you to meet a the coffee shop and tells you they want to talk “off the record”. If you have to do these things on a regular basis, there is a serious problem. You're afraid of getting hurt. You don't trust the people you work with.
There is only one piece of advice I can give about secrets: don't have any. Talk openly about everything. If you meet in a coffee shop, or in a back room, you never know who will overhear you. If something can't be said openly, it shouldn't be said at all.
“Communication increases: paranoia decreases”. [Wired 08/04 p, 124]
All companies need someone driving the company forward. People look to a leader when they are confused or the company is in crisis. Most companies have leaders that the employees don't believe in so most companies will flounder in a crisis.
If you ask 10 different people, you'll probably get 10 different definitions of leadership. Is a leader someone who makes more money than others in the company? Is a leader someone who enforces peace by obedience? Is a leader a mouthpiece to the press to is sacrificed when the company makes serious mistakes? Does the leader have the most voting shares?
If you've read this document, I've discussed many types of false leadership. You probably have a good idea what a leader is not. A leader is not a position or a job. It is not a CEO, manager or administrator. You don't get pay for being a leader. People don't follow your orders without question when you are a leader. A leader is not an empty symbol to calm shareholders.
A leader is someone with the skills necessary to help your group overcome an obstacle. A leader is a servant of the group: the group is not a servant to the leader, but it defers to the leader when the group has faith that the leader uses sound judgment in their area of expertise. Management has nothing to do with leadership. In a baseball team, the manager and the coach are two different people. There's no reason for a manager to be above a coach.
A leader takes action. If someone brags about what a great leader he would make but you never see them do any work, that person is not a leader. A leader must willing to “get their hands dirty” and serve the team, putting their personal interests and comfort aside.
A leader teaches. They explain what they do, accept comments, and try to develop their skills in others so the entire team becomes stronger and adaptable to changing circumstances. A leader doesn't try to prove they are the best nor do they try to stay the best at the expense of the team.
A good leader leads through the admiration and respect of the team.
A bossy leader will usually fail with little or no pay. People will listen to a bossy leader that they don't respect to keep a good wage. In a non-profit setting, people never do as they are ordered. [S p.28] Bossy leaders don't work in startups.
Some startups get bogged down arguing about who will be the leader. There's no such thing as “the” leader. In an army, a general may be the top rank but the general doesn't always lead. When the army is on the march, it cannot cover any more ground that the supply officers can move the supplies. In a medical emergency, the medics make the decisions. The general cannot know everything: he's got to delegate responsibility to the people who can make things happen, and trust them to do their job. When a single person demands to be leader, they are a dictator who doesn't trust the people they work with.
A startup is formed by people of different personalities, interests, and backgrounds. A team exists because a group can accomplish what one individual cannot do alone. Determine what skills sets you have and be humble enough to delegate responsibility.
As DeMarco says, “Leadership is everybody's business”. [p.145]
If a startup survives and begins to grow, there is a danger of creating a clique: a small number of friends who run everything and shut out the interests of the company.
While attending a writer's group, I was asked to upgrade and maintain the group's web site. I was glad for the opportunity to put my skills to use and to use the small amount of money for the position to cover my membership dues and traveling costs. I published my email address and asked the board of directors to send me anything they wanted on the web site. After a few months, I was told that I wasn't working out and that a spouse of one of the board of directors would do the web site.
After some inquiries, I found out that the decision had nothing to do with me. The board of directors of the club always shut out new members from participation. Although there was general elections for the board, the nominees were always selected from the same group of people. When jobs came up, they were always assigned to the same people. As the writer's group grew in size, so did the fear of failure and the distrust of members. Eventually the board of directors elected a destroying angel as the president of the club.
To handle growth, you have to believe in the people who work with you. You have to give people a chance to prove themselves. Where leaders embrace, a clique shuts out. Members of a clique are not leaders. Cliques are formed by fear and the damage is greater than what you are afraid of. If you are not afraid, then there will be no clique.
Talk is cheap, but the people who will really put in the hours and remain loyal aren't always clear at your first get-together.
As an example, when I took the 3rd year project course at Brock University, our team started out with six members. One member began to crack under the stress and began to claim that one person after another was incompetent and the team leader told each that they were "out" of the team. Finally I received a phone call from the team leader who said, "Ken, I'm concerned. I'm the only one left!".
Loyalty and team work are difficult to plan for and difficult to evaluate. Nevertheless, these are important qualities to the success of any project. What good is a skilled developer if he's not loyal or if he tries to sabotage the team? Loyal people must be rewarded. I was once told vehemently that I had to "kick out" one of the members of PegaSoft because he was incompetent with money. It was true that this guy had serious problems dealing with financial issues. He came from a poor family and wasn't experienced with managing money. But it was also the longest standing member of PegaSoft (besides me), he showed up to virtually every meeting. I could live with his financial ineptitude if I could count on him to weather a crisis.
If a man turns $100,000 to $200,000, people will say he is a financial genius. If that same man turns $1 into $2, nobody cares in spite of the fact he accomplished the same feat. Financial skill and a good product is not enough: you need money in order to create money.
As a farmer's son, I grew up understand the value of a business. When my father remarried and my stepmother saw the annual cheque from the winery for over $100,000, she announced that they should go out and buy a luxury car. My father rolled his eyes and explained that this single cheque had to pay all the employees, all the equipment and pay off the bank. $100,000 minus $100,000 is zero dollars profit. It hasn't been a good year.
The size of the profit doesn't matter when the size of the bills were equally large. To hire several people, rent a building and equipment and pay the bills until the first project is complete can cost several million dollars—a large amount of money for a person, but a small amount for a business.
A software company doesn't need a lot of money to get started—many people will work for the fun and a share of the profit—but it needs some money all the same. Steve Jobs sold his VW to get money to start Apple Computer. Without that money, they wouldn't have been able to get started.
At PegaSoft, we charge a small yearly fee for membership. Someone pointed out that a lot more people would join my company if membership was free. Anybody who won't pay to come to a meeting doesn't believe in the company--do you really want people like that freeloading while other members pay all the costs? A company has expenses, even if it's just to take out newspaper ads to try to get new members to come to meetings. And if one person gets a free ride, that means the others have to pay for the cost. And that's not fair.
Some startups try to undercharge for their products to compete. They don't pay their employees and work long hours to fill orders. Afraid to turn a profit, they give away their efforts for free. I met a woman running a small-time Internet business. She said that she was desperately trying to turn a profit, getting up at 2am in order to fill orders and get them ready for the mail. With so much demand, I asked her why she was undercharging. With more profit, she could pay people to help her and get her business truly operating. She scoffed and said that if she charged more, her customers would leave her. Giving stuff away for free is not running a business, just the illusion of one.
An author recently said to me that she once calculated her time, money and materials and found out that she was selling her writing services for minus $7 per hour. Although she was no math whiz, she knew enough that if she didn't charge enough for her services, her writing career would never get off the ground.
Success means that you have to believe in your idea, company and team. It's not like sports stars on talk shows that say that anybody can get into professional sports. Timing and dumb luck are as big a role in success as any other factors. Tech companies should plan for failure because any venture means many risks will need to be taken. [S p.188] There's no such thing as a “sure thing”.
Never be so over-demanding that your team doesn't enjoy themselves. Don't be too quick to reach out and blame the team for a first failure unless you're sure they're the problem. Enjoy your startup.
Cringely says that startups fail because of "stupidity, bad luck, and greed". Stupidity is not thinking things through. Bad luck means obstacles arise that were beyond your control. Greed is more difficult to deal with.
Does anyone have enough money? No matter how rich you are, there always some goal just out of reach if you just had a little more money. Greed is a deception: to be content, you have to be content with what you have now and make the most of your opportunities today.
I've met many who single-mindedly pursue money with no clear idea of what they wanted. I had a heart-breaking conversation once with a professor who was retiring, and he said that for decades he was so concerned about providing for his family that as he looked towards retirement, he didn't know who his children were: they were like strangers in his house.
Greed is a bad reason to start a computer company. You want to be rich? Work at a gas pump and save money every week.
Greed is a bad motivation for a computer company. It's not reasonable to expect success from risky ventures and volunteer help—at least, not for many years.
As early as possible, determine how shares in the company will be divided up. If your company members angrily argue over shares when there is no profit, imagine how much they will fight when there's actual money to distribute.
I live in a small town of 150 people. The main industry is farming. Why would I try to start a computer company here? Will I find interested developers out of 150 residents? Will I find contracts or investments or good Internet service? Of course, not.
For many living in urbanized countries, location may not be something that you think about. But if you're living in a small town, be reasonable about what you can accomplish. You can find better team members if you have a larger population pool to draw on.
I hate driving and I hate smog, but I hold PegaSoft meetings in the city of Toronto—two hours away. It's difficult to meet with venture capitalists or interested businessmen who tend to live in a big city and went to meet face to face. If you live in the country, be prepared to spend money on gas and a hotel in order to meet with the people who can make your company a success. Don't expect people to drive to your town if you won't drive to theirs.
A successful startup requires people of many different talents, but there are limits of what you can accomplish on your own. If you are interested in selling commercial software, it is difficult to get your products on the shelves and exposed to the public. In the early days of the Apple II or IBM AT PC, there were many hobby magazines with cheap ad space. Now to get exposer in a major magazine, even a Linux one, costs thousands of dollars.
Somebody must be selling the company, promoting your stuff. You must contact other companies, distribution companies, etc. to make a profit. Don't start out by trying to sell to a countrywide chain like Wal-Mart--instead, focus on companies that are smaller, who can take you seriously and see the value of your project for what it is. Use that credit to sell yourself to larger institutions. Don't start with a national company and then give up in despair when you discover they are not interested.
** Enron got into the news in 2002 when it inflated its earnings and bid on contracts that it couldn't fulfill. Linux advocates made bold predictions of Linux on the desktop. But both were foretellings using impatient lies and empty bravado.
Contracts and Consultants
Everybody thinks they can build "the better mousetrap", yet most computer startups fail. Finding good projects is easier than finding good people. Do you have a clever idea for an original software project? Why only one? Thomas Edison said, "I am long on ideas, but short on time. I expect to live to be only about a hundred." [BR3]
There are thousands of possible projects to undertake: try to look at more than one alternative to find one with the best chance of success. According to an “I, Cringely” aritcle, one estimate at the success of an Internet-based Application Service Provider is 1 in 50. [Cringely, ASP] Those are not very good odds—choose a project that presents the best opportunity.
Work smart, don't work hard. If you want to build the world's tallest building, make sure that the office tower under construction down the street isn't scheduled to be 10 feet hire than yours. Never bring a knife to a gunfight. Never take on a project that competes directly with a company larger than yourself. Try to find a niche that you can exploit one that other companies have overlooked. Orville Reddenbacker, the popcorn millionaire, chose to make his money in popcorn precisely because there was no competition. Everybody wants to compete where the money is, but that's greed not shrewd strategy. Maximize the odds of success by outmaneuvering people looking for quick dollars.
Focus your efforts: if you start 10 projects and don't complete any of them, all 10 are worthless. Pool your resources on a select few and if they fail, move on to others. If everybody does their own thing, then you don't have a company.
If you are going into business to change the world, be focused and organized. Don't try to change everything at once but chose to follow a series of projects that lead up to your ultimate goal. When a project has a simple, clear message, people are more likely to hear it than any radical "manifesto". Only grind one axe at a time. GNU would never have succeeded without GCC first.
Since quality work is necessary, don't do too many projects simultaneously. Pool your resources on a select few and if they fail, move on to others. If everybody does their own thing, then you don't have a company. DeMarco says, "Make less and choose more carefully what you make." to improve quality. [S p.119] Unless you have a solid foothold in one area, doing projects in other areas is a lack of vision. [S. pg. 41]
Some people deliberately chose projects that will fail.
If you are going into business to change the world, be focused and organized. Don't try to change everything at once but chose to follow a series of projects that lead up to your ultimate goal. When a project has a simple, clear message, people are more likely to hear it than any radical "manifesto". Only grind one axe at a time. GNU would never have succeeded without the focus on GCC first. To give a project the best odds of success, minimize the risks that will cause it to fail. [S. p.191]
Don't try to fool customers with a bad project. With an Internet-based project, you might be tempted to hack together some badly written code and put a good looking web page in front of it. Cringely believes that this will never work for a computer startup. "Success in a large organization, whether it's a university or IBM, is based on appearance, not reality" but in a computer startup "all that counts is the end product." For an Internet project, having good software is essential since that's all you are selling. If you think sloppy work is a shortcut to success, think again. A reputation of bad workmanship is difficult to repair.
Scacchi quotes a ITT Advanced Technology Center report that the median size for a software project is 22,000 lines."
When To Try Something Else
Since most projects fail, be smart and set benchmarks. The more you work on a project, the more you will want to see it through even if it's a failure. Instead, if a project is not meeting the benchmarks, be prepared to switch to a new project with a better chance of success. Cut your losses before you dig a pit you can't get out of.
DeMarco recommends evaluating the associated risks:
1. List and count each risk.
2. Have an ongoing process for discovering new risks.
3. Quantify a transition indicator for each one that will tell you that the risk is beginning to materialize.
4. Designate a transition indicator for each one that will tell you that the risk is beginning to materialize.
5. Set down in advance what your plan will be to cope with each risk should it begin to materialize.
[S p. 200]
The initial release of a product should be as rock-solid as possible. Subsequent releases can tolerate some bugs in exchange for timely release of new features. For example, Netscape 6 was essentially a new product, but it was released prematurely with bugs and it died.
Open Source is not a cure-all and can sometimes have bad results. When IBM created built their first Intel-based PC, they build their PC from off-the-shelf parts. Their openness opened the way for cheap clone machines that could be built by anybody. Although the architecture chosen by IBM went on to dominate the industry, the industry wasn't dominated by IBM. They quickly lost their dominance and people lost their jobs. On the other hand, IBM licensed away the CPU to Intel and the operating system to Microsoft. losing control of the core technology to their own dream. Poor licensing made the 1980s IBM PC's a sort-term success but a failure.
If you are going to spread a way of thinking in a guerrilla software war, you need to keep your standards open. As the lesser market holder, you can't force people to use your software or standard by peer pressure. You have to make it easy to play with it, test it and adopt it. It's not the place for licensing fees. Apple was considering licensing it's GUI back in 1985. If it had, Apple may well have become the industry giant instead of Microsoft. Instead, they alienated themselves in the marketplace.
Because of contractual obligations or cutting edge technology, closed source licensing may be an option. Part of Linux's success it that lots of people know UNIX, so they duplicate what they know. However, innovation and new ideas are difficult to get moving in an open source project because people are lazy and they don't have a unique vision.
Don't get greedy with your licensing agreements and shut out the little guy. There was the case of a Nintendo developer who created an updated version of the game Ultima I for the Apple IIgs, the last computer in the Apple II line. He decided to do the right thing and get permission from Origin Systems and immediately received a bureaucratic wall of licensing agreements and up front payments. In the end he managed to get the email address of the company's founder and the author of Ultima I who told him to go for it as long as he sent the author a copy of the program so he could try it out on the Apple IIgs in his office.
Think about the nature of your work and what you are trying to accomplish. Use an appropriate license for a project.
Software takes a long time to develop, and in a short time is worthless. Know the lifetime of your software and adjust your prices accordingly.
If your software is obsolete and you are no longer maintaining it, release it as open source. Those few customers who are forced to use your unsupported software will appreciate it. If you hold onto the source code despite the fact you cannot turn a profit, you're not looking accepting the reality of your software. People want to avoid humiliation: by holding onto outdated software, they won't have to see their mistakes if the software becomes successful under open source. But if you've done everything you can reasonably can to support the product, if it's released open source it will probably be a dud.
Mendacity can play havoc on someone starting out a new software company. When there's a lack of trust in a small group, communication breaks down and the members become discouraged. Somebody who doesn't pull their weight will not likely leave quietly, but will spread rumors, blame others, try to break the group apart and steal in revenge.
Small groups are reluctant, therefore, to formulate grounds rules for membership. But rules are necessary to both outline expectations, goals and for discipline. When you are your own boss, there is no one to cover for you, no room for lazy or dishonest people.
Some believe that an absence of rules makes the group happier because there's nothing tying members down. No rules means one rule: "might makes right". There is no such thing as a company that has "no rules"
An example of this is the writing club I belong to, the Writer's Circle of Durham Region. Not wanting to impose rules, the Writer's Circle had breakfast meetings where attendance was optional. But as the size of the club grew, more and more members joined who only made it out sporadically to meetings, forcing the club to pay for uneaten meals. Eventually the club had to lay down rules regarding attendance.
When my house was broken into, the police laughed at me. And it's a well-known fact that the cost of legal fees outweighs any justice so that the justice system has become an expensive way to defer blame or bankrupt the little guy. But, after all, the Bible warned about this a long time ago: why approach a human court that owes no allegiance to you or your principles, a court that cannot be trusted, to resolve your own personal matters. We live in the world, but we are not a part of it. It's up to you and your partners to handle these issues. You can't outsource justice.
Don't expect justice to be served by outside institutions. When the pressure is on, reality seems to warp. A case I had with my 4th year Computer Science project at Brock University. Under the pressure of my project demonstration, I blurted out that the project wasn't completed because the professor didn't warn me how big it would be. Later, he asked why I said that when he actually told me that it was too big. Ashamed and embarrassed, there was a part of me that wanted to believe that he was to blame and not me. Ever since then, I write down everything I can and verify the facts since I don't want to make the same mistake again.
Although rules and goals are essential for giving a sense of what is expected, too many rules or conflicting goals are the same as having none at all. Can anyone say that they know or understand all the traffic laws? All rules and policies must be reviewed at set times or have limited durations.
Venture Capital and Investment Banks
Chose a project that falls within the skills of your team. For that reason, Cringely advises startups to "Do a product that you want to do." That is, don't let venture capitalists make the decision. Stay in your expertise and use the talents that make your startup unique from other companies. Don't take on impossible but impressive looking goals.
If you invent the cure for cancer, you don't get venture funding. If you invent monkey angel dolls, VCs will back up a truckload of money to your door. Don't expect to get funding for good projects because VCs don't fund good projects.
To buy a fast car and get even with people who laughed at you in school is one of the stupidest reasons to start a company. School is over: get on with your life. Never let the past make your decisions for you. Instead, design your company for the future, not the past.
Impress people with your contributions not possessions.
Another Cringely recommendation to startups is "Don't take venture funding too soon." Use it only to launch a product.
Most computer companies don't need a lot of financial backing. Microsoft only went public to increase its own wealth.
A venture capitalist wants to turn a fast buck on an IPO.
How to spend the money. One consultant said he would buy a building in North York (a wealthy area of Toronto) to make his enemies jealous. First, enemies don't get jealous. Second, how much is revenge worth? Why not instead purchase property in some of the beautiful recreational areas around Toronto? Or use the money to buy better equipment? This guy wanted to rape the company money for selfish revenge instead of contributing to the success of the project.
Apple lost it's "look-and-feel" court case against Microsoft when it signed a contract which gave Microsoft "a non-exclusive, worldwide, royalty-free, perpetual, nontransferable license to use in these derivative works"--essentially saying Microsoft could copy from the Mac at will now and forever. The original contract was only for the Windows 1.0, allowing Apple and Microsoft to review their progress and draw up a new contract.
Investing your company in a business deal with a larger company can result in great profits as it did with Microsoft, or massive failure as it did with Apollo Computer. Apollo, Sun's main rival, was decimated when it sunk resources into a deal with Apple only to have Apple pull out and leave Apollo with big debts and lost market share. As they say, sometimes a business deal isn't worth the paper it's printed on.
Not all business negotiations have to be hostile. When IBM and Apple developed the RISC processor for the PowerPC line, they brought in Apple's former chip supplier Motorola so as not to cause hard feelings.
Why are there buyouts instead of alliances? For one reason, mendacity doesn't allow you to deal with other companies for fear they will rip you off.
According to Chris Crawford, famous game developer, creative solutions come from thinking about diverse things. If you think only about the competition, you'll only be able to clone what they have done—that's where your mind is. To have original ideas, you need original ideas and basic research. Read or research areas outside of computers. To create a new space game, think about molecular biology or ancient Egyptian warfare: look for patterns that can create a different, interesting game. If you research your space game by playing existing games on the market, you'll only reinvent them...maybe with better graphics or sound, but it will still be an unoriginal copy. [Crawford]
Innovation is always inverse to production. If your staff are focused on producing a product with strict deadlines, they won't be able to divert time and energy to research creative solutions. Likewise, if you staff only work on creative solutions, they'll never divert enough time and energy into the tedious work of testing and completing a usable project.
Marketing an innovative product is always difficult: since it is not familiar, people will be afraid of it. They won't know what to compare it to and most customers don't want to take the time to explore and test the benefits of an unknown product. When shopping carts were first invented, shoppers didn't use them: women who pushed baby strollers all day didn't want to push things, men didn't want to appear too weak to carry a basket. In addition, grocery stores were afraid of being sued because of kids playing with them. It wasn't until the inventor of the shopping cart created a movie showing shopping carts in use that the fear of store owners was reduced and they began ordering them. [AABR, pg 262]
Because people are afraid of original projects, create your web site or sales information accordingly. The people visiting your web site will be reluctant to be there, so make it as easy to read and navigate as possible. The people will be afraid of wasting time testing your product so provide easy to read summaries, pictures, samples, case studies, review quotes or walkthroughs—you want them to see how the software will benefit them with only a few seconds of surfing. If they leave your web site, the probably won't be back.
In commerical software, Crawford has pointed out that the retail software business is designed to inhibit innovation and shut out small companies. In Crawford's example, suppose you create a really great commerical game aimed at housewives. You have to convince a publisher it's worth the risk to produce the game. They have to convince a distributor to carry the game. They have to convince retailers to put it on their limited shelf space. Even if you can get past all these difficult obstacles, housewives (your target audience) don't visit game stores. [Crawford]
If you sell software through your web site, it's difficult to charge money for an easily accessible, easily copied product. Software, before the Internet, was a sellable commodity because the customer was not paying for the software but for the delivery of the software on a physical media. With the Internet, there's no reason to pay for physical media, but software is still sold by the purchase of a physical thing. If you're writing a game, the game is not sold by itself--anybody can pirate the game. It's sold by the rewards of purchase--the decals, T-shirts and cloth maps and parchment manuals. They are the rewards of purchase and the proof of membership.
In the opening class on economics, the professor said that the purpose of a business is to make money. He added that there were other motivations, but to earn money was the root of it. Unfortunately, many businesses didn't hear what the professor said. Making money was the only thing that was important.
The purpose of a company is to band together to provide useful goods and services that could not be done by ourselves. Money is the unit of change to trade one form of goods and services for another.
In today's world, businesses are far more than places to earn money. With the advent of television, work becomes the place where people socialize and the odds are high that you will marry a coworker because most people cannot except the workplace. To some, work becomes a way of life, and a religion.
When Apple's System 7 was behind schedule because of too many new features, the developers were pushed to work 7 days a week and 12 hour days leading to nervous breakdowns and divorces. [Apple pg. p183]
The team slice consists of the programmers and the project leaders. It does not include marketing personnel, accountants, etc.
The support slice consists of all staff directly supporting team efforts: marketing personnel, researchers, graphic designers, etc.
The administration slice consists of the remaining staff who are necessary to run a successful, accountable business.
As a result, there is no CEO, CFO or CTO as used in traditional business models.
A Founder is a person who joined the company before January 1, 1999 and who contributed work to the company. Existing Founders can add new founders before January 1, 203 if the person is unanimously elected. Founders take responsibility for startup debts from venture capital angels.
Ken receives a double portion because PegaSoft he started the company, providing the corporate framework and much of the initial operating money and labour.
Before expenses, 15% will be taken off the top per annum as the Founder's portion, to be split evenly amongst the surviving Founders. In the event of the deaths of all Founders, the Founder's portion will be eliminated. Note that this is 15% form the company, not from the individual projects.
Startup Compensation: 50% after expenses will be split amongst the founders while the company remains 10 members or less.
PegaSoft can choose to recognize people inside or outside the company by making individuals "PegaSoft Fellows". Employees of more than 10 years qualify as fellows, and this status stays with them even if they leave PegaSoft's employment. Fellows enjoy special privileges to be determined.
The terms of losing fellowship status is to be determined, but must be an exceptional case and be by unanimous decision.
All Founders automatically enjoy Fellowship privileges.
Fellowships are to be awarded on a yearly basis.
Traditional hiring practices fail to provide quality, capable people.
Traditional rule and discipline systems neither enforce justice nor provide conflict resolution. Common Law such as used in countries like Canada is impractical and raises other problems. Systems, especially complex ones, do not provide justice. Yet without some agreed-to rules, there's no way to resolve conflicts and track problems.
The simplest method to combine tracking, discipline and accountability is a demerit system. A demerit constitutes a "verbal warning", but has less weight.
[Choosing people for discipline committee.]
This is an outline for the awarding of demerits, bonuses and avenues of appeal.
Demerit are loosely considered verbal warnings. By treating them as points, they can be tracked and reviewed. Demerit points are accumulated over a two week period. Demerit points can only be submitted once a day for a particular offence.
Three demerit points in a two week period, or two consecutive periods with the same offence, results in an Hours Punishment--work overtime for two hours without pay, or lose two hours worth of money from the next paycheck. Announcement of hours must be made verbally or in writing. People receiving hours must be posted to the company but the offence doesn't need to be listed.
Three consecutive periods with Hours, or four Hours in six months, the employee will be brought up for an investigation of the cause and more severe disciplinary action. Announcement must be made in writing. Only when punishment is handed out is the results announced publicly.
You cannot submit a demerit for not giving a demerit to someone else. [?promotes washington two-step?]
One person from each slice must be appointed to submit demerits for a particular person who doesn't have a direct boss, such as a slice leader. Everybody must be watched by somebody.
All Hours Punishments or more severe disciplinary action must be announced to the company.
Example: Bob arrives half an hour late, he receives a demerit. If he arrives half an hour late from lunch the same day, he cannot receive another demerit for being late. If he's late 3 mornings in a two-week period, he receives an Hours Punishment.
All serious infractions should be brought before the discipline representatives(s).
In general, infractions should be handled internally (i.e., outside of a secular court). Apologies made to the people involved negate punishment. Reimbursement for damages must be made, or if damage is not monetary, service to the plaintiff. All reimbursement must be done by the defendant--that is, no automatic paycheck deductions.
The recognition of outstanding work is done over the same two week period as the demerits by awarding bonuses. Like demerits, bonuses can only be submitted once a day for a particular action. The can only be awarded by someone's direct boss. They provide a way of tracking good behavior for yearly reviews.
Three bonuses in a two week period, or two consecutive periods for the same action, results in an Hours bonus--two hours off without pay or two extra hours on next paycheck. No more than one bonus hour can be awarded per two-week period.
Bonuses do not cancel demerits, or vice versa.
Bonuses cannot be challenged by other employees, but an Hours Bonus can be appealed.
At the end of the year, every four unused bonuses can be "cashed in" for an Hours Bonus. A year with no Hours Punishment automatically awards an employee with an Hours Bonus.
This system avoids the red tape associated with most bonus systems. It provides away to track outstanding work for yearly reviews. It provides a level playing field for the awarding of bonuses.
Startup Compensation: a founder can award an hours bonus to anyone.
Annual Closures
Apple Computer had Christmas closures once holiday orders were filled. PegaSoft will be closed from the week before Christmas until the third week in January.
Startup Compensation: The company can work through the winter closure if the founders unanimously agree to it and the company is 10 members or less.
Clothes must be in good condition (no rips or tears). No novelty T-shirts. Clothing must cover from the knees up and include the chest.
The standard probationary period is six months. All new employees must go through a probationary period. Anyone returning from a suspension is subject to the standard probationary period.
All employees start with two weeks vacation (10 working days off). +2 after one year, +2 after three years, +1 after 5 years.
Company closures do not count as vacation days.
All vacation days are floating but must be approved by a person's boss and personnel.
Personnel is free to make changes on individual cases.
As a flat project driven company, most hours should be floating. Personnel will determine the minimum staffing requirements for each department under normal business hours.
The normal workday is 7.5 hours.
Because of the company structure, salaries tend to be low. Developers are compensated by having a share of the rewards of a project. Seniority will typically be reflected in increased profit share.
Social and Recreational Responsibilities
PegaSoft will supply recreational facilities for on-site employees. Employees are free to use these services at any time provided it doesn't interfere with assignments. [Microsoft]
The annual PegaSoft summer retreat is to be held on a summer weekend with all employees and fellows invited. This is considered a working retreat, with employees being paid and attendance mandatory, but it is also to be a fun event for their families.
The personnel staff are expected to organize parties, fund raisers and recreational events throughout the year.
Use of Company Equipment/Supplies
Employees are allowed to use $7.oo of supplies and equipment per week for personal use (e.g. photocopying). Employees are not to be monitored unless there is evidence of abuse. Telephone and Internet access is unrestricted unless it interferes with duties (earning a demerit).
All development teams are considered to be under the team slice. The administration and support slices, in the company model, are considered to be supporting the team slice.
The team slice is overseen by a team slice leader.
The research pool, that is programmers doing research in the area but not actively developing a particular project, is also under the team slice.
Tech support in the company is to be provided, whenever possible, by the teams. This is to avoid the situation at Brock University were the costly and fruitless rivalry between computer services and the computer science department.
PegaSoft has two classes of projects that a team can undertake: major projects and minor projects. Each type has specific goals in terms of time, profit, and team commitment. These goals are used to determine what kind of project it is, and if the project is worth undertaking.
A major project must meet the following goals:
- the time between the start of development and the release of the program must not exceed two years
- the program must be marketable--that is, you must know who might buy your program and why
- the estimated gross profit must be at least $50,000 per team member, or if this is a major upgrade, the previous version must meet this requirement
- there must be at least two programmers on the project
- the team must be prepared to offer technical support and minor upgrades for their program for two years