In many industries, there is a non-stoppable trend towards larger and larger companies.
In retailing, Wal-Mart is crushing their competitors, who are also extremely large but they do not have the economies of scale that Wal-Mart. (In addition, Wal-Mart has better management.)
In some cases, large companies succeed because consumers want the perceived value of purchasing from a large company. Dell and HP basically own the server and desktop computer market, and there is no reason to assume this will change in the next ten years. People feel that Dell will not go out of business.
One of the industries hurting most right now is the automobile industry. GM and Chrysler say they will run out of money unless the U.S. government provides several billions of dollars in aid soon. Ford says it does not need government financing but it has pledged all of its assets in a collateral-based loan. Yet almost no one is willing to buy a car from a small automobile company, they are viewed as risker than GM or Chrysler.
In choosing software companies, one factor sophisticated purchasers look at is, "How likely is this company to be in business 5 and 10 years from now?" When you commit to a software package, you are making a huge investment in time and energy, far beyond the purchase price of the software. You want to know that in the future, the software company will continue to enhance the product and will be around to provide technical support. This is why companies, for example, purchase software from Oracle, even though Oracle is ridiculously. People expect Oracle to be in business ten years from now. Yes, open source to some extent mitigates this risk, but putting aside very successful open source projects (Linux, Apache, MySQL, how do I know that they will be around ten years from now. Do I really want to writing the code myself?)
Law firms have consolidated tremendously in the past decade. Large clients want one firm that can provide experise in dozen of areas of the law and can handle complex transactions that span the globe. In Boston, ten years a 100 lawyer firm was considered to be a large sophisticated firm. Now such firms (at least on the corporate side) have merged with other firms to create 500 attorney firms.
In public accounting, the Big Eight has consolidated into the Big Four. Clients want the brand name that only a Big Four firm can provide.
There is no clear trend here. In some industries, consolidation is taking place and will never stop. In other industries, it is the opposite effect.
As noted above, the VC firm Charles River has a seed fund that invests $250,000 per deal:
www.crv.com/AboutCRV/CRVQuickStartFAQ.html
This is a brilliant idea! Ycombinator has been a true innovator -- they saw a gap in the capital markets, they had the guts to go out and fill it, now a lot of guys are copying what they do, but they will always enjoy first mover advantage. Charles River is doing the same thing, filling a gap that everyone knows exists but no one so far has actually filled.
One of my reactions to the Ycombinator standard approach -- $15,000 for 3 months of funding -- is that 3 months is a very show time frame. I am pretty good at what I do, I frankly have a lot more experience than the kids who get Ycombinator funding, I typically work about 70 hours a week, and I know I am quite productive at what I do. I don't go for home runs but at the same time, I don't go for singles. I go for doubles, that is a risk/reward range I am comfortable with. I find it takes about 12 months to get it running reasonably well, and I have done it several times before. I would not want to work on a 3 month time schedule. If I did, I would be going for singles and singles just are not that interesting.
To me, a single is a deal where you end up with $1 million cash out after two years. A double is $5 million after two or three years. A triple would be north of $10 million, no later than four years. Anything north of $25 million is a home run. Anything above $100 is a grand slam home run.
I do not need financing, I just fund them myself, but if I did, the Charles River deal would be much more appealing. Assuming I needed the money, with $250,000, I could get two partners, we each pay ourselves $60,000 a year, perhaps burn up another $50,000 in expenses in one year, for a total burn of $230,000,leaving us $20,000 margin (admittedly, not a lot). But in one year with two really good partners, we could do a hell of a lot of work, perhaps a triple rather than a double.
I really like what Ycombinator is doing but so many of their investments seems unambitious -- two smart college students who whipped something up in a month. Perhaps they might build a company from that, but in some cases, I am skeptical.
My comments relate to Graham’s Cliff Notes summary of his essay, rather than his underlying essay.
Every profession has a pecking order. The fact is some people are better at doing X than other people are. There are at least four factors that could determine where one sits in the hacker pecking order:
1. How smart you are
2. What you have contributed
3. Do you have the guts to start a start-up?
4. If you do, were you successful?
On that score, Graham does very well in the pecking order. He’s clearly brilliant. He has made numerous contributions -- spam filters, author of the two leading books on LIPS, ARC, just to name a few. He had the guts to start a start-up and he was successful, having sold it to Yahoo. The only thing he is missing is being co-founder of, say, Google or Microsoft or Apple or Adobe.
Graham says people think his group are elitists, implying the opposite, that they are not. Of course they are elitists! But why is that bad? Graham does not like Java is that it was designed with training wheels, to prevent programmers from doing stupid things. That may be necessary for some programmers, but it is not necessary for Graham and his target market -- the very top programmers, who do not need training wheels, who do not need to have the language prevent them from doing stupid things. (Have you ever met a LISP programmer who was not very very smart?) His primary partner, Robert Morris, is a professor at MIT. These are two uber elitists.
Graham likes to write essays that challenge orthodoxy and he is very good at doing so. See, for example, his essay “Mind the Gap,” in which he argues that a higher rate of income inequality may be a good thing rather than a bad thing (www.paulgraham.com/gap.html). It’s hard to think of a more controversial proposition than that. So me thinks that Graham’s next challenge-conventional-thinking essay should be why elitism is good, not bad. And he does not have to start from scratch, he can read “In Defense of Elitism” by William Henry (http://www.amazon.com/Defense-Elitism-William-Henry/dp/03854...).
So why does Graham want to portray his group as non-elitists even though they clearly are? Because elitism has a bad rap. Some elitists are smug and arrogant but I do not think Graham is. I am pretty smart, or so I have been told. But I do not think I am as smart as Graham. When I read Graham’s essay, I feel, “OK, this guy is smarter than me, but he is not arrogant about it, he is laying out his arguments in a clear concise logical manner so that others, almost all of whom are not as smart as he is, can follow his logic.” That is not arrogant.
Some people criticize elitists if they do not give something back. Graham is clearly giving something back. His essays are free. If you don’t want to buy his book, you can read all of them (I think) online for free. He is giving ARC away for free. Although Ycombinator is hardly non-profit, I am still shocked at little equity they take (an average of 6 percent) in exchange for what they provide.
Graham argues that Ycombinator would not exist if all it did is fund “rare geniuses.” That is true but misleading at the same. I have been an entrepreneur all my life, so far successfully. The thought of working for the man, I would just jump off a bridge and die. But starting and running a company is a bitch, even for someone like me who is genetically programmed for it (and for nothing else). Our society glorifies entrepreneurship, which is a sign that our society is more advanced than others. In one of his State of the Union addresses, President Reagan urged every schoolchild to consider starting a business. The fact of the matter is that few people are suited to starting a company. You have to be good at a whole lot of different things; most people at best are good at one thing. You work all the time. There is currently a best selling book, The Four Hour Workweek. Yeah, right. Try the 14 hour work day.
Few people are suited for this, and that includes gifted hackers. Few people have the balls to start a company and that is a good thing -- perhaps 2 percent of the population is suited to being an entrepreneur, I doubt it is as high as 5 percent. Perhaps among hackers, or gifted hackers, the percentage is higher, but I can’t imagine it is 10 percent. Most people are better off working for someone else. That fact, by the way, creates a lot of opportunity for the few who are suited to start companies.
Ycombinator’s thesis is that there is an untapped market of young entrepreneurs who could start companies but do not and will if they are funded by Ycombinator. Yes, there is such an untapped market, but it is not as large as Graham implies. Let’s assume Ycombinator funds thirty companies a year, twice a year, and each company has three hackers. That’s 180 people. A rounding error of a rounding error, since every year more than 4 million people reach whatever age is the minimum age for starting a company.
As far as I can tell, Ycombinator has been successful. There are several reasons for this. There were the first to market, that always helps, people think of them first. Graham’s awesome reputation and his Web site are extraordinary marketing machines, so they get the pick of the litter. On the other side are several brilliant people with a variety of backgrounds who are not virgins, they actually have done it before; in short, they have the perfect background to pick the right people to back. So Ycombinator may not limit its investments to ideas proposed by “rare geniuses,” but the caliber of the people they back is so close to genius it may be hard to ascertain the difference. More than anything, Ycombinator has been successful because they are able to attract an extraordinarily talented group of applicants from which they choose a few to back. If Ycombinator started funding the “average” hacker rather than rare geniuses or close to it, I guarantee Ycombinator would be a flop. Graham’s statement that they are not limited to funding “rare geniuses” is a misleading as the Dean of Admissions of Harvard or Stanford stating that everyone and anyone should apply for admission to those universities. The fact of the matter is that it is damm hard nowadays to get admitted to Harvard and it is damm hard to get funded by Ycombinator, even if you are really really smart.
I find it totally confusing as to how to post a comment. You really need to have some instructions, it is completely unobvious. I ended up posting the same thing three times because the comment never seemed to show up and then deleted the second and third ones.
My comments relates to Graham’s Cliff Notes summary of his essay, rather than his underlying essay.
Every profession has a pecking order. The fact is some people are better at doing X than other people are. There are at least four factors that could determine where one sits in the hacker pecking order:
1. How smart you are
2. What you have contributed
3. Do you have the guts to start a start-up?
4. If you do, were you successful?
On that score, Graham does very well in the pecking order. He’s clearly brilliant. He has made numerous contributions -- spam filters, author of the two leading books on LIPS, ARC, just to name a few. He had the guts to start a start-up and he was successful, having sold it to Yahoo. The only thing he is missing is being co-founder of, say, Google or Microsoft or Apple or Adobe.
Graham says people think his group are elitists, implying the opposite, that they are not. Of course they are elitists! But why is that bad? Graham does not like Java is that it was designed with training wheels, to prevent programmers from doing stupid things. That may be necessary for some programmers, but it is not necessary for Graham and his target market -- the very top programmers, who do not need training wheels, who do not need to have the language prevent them from doing stupid things. (Have you ever met a LISP programmer who was not very very smart?) His primary partner, Robert Morris, is a professor at MIT. These are two uber elitists.
Graham likes to write essays that challenge orthodoxy and he is very good at doing so. See, for example, his essay “Mind the Gap,” in which he argues that a higher rate of income inequality may be a good thing rather than a bad thing (www.paulgraham.com/gap.html). It’s hard to think of a more controversial proposition than that. So me thinks that Graham’s next challenge-conventional-thinking essay should be why elitism is good, not bad. And he does not have to start from scratch, he can read “In Defense of Elitism” by William Henry (http://www.amazon.com/Defense-Elitism-William-Henry/dp/03854...).
So why does Graham want to portray his group as non-elitists even though they clearly are? Because elitism has a bad rap. Some elitists are smug and arrogant but I do not think Graham is. I am pretty smart, or so I have been told. But I do not think I am as smart as Graham. When I read Graham’s essay, I feel, “OK, this guy is smarter than me, but he is not arrogant about it, he is laying out his arguments in a clear concise logical manner so that others, almost all of whom are not as smart as he is, can follow his logic.” That is not arrogant.
Some people criticize elitists if they do not give something back. Graham is clearly giving something back. His essays are free. If you don’t want to buy his book, you can read all of them (I think) online for free. He is giving ARC away for free. Although Ycombinator is hardly non-profit, I am still shocked at little equity they take (an average of 6 percent) in exchange for what they provide.
Graham argues that Ycombinator would not exist if all it did is fund “rare geniuses.” That is true but misleading at the same. I have been an entrepreneur all my life, so far successfully. The thought of working for the man, I would just jump off a bridge and die. But starting and running a company is a bitch, even for someone like me who is genetically programmed for it (and for nothing else). Our society glorifies entrepreneurship, which is a sign that our society is more advanced than others. In one of his State of the Union addresses, President Reagan urged every schoolchild to consider starting a business. The fact of the matter is that few people are suited to starting a company. You have to be good at a whole lot of different things; most people at best are good at one thing. You work all the time. There is currently a best selling book, The Four Hour Workweek. Yeah, right. Try the 14 hour work day.
Few people are suited for this, and that includes gifted hackers. Few people have the balls to start a company and that is a good thing -- perhaps 2 percent of the population is suited to being an entrepreneur, I doubt it is as high as 5 percent. Perhaps among hackers, or gifted hackers, the percentage is higher, but I can’t imagine it is 10 percent. Most people are better off working for someone else. That fact, by the way, creates a lot of opportunity for the few who are suited to start companies.
Ycombinator’s thesis is that there is an untapped market of young entrepreneurs who could start companies but do not and will if they are funded by Ycombinator. Yes, there is such an untapped market, but it is not as large as Graham implies. Let’s assume Ycombinator funds thirty companies a year, twice a year, and each company has three hackers. That’s 180 people. A rounding error of a rounding error, since every year more than 4 million people reach whatever age is the minimum age for starting a company.
As far as I can tell, Ycombinator has been successful. There are several reasons for this. There were the first to market, that always helps, people think of them first. Graham’s awesome reputation and his Web site are extraordinary marketing machines, so they get the pick of the litter. On the other side are several brilliant people with a variety of backgrounds who are not virgins, they actually have done it before; in short, they have the perfect background to pick the right people to back. So Ycombinator may not limit its investments to ideas proposed by “rare geniuses,” but the caliber of the people they back is so close to genius it may be hard to ascertain the difference. More than anything, Ycombinator has been successful because they are able to attract an extraordinarily talented group of applicants from which they choose a few to back. If Ycombinator started funding the “average” hacker rather than rare geniuses or close to it, I guarantee Ycombinator would be a flop. Graham’s statement that they are not limited to funding “rare geniuses” is a misleading as the Dean of Admissions of Harvard or Stanford stating that everyone and anyone should apply for admission to those universities. The fact of the matter is that it is damm hard nowadays to get admitted to Harvard and it is damm hard to get funded by Ycombinator, even if you are really really smart.
The readers may be interested in "Regional Advantage: Culture and Competition in Silicon Valley and Route 128" by AnnaLee Saxenian, a professor at Berkeley:
The book was written in 1994, one year before the Netscape IPO, but even 13 years ago, the author concluded that Silicon Valley offered enormous advantages over Massachusetts for entrepreneurship. Two points in particular are worth noting:
1. Professor Saxenian paid particular attention to the (back then) highly successful minicomputer companies: Digital Equipment, Data General, Prime Computer, Wang. All of them were vertically integrated companies that attempted to do everything themselves -- R&D, product design, manufacturing, sales, and after-market service. The SV companies were more likely to focus on their core competencies and then network with other companies to provide the missing pieces.
2. Employment laws make a significant difference. Under California law, a non-compete contained in an employment agreement is void as a matter of public policy. In the employment content, California courts simply do not enforce them, and thus companies do not even bother to ask for a non-compete in California. Under Massachusetts law, a court of equity will consider enforcing a non-compete agreement if it is reasonable in terms of scope, duration and geography. I say "consider enforcing" because to a large extent, it depends on which judge decides the case; some judges are more sympathetic to the employer seeking to enforce such an agreement while others are more sympathetic to the employee who needs to earn a salary.
The upshot of this is that in California, labor is extremely mobile. People quit their jobs on Monday and start a new company or work for a new company on Tuesday, and there is nothing the former employer can do in terms of enforcing a non-compete clause. (I am ignoring intellectual property issues that may protect the former company.) Because the new employer know that California courts will not enforce non-compete agreements, it is willing to hire employees that would not be hired in Massachusetts, and investors are willing to fund start-ups that would not be funded in Massachusetts.
In Massachusetts, due to the legal uncertainty and the possibility that a non-compete will be enforced, lots of employees are not hired, lots of start-ups are not started, and lots of start-ups are not funded, that would be in California.
Simply put, Massachusetts could increase its competitive advantage (or more accurately, decrease its competitive advantage) by changing its law concerning non-competes.
One issue that Saxenian and Paul do not address is the quality of the business laws and its court system. Ceteris paribus, a state where business laws make sense and where the courts enforce contracts and handle business disputes quickly and fairly will have an advantage over other states.
The "gold standard" is Delaware. The Delaware Corporations code, for example, is clear, well written and is in almost cases unambiguous. The Delaware Chancery Court (a court of equity that hears all business disputes) is respect through the world for its smart judges who show no favorites, enforce reasonable agreements, and decide cases quickly. Their supreme court (Delaware is small enough so that all appeals go directly to the state supreme court) is equally well regarded. To some extent, companies outside Delaware can opt-in to the Delaware legal system by incorporating in Delaware, giving Delaware jurisdiction to corporate governance issues, for example.
Neither Massachusetts nor California can compare to Delaware. In comparing Massachusetts and California, I believe Massachusetts has the advantage. At the Superior Court level, at least in Middlesex and Suffolk counties (the only two counties with which I am familiar), the judges are very well regarded. Massachusetts has a ton of law schools who produce an excess number of lawyers who do not want to practice law. So when a judgeship opens up, a lot of very good, well educated lawyers apply, and the committee has its pick of the litter. At the appellate level, the judges are also well regarded.
A recent development in Suffolk county has been the establishment of a court -- called the Business Litigation Session ("BLS") -- whose sole mandate is to adjudicate business disputes. Like federal court, the BLS handles the entire case from the time it is filed to trial. Allan Van Gestel is the judge appointed to the BLS. He is highly experienced in business disputes and is very well respected by lawyers who litigate business disputes. So far, none of the other counties have create a BLS, even though everyone agrees the BLS has been quite successful.
In California, the court system is less well regarded. The caliber of the trial judges is very uneven. Appeals in California are often unpredictable -- California is such a large state that different appellate divisions in the state simply disagree with each other, and thus there is not the predictability there is in Massachusetts. And the California Supreme Court does not step in enough to decide the disagreements among the appellate divisions, again because California is such a large state that there are simply too many appeals to the California Supreme Court, a small fraction of which are accepted for review.
In retailing, Wal-Mart is crushing their competitors, who are also extremely large but they do not have the economies of scale that Wal-Mart. (In addition, Wal-Mart has better management.)
In some cases, large companies succeed because consumers want the perceived value of purchasing from a large company. Dell and HP basically own the server and desktop computer market, and there is no reason to assume this will change in the next ten years. People feel that Dell will not go out of business.
One of the industries hurting most right now is the automobile industry. GM and Chrysler say they will run out of money unless the U.S. government provides several billions of dollars in aid soon. Ford says it does not need government financing but it has pledged all of its assets in a collateral-based loan. Yet almost no one is willing to buy a car from a small automobile company, they are viewed as risker than GM or Chrysler.
In choosing software companies, one factor sophisticated purchasers look at is, "How likely is this company to be in business 5 and 10 years from now?" When you commit to a software package, you are making a huge investment in time and energy, far beyond the purchase price of the software. You want to know that in the future, the software company will continue to enhance the product and will be around to provide technical support. This is why companies, for example, purchase software from Oracle, even though Oracle is ridiculously. People expect Oracle to be in business ten years from now. Yes, open source to some extent mitigates this risk, but putting aside very successful open source projects (Linux, Apache, MySQL, how do I know that they will be around ten years from now. Do I really want to writing the code myself?)
Law firms have consolidated tremendously in the past decade. Large clients want one firm that can provide experise in dozen of areas of the law and can handle complex transactions that span the globe. In Boston, ten years a 100 lawyer firm was considered to be a large sophisticated firm. Now such firms (at least on the corporate side) have merged with other firms to create 500 attorney firms.
In public accounting, the Big Eight has consolidated into the Big Four. Clients want the brand name that only a Big Four firm can provide.
There is no clear trend here. In some industries, consolidation is taking place and will never stop. In other industries, it is the opposite effect.
James Mitchell www.bostonconvivium.com [email protected]