The Future of Web Startups
paulgraham.com
paulgraham.com
Yeah, but you get a 6 month grace period after school's over before repayment begins, and you usually have several financial hardship forbearances you can take to postpone repayment. Even then, you can go back to school for one semester, and reset the grace period and financial hardship forbearances.
So, you have at least 18-24 months of forbearance/deferment before you have to begin repayment, and that can be hacked indefinitely by taking 6 credits at your local community college.
Also, free interest or low interest as "aid" is a joke. You still have to pay the entire amount you borrowed.
I think the only time your strategy makes sense is if you can pay $1,000 to enroll in community college for one semester in order to stay on your parents' health insurance from June to December, which comes out to $140/mo, and the quotes you received for health+dental+vision were $500 a month, therefore saving approximately $2,500 per semester, or $5,000 a year. (And I believe if you're from Massachusetts, you can stay on your parents insurance as long as you're a dependent in income level even if you're not a student.)
" In fact, most graduates want to find stable jobs to pay off their debt -- startups are not really on most graduates' radar."
That's irrational. If they're already way in debt, shouldn't they want to take more risks? Why the hell would you do something boring for ten years to pay off your debt, when you could do something fun for one year and have a 10% chance of more than paying off your debt. Also, employers are impressed by people who start their own businesses.
" Businessweek had a survey of the top ten desirable places to work for new graduates -- three were govt. agencies, one accounting firm, google, etc."
Right, because Businessweek can't say "The best place for you to work is that company you start in your apartment with your pals," because 1) that doesn't appeal to BW's audience, and 2) startups are small and varied, so it wouldn't make sense. It's like taking a survey of the most popular restaurants (McDonalds, Burger King) and using that as proof that the swanky new sushi place down the street is an awful place to eat.
No those people would not drop out earlier, because they likely did not know they wanted to be startup founders until they had enough technical knowledge and/or more importantly built those relationships with students around them which led to the creation of a startup in the first place. Perhaps your experience has been different in which case I cannot argue with that.
"when you could do something fun for one year and have a 10% chance of more than paying off your debt"
For precisely that reason -- that the likelihood of success is only 10%. And it often takes more than a year to build something meaningful.
"Also, employers are impressed by people who start their own businesses."
This is not always true. It depends on the role the employee is being hired into.
"For precisely that reason..."
Did you miss what I said above? If you're already broke, why is getting broker a problem? What do you have to lose if you have less than nothing?
Can you give me some detail on your 'depends on the role' comment? I didn't realize this; when I recruit people (IT/quant finance), employers respond more favorably to entrepreneurs than to people who worked for someone else. Where did you hear otherwise?
Yes that's exactly what I'm saying. You meet like-minded people there who are excited about technology - potential co-founders. You learn about technology.
I've been through 3 universities, and the relationships I have built at these institutions have helped me to this day. I'm running a startup btw.
Is this like saying "If you're going to be a programmer, you'll do better to buy a computer. But if you buy a computer, you'll have to get money, which leaves you less time to be a programmer!"?
I would disagree with this: "1) that doesn't appeal to BW's audience" starting businesses _does_ appeal to businessweeks audience. ;)
I wouldn't be surprised if BW's readers were less entrepreneurial than average. It's a magazine about big business, not about doing business.
1. They took basically zero risks before, so it would not be "more".
2. And screw up their credit history/get creditors chasing them?
> So wouldn't those people drop out earlier?
Because they couldn't predict the future? Not everyone who does a startup knew their path since they were 2 years old.
While Paul focuses on web based startups, I think this phenomenon will apply to more areas as technology enables individuals to create their own niche. This is already true for some sports and certain types of businesses.
Like you correctly mentioned, the debt problem is more pronounced for Americans and some other nationalities, most developing nations do not have this problem and their entrepreneurs will start startups especially in developed and lucrative markets such as the US. This inturn will force the US mechanism to adapt. This effect of globalization has already had effects in many other industries.
...because we've inspired everyone to do startups. There will be nobody else to hire :)
We're back to the pre-industrial revolution era. Where everyone didn't hire anyone and had their own backyard industry.
Woohoo!
I was surprised to find it in a PG essay, but on reflection it seems really appropriate in context. Though it did distract me enough that I had to read the last section twice.
Are you suggesting that investors should be racially biased?
I'm playing devil's advocate far further than intended for sure, but the laws regarding employment, although bureaucratic are there for very good reasons.
"I'm playing devil's advocate far further than intended for sure, but the laws regarding employment, although bureaucratic are there for very good reasons."
What is that reason?
That's what our society has decided, therefore employment laws, while imperfect, have good reason.
For reference, see 19th century factory conditions.
If not, why the distinction?
I'm not sure we need laws that commit them to be though for the simple fact that it would be very hard to actually enforce those laws.
Employers are a completely different situation, and thus are bound to a different set of expectations.
See also: http://online.wsj.com/article/SB118134633403829656.html?mod=...
And I don't understand what you mean about laws being proof that people feel this way. I assume that the proof that people feel a certain way is that they act that way. For example, if I think being a coal miner is too risky, I won't be a coal miner; if I think the rewards are fair compensation, I will be one. Why do you need a law to deprive me of my freedom to make those decisions?
>For example, if I think being a coal miner is too risky, I won't be a coal miner.
Unless of course you're not in the position to make that choice. You are speaking from a position of privilege and assume all share that privilege. It leads to only a half understanding of why we've instituted the systems we have.
>Why do you need a law to deprive me of my freedom to make those decisions?
Wow, that's a warped view of freedom. Too many Ayn Rand books to be sure. If you were right in your theory, the nation you live in (assuming you're American) wouldn't exist right now. Think about that a little.
" You are speaking from a position of privilege and assume all share that privilege."
Pleasant nonsense. Why should my status change the validity of my arguments? I have an idea -- let's ask a black high school dropout (http://www.tsowell.com/) if he agrees with me! Then let's decide that an individual's 'privilege', whatever that means, isn't as important as making a reasonable argument backed by data.
"If you were right in your theory, the nation you live in (assuming you're American) wouldn't exist right now."
If I were right in my theory that depriving people of freedom in order to force them to make decisions they're going to make anyway (at best!) is a bad idea? I don't see how my rightness would affect American politics. Do you mean that if more people agreed with me, the world would be different? Are you confusing my perscriptions and historical descriptions? Please clarify.
That got me to thinking though. While I'm attracted to the sentiment of freedom above all else, what I think is missing in his equation is that communities tend to be sort of self-reinforcing. It would be very difficult, if it were accepted by others to not hire black people, to be the first one to do so. Now, laws against that kind of discrimination don't change the way people feel, nor make their racist sentiments go away overnight. However, they do drive those thoughts underground, and break the cycle of "we've always done it that way" by making it no longer acceptable. And with time, that improves the climate and makes it easier for attitudes to change.
At least that's one thought that came to mind. I really admire Friedman's purity of thought, but think that perhaps it's a bit naive. It would certainly be an interesting subject to discuss in person with a person of that libertarian mindset.
I dont foresee increased regulation to be a big problem. Hindering the development of new businesses would be shooting the economy in the foot.
That's just a bunch of bunk. Nothing was funnier during Web 1.0 than guys saying we were "rewriting" the rules of business - like, profits didn't matter. Truth is, from a business point of view, starting a web service is no different than starting a dry cleaners: give people what they want, at a reasonable price (in our business, that usually means ad-supported)... and they'll come back (and click on ads in the process).
Yes, it may take some technical innovation... but it doesn't take any business innovation. Everything you say in your essay is true; so seriously, why do I need you? You expect me to go through this whole dance to come to Cambridge for, what, the possibility of five-thousand bucks? And then move to SilCon for three months so you can keep tabs on us once a week? Like, are you serious?!
Your essay is particularly prescient because you might have inadvertently predicted the ultimate demise of your current business - i.e., venture funding. No, I won't fill out your form; no, I won't travel to Cambridge for the possibility for you to give me five-thousand bucks. No, I won't move to the Valley so you can keep tabs on what I'm doing. I'm a big boy, and I can take care of this end myself, thanks. Good luck with giving away your cash.
James Touchi-Peters, Founder/CEO CONK!/Arrow Media Minneapolis USA conkhead@conk.com http://conk.com
It's like any other long tail, in this case the prototypical founder (late 20', single, technical guy) will not be as common - it'll become a more-diverse population. There will be younger founders and there will be older founders, too. Nerdier, and hipper. Etc.
For example, I'm a 40 yo front-end developer (started in 1995 building interactive apps at ESPN). I'm now starting a company, doing all the backend development too, thanks to Ruby on Rails. This was never available to me before 2006, really.
My bet - the trend will be away from the core developers, as the back end tools become commoditized and packaged, and the prized employees will become the designers - UI and graphical. Brands and positioning will start to play a larger part, and these contributors will have the most to say about this part of the company. I predict the trend will go away from the technical, as people will need to wear many hats in a small company.
As for the young, I don't believe they have an advantage in a crowded startup environment. The lucky bet, or obvious missing younger-demographic website will be created by the kids, but this will happen less and less as the low-hanging fruit is gobbled up. Instead, I think the older (or rather experienced) will have an advantage in a crowded startup world. Here, it'll be the niche positioning, establishing business relationships, and pursuing effective growth strategies that will make the difference.
1) The university promises admission the top student from the Japanese language program of a certain high school each year. Situations like this often arise from alumni putting crazy earmarks on the donations to their colleges.
2) The university guarantees admission to a certain percentage of state residents each year in exchange for state funding.
3) The university wants to buy more land but the town tries to block the deal because it doesn't want the property taken off the tax roll. The university brokers a deal guaranteeing admission to a certain number of students from the local HS in exchange for the land.
4) A top prep school comes to some more formal agreement with a university for purely strategic reasons.
Situation #1 seems to be the most common in terms of connections to prep schools, although it's hard to say since the relationships aren't exactly published.
pg, I am a little alarmed by this flat recognition that is awarded to Google. It might be actually true. But its not an optimal way to think about things in the long run. As a teen ager I was a Michael Jackson fan, It took me a good while longer than other people to realize that he has turned into shit.
It's like saying "And being Wal-Mart, they're figuring out how to squeeze the last nickel out of their suppliers." It's just how they're interested in doing things.
You say one goes to Silicon Valley to be in a community of startups, meet cofounders and talk with smart people. Then you say that in the future, in college, we will be more concerned with meeting cofounders and communing with smart people than we will be concerned with grades. Why won't yo want to go to the school with all the smartest like-minded people? Thats what Harvard is to business people. A community of like minded people that band together to help each other. The same will be created at the new breed of startup oriented tech schools. The admission standards might be a tiny bit different though.
I think it's right that startups will be acquired earlier and earlier, but that makes me wonder: might those earlier acquisitions (which sometimes decrease productivity to 1/13th what it was) sabotage the success of the startup either because once acquired the founders don't have to work as hard to keep their heads above water or because when not constrained by a tiny startup budget the founders won't worry as much about finding efficient ways to do things--which in the past made them better startups/companies?
A company that bought a bunch of hot startups early on and didn't destroy their productivity would be a force to reckon with.
Your each and every article is giving indication that married people are not good enough to start startup. Come on man - I am reading your articles to inspire myself and to keep up my tempo but when I read age analysis then that is very demotivating.
Well I think I need to find place where there are success stories without age limit ;)
I understand that kids and wife can take quite a bit time. But if there are people like me who dreams about own business for years and wasn't able to start it because never found right channels and due to that now I am above 30. Now I have vision and courage but also I am married. Does it mean that I have to stop seeing dream? No way man. I am here to prove you wrong. Will let you know whenever I am succesful :)
Thanks for all great articles.
I do think taking that last exam and finishing university was a magical moment. Following a commitment through to the end even when things are tough or tedious is valuable experience. You can get this other ways but there is a huge difference between someone who stuck in there and finished university and someone who called it off just three credits short.
If the number of startups founded goes up by an order of magnitude or two, then the need to move to Silicon Valley in order to be in a startup-rich environment goes down.
Now, I do think that startups will still move, but I think that the number of hubs they can choose among to move to will be increasing as well.
I also have a feeling that beyond a certain density, the largest hubs will eventually hit diminishing returns in terms of their utility for startups (which will prompt more hub specialization).
Lots of folks have tried to create startup hubs. What will be different this time?
And in any case I doubt the number would increase. The power of hubs is self-perpetuating. So no matter how many startups originated in St. Louis, there wouldn't be a lot there when a given new one started, because many of the previous ones would have already moved.
(Suppose it happened that 50% of the next generation of big movie stars were from St. Louis. That would not make St. Louis a rival to Hollywood.)
Also there's the empirical argument: I can't think of one industry in history whose growth made the established centers irrelevant.
If the trend if that many more small startups with bold ideas start to win, other VCs have to become more bold, and so the difference between, for example, the Valley with their aggressive VCs and other areas would diminish.
I do agree with your basic premise though.
How about textiles, after the spinning wheel came to Europe and the cottage industry developed?
If there are 'only' 5-6 startup hubs today, not too long ago there was only one. Since the other 4-5 hubs managed to grow in spite of SV's existence (and dominance), I do expect this number to increase further, by one order of magnitude at least.
I don't think the star-system is necessarily a relevant analogy because it is focusing on individuals, who as actors, actually have quite a few more options than just Hollywood. Rather, what would happen if 50% of the next generation of movie producers and production companies were from St. Louis (or Bollywood)? Or, for a different scenario, 50% of the next generation of SFX and post-production companies? Do they all have to move to Hollywood? Can some of them move back? For some types of companies, It's actually demonstrable that they don't all make the move, and a rash of soundstages and post-production facilities are sprouting up in secondary hubs across the US because Hollywood is too expensive and crowded.
Your empirical argument is well taken. By no means was I arguing that SV would become irrelevant to the current industry because of growth per-se, but you're assuming that startups are (and will remain) an 'industry', and second, that they will remain the same industry.
Personally, I anticipate quite a bit more disruption and creative destruction than that. The US rust-belt is testimony that further growth in an industry can under the right circumstances pass existing hubs by, eventually making them irrelevant because the industry itself has changed in fundamental ways. It is also possible that 'startups' might get distributed into the broader economic fabric as thoroughly as 'web design' or 'desktop publishing'. This may actually be required if companies in all industries (in industry-specific hubs everywhere) are to grow by acquiring startups rather than hiring.
The company that manages to blend digital identity and reputation with the information power of networks is probably going to be the next MS. I've been interested in this problem for a long time now. Someone should really submit a proposal as a YC idea.
I do agree, though, that Japan does have a reputation of being not very inviting to foreigners that is to some degree deserved. I don't know as much about S. Korea in this regard.
Japan could probably do much to solve some of their looming economic problems (low birth rate and rapidly aging population, for example) by allowing more people to immigrate and integrate into Japanese society and culture. Especially entrepreneurial, technology loving types who could contribute a lot to Japan's high tech economy.
I know a couple other startups in my town (one of your big 3) who do similar things to my startup. Thanks to the public (link-happy!) structure of the web, we already let people go from one to the other: a loose ad-hoc integration. Will this be the future of cooperative web software development?
Sure they can merge, but this is not an "exit". As far as I know, the purpose of a startup is a successful exit of some type.
I'm not quite with you on universities.
I haven't been to US education system, but I don't think you get to know stuff like Fourier series or why inverting a matrix is unstable, etc., until you get through a university. And such stuff is important in high tech. Unless you're into web sites that host user generated videos. Being in video hosting pays off, for some, but it's not what I'd call high tech. I mean, it isn't a new UNIX, or a new Internet, or a new CPU design.
Those bigger, more interesting things were done by people who'd been to universities, and I think there's a reason to it.
Those startups started by undergrads, they may very well make money and produce cool apps and products, but I doubt they will drive technology.
I recently found I wasn't keeping up with new languages any more. Not because I've become too lazy, but because there aren't any new concepts in all those Pythons. Those are products of smart hackers, and new concepts are done by scientific kind, who seem to be out of fashion today.
Thanks for your writing anyway.
Alexander
While I like having in my mind all about good stuff from Silicon Valley, VC and tech folks, I would encourage Paul to write about bad things that can happen to hackers when they deal with VCs or Angels like dropping out of the Board, etc. it would be good to maybe prepare us tech folks to handle the truth that not all people will be good people around us that will try to invest in us, or are there only "good people" in Silicon Valley?
Through these projects I learned new ways of making things using current technologies like Ruby on Rails, Python, Mapreduce, Amazon S3 (and friends), Lisp, and other technologies that someone simply fulfilling the graduate requirements would likely not have learned. I also met great people along the way who were far more impressive than any programmer, designer, or manager I've met in the industry.
Oddly enough, now I'm in grad school, hacking the iPhone for a degree.
Eventually, I imagine that my continual curiosity will bring me to the startup arena, or perhaps to the international sphere.
ak
However, there are still advantages to a biz being close to itself and its resources.
The combination of those to things means that startup hubs will become stronger.
Hello, Paul. I'm selection bias. We haven't met formally, but you're obviously acquainted with me.
I can see this being very beneficial, both for the VCs and the startups. The standardization of terminology in any discipline or field allows the people in it to have their lingo easily referenced and indexed, increasing efficiencies. Who wants to wade through a bunch of superfluous gibberish to get to the point?
In grad school, I remember having the realization that every discipline has it's own little shorthand lingo. I remember thinking about how confusion can result when there are clashes/overlaps in, for example, acronyms.
>>An angel who wants to insert a bunch of complicated terms into the agreement is probably not one you want anyway.<<
Exactly: reducing complexity, not increasing it, seems a better objective.
I've long ago lost track of all the different meanings I've run into for the word "feature", for example. Practitioners of various academic disciplines (or even in the same discipline but for a different theory) use this term for all kinds of things with absolutely no connection to each other. I nominate this for most overloaded academic term.
Any other nominees?
While it depends on a nature of the business you are starting, in many cases "build something quick" advice does not work as well as advertised IMO: for most non-trivial ideas "something quick" still means a few months of full time commitment which means tens of thousands of dollars that many "younger nerdish" founders still do not have.
I guess what I am saying is that a million dollars is no different from fifty thousand when that's the money you need but do not have. Sure it probably easier to find an external investor with 50 grand, but is it really? Especially for "nerdier" types?
This is why YComb exists, but that is only about 30 deals a year.
Furthermore, I'd like to point out that even if local YC clones perform worse than YC, there may still be a market for them. State governments pay much more than YC (monetary) costs to encourage YC effects. Local VCs demand more startups, and people to tell them which startups to invest in. And as I argued above, many YC-caliber hackers would be interested in a local YC knockoff, but not YC itself.
The one thing I find interesting is the overall theme that seems to pervade many of these essays--mainly that all trends are moving to increase the importance of hackers. While this is true for now, I think the more the web evolves and the easier it becomes for non-hackers or even technologists to build apps and start companies then the advantage potentially moves strongly to those that understand consumers and markets. Which may well be hackers, but also may not depending on the app--as has been pointed out in the past by Josh Kopelman the Tech Crunch 50,000 of 2006 (or even 550,000 of 2007) is not the mass market.
Thanks again--nothing I read provokes more intense thought than your essays.
I actually think Microsoft does.
One last thing: Recognize that success is a shape-shifting beast.
Look out, oldsters with all your real world money and stuff, Reality 2.0 is on its way, and it doesn't look good for slow-moving dinosaurs. I, for one, welcome our new 20-something overlords. I cross-posted on your piece to http://blog.innovators-network.org The Innovators Network is a non-profit dedicated to bringing technology to startups, small businesses, non-profits, venture capitalists and intellectual property experts. Please visit us and help grown our community!
Best wishes for continued success,
Anthony Kuhn Innovators Network
I don't think that's the only reason founders want to keep risk low.
The most important reason is that money has diminishing effect on happiness. People want to be able to guarantee a "moderate" life. What comes beyond that is not as important basic needs.
If you are instead saying that a big company is not designed to scale as fast as a small company/startup then that is perhaps inaccurate too, especially if you were to determine scaling as a function of headcount and/or revenue.
Hrm... one of the things that makes markets work so well is that it brings "the wisdom of crowds" to bear, and doesn't require small, homogeneous groups to attempt to pick winners. So something about that section raises a flag in my head - if there are more startups in the future, maybe more of them will have to go it alone, instead of this "easy out" of being bought?
It would also be a great change to see technological universities align their teaching to help students startup companies instead of preparing them for 9-5 work.
The whole essay is based on this idea. But Paul gives no reasons, why there should be a such a trend.
Cheap CAD tools, cheap custom boards, processized manufacturing.
Most people drastically underappreciate how much China has helped innovation in this regard.
You have software to rate the applications?
If the number of startups increases dramatically, then the people whose job is to judge startups are going to have to get better at it. I'm thinking particularly of investors and acquirers. We now get on the order of 1000 applications a year. What are we going to do if we get 10,000?
That's actually an alarming idea. But we'll figure out some kind of answer. We'll have to. It will probably involve writing some software, but fortunately we can do that.
So I couldn't agree more to your 8th point on the future of colleges!
I am a professor at a German University of Applied Sciences for 7 years now and a serial entrepreneur and business angel of 9 start ups.
Pedantic Typo Report
The Future of Web Startups, Section 2. Standardization.
"In essense, let the market design the product."
"essense" vs. "essence"
That is all.
thank you for your communication skills.
I found the warning that it's getting more risky to just sit on ideas and not act inspiring. Will be doing extra coding tonight!
I have enjoyed all ur essays and am fascinated by how insightful they are. I am graduating from college in december with a bachelors in electronics. Could u please recommend some of the books which have helped evolve ur thinking to this level.
Thanks
Gaurav
In the future maybe they could offer this to entice more young people to start companies.
I know of one company that already has this, with a different title.
Ideas for new subjects:
- "google is dead";
- "how to fund painters";
- "all things considered, java rocks";
- "vaporware, or why I gave up arc";
- "return of the web 1.0"
There are plenty of weird ideas that only wait for PG to write clever stuff on them! Seriously, I don't know why people like FOWA, whoever that is, actually pay you to listen to the same old story...
And oh, the part about ideas for new subjects... ok. I am wondering why I didn't downmod you now.
i'd love to be on the bandwagon already, but i don't meet a lot of the qualifications. if the bar is lowered just a LITTLE bit, i am so there.
by "don't qualify," i mean i'm not really into the game the same way as the companies ycombinator seems to fund. for example, i'm not that excited about writing web apps, it just doesn't appeal to me. i think there are important sectors that could benefit from a bunch of startups that paul has completely written off, because he thinks it's less lucrative, i guess. well, i'm willing to take a LOT less money to do something i'm actually interested in, rather than learning ruby on rails, say.
this is paul's world. i like reading about the energy and creativity involved in web startups. i personally think that narrowing the field to just web startups is short-sighted, but that's not my call to make, certainly not here.
Interestingly, those also seem to be some of the higher profile projects.
i'm personally more drawn to "real" software that runs on desktops and servers. web stuff is nice, but it doesn't fit the way i think as a programmer.
I haven't spent one penny or any time on/in college!