How not to die

http://www.paulgraham.com/die.html

by swombat • 17 years ago

145 47 17 years ago

paulgraham.com

sound2man• 17 years ago

Work hard and never give up. Got it.

aditya2• 17 years ago

or, work hard and know when to give up. much harder.

ErrantX• 17 years ago

I've started 4 companies. 3 died (but not badly) one petered out and I sold the core tech for a pretty decent figure (so I call it a semi-success even though it was the least likely to succeed)

I dont know: I came to HN after I began hacking and doing startups (I never even called them startups till I got here) and my assumption was that 99% failed anyway as a part of the process of learning.

I figure you have to either be super lucky (and make a Twitter, Facebook or Google hit) or you have to fail hard a few times before eventually succeeding. I chatted to some London VC's the other week and they said they trusted founders from previously failed startups a lot more than newbies - simply because they had learned many of the hard lessons and wouldn't then make them at the crucial moment their company could be worth something 6 figures or more.

jpwagner2• 17 years ago

VC's want the choice of whether or not to invest in you.

So naturally they trust your attributes alot more than others' when they are talking to you.

chrischen• 17 years ago

> I dont know: I came to HN after I began hacking and doing startups (I never even called them startups till I got here) and my assumption was that 99% failed anyway as a part of the process of learning.

Same here I've been doing startups since 13, didn't even know about the dot-com boom bust at the time.

jacquesm• 17 years ago

I don't know anybody that 'scored' that didn't fail a couple of times first. The ones that hit a home-run the first time out are the ones that you hear about all the time but they're the exceptions.

It's not a rule, but hard work definitely pays off - most of the time.

iamwil3• 17 years ago

the previous discussion when it was last submitted.

http://news.ycombinator.com/item?id=48294

swombat• 17 years ago

I thought it was relevant (and far enough away time-wise) for a repost, given the news from SnapTalent today.

zackattack• 17 years ago

http://img.skitch.com/20090818-ca7b431u5urtm7rpdh1ag7uwnx.jp...

mrshoe2• 17 years ago

I'd read that before, but I think I needed it today.

Maybe someone should set up a cron job to submit that talk to HN every 3 months.

jpwagner2• 17 years ago

Or to randomly choose one of PG's essays every month.

This particular one was last posted almost two years ago.

edw519• 17 years ago

I have cut & pasted all of pg's essays onto my Palm pilot. Along with my bridge program and a bunch of other great content, I never have an excuse not to learn something.

Whenever I sense that I'm getting stagnant, I always go back and reread old faithful, "Good and Bad Procrastination".

http://www.paulgraham.com/procrastination.html

(When someone invents an ebook reader that plays bridge, I'll make the switch.)

mrshoe2• 17 years ago

Startup idea: present the user with 8 adjectives and ask them to pick the one which best describes their mood, then give them an optimal article to read as they start their day.

We'd all be more successful if we could capitalize on our good moods and not let our bad moods drag us down for long.

andreyf• 17 years ago

Two years? Has it really been that long? I'm old :( and in need of moving on to new memes - any suggestions guys?

iamwil3• 17 years ago

There's a link to HN library at the footer. You can browse that at your own leisure.

ujjwalg• 17 years ago

I am finishing up my PhD and I have a very direct analogy to what PG mentions.

There are multiple research groups with different success rates and what I have noticed is groups with extremes are not doing that well. By extremes I mean professors who try to micromanage and get a daily update and professors who do not care at all. In my group, we have a biweekly meeting for everyone to present what they did in 2 weeks time which makes all of us work hard but on our own terms. I think incorporating similar things in education/research will be very helpful in the long run.

Good article PG!!

paul9290• 17 years ago

Sorry I think the part about starting up other things is ill-advised.

TWITTER WAS A SIDE PROJECT!

aditya2• 17 years ago

When they knew Odeo didn't have enough traction: http://www.stubbleblog.com/index.php/2009/06/the-real-lesson...

nostrademons• 17 years ago

Side-projects are always confusing because the answer to "Should you do a side project?" is always "Yes - if that side project will eventually become your main project." Unfortunately, it's almost never possible to tell that a priori. So it always comes out as "You should do side projects, except in the cases where you shouldn't", which isn't terribly helpful.

steveplace• 17 years ago

In addition, once ev realized the side project was better than his current startup, he shut down the original, gave back all the money to his investors, and raised for twitter.

zaidf• 17 years ago

This ranks right up there with Steve Job's Stanford speech.

It was even more powerful when pg actually delivered it. I wish a video from that night's dinner existed.

iamwil3• 17 years ago

The part I found most useful is where he quotes Paul Buchheit, and about pleasing that first small set of users. He doesn't tell you where to find them, but at least once you've found them, you can hang on to that through thick and thin to protect your morale (given that morale is what motivates founders not to quit)

martincmartin• 17 years ago

Am I the only one that thought PG was going to come out as an extropian?

Estragon3• 17 years ago

So, how are octopart and wufoo doing? (They both seem to be around, still. Have they been in the news, lately?)

pg3• 17 years ago

They're both doing well, actually. Both are profitable. Wufoo remarkably so, but they've had longer.

prakash• 17 years ago

Quick question: Do you make a note of how successful a startup would be when you fund it, and then revisit those notes a year or 2 down the line? How do you make sure you have the right feedback mechanisms in place?

Estragon3• 17 years ago

wufoo's demo video peters out just where it gets interesting to me, and it confuses me about their value proposition: once you've got the form made, what do you do with it? Do they just give you the html and point you to where the URL for the CGI ought to go?

diN0bot• 17 years ago

i asked the MIT Electronics Research Society http://mitERS.mit.edu if anyone used octopart and what they thought. here are the responses so far (sent to public list not just me personally):

"It's a great idea, it's useful, but it's a work in progress and not yet a replacement for searching distributor websites." -henry

"If you know what you're looking for and Octopart has incorporated it, you'll find it in seconds flat, and it's fantastic. Unfortunately, it seems spotty in terms of catalog completeness, and it is not as comfortable as browsing/guided search on, say, DigiKey or Mouser, so it gave me mixed results." -darwin

cmos• 17 years ago

I agree and disagree. Octopart is very revealing if you do low volume manufacturing had have been caught in the 'digikey convenience turbine'. Like any search engine, it's useless if you don't have the part. But there are a lot of core parts to a generic circuit board listed, so I end up using it quite a lot.

andres• 17 years ago

thanks for asking mitERS for their input. please let everyone know to email me with suggestions on how we can help them search for parts (andres@octopart.com).

keyist• 17 years ago

Do not go gentle into rewrite / Rage, rage against the dying of the site.

/me ducks

nopassrecover• 17 years ago

I thought it was customary to include the year at the end of old/reposts.

gjm112• 17 years ago

I fear there may be less to this than meets the eye.

It reminds me of a foolproof recipe for (human) immortality, which I think I read in a book by Raymond Smullyan. It goes something like this: every day you say "I will say this sentence again tomorrow". If you always keep that promise, then you are assured of never dying.

The point, of course, is that "if X, then Y" doesn't mean that by doing X you can ensure Y, even if on the face of it X is something under your control.

Similarly: yes, indeed, if your company continues interacting with PG and other Y-Combinator-funded companies, then it will not die -- because if it dies, it will no longer be able to have those interactions.

pg3• 17 years ago

But that's false. People don't lose the ability to email us once their company dies.

gjm112• 17 years ago

The companies no longer have that ability, even though the individuals do.

Don't get me wrong; I think "keep doing something and stay in touch" is an excellent piece of advice for someone doing something difficult where failure is strongly correlated with losing motivation.

But I wonder whether you're overestimating the causal role of demotivation here. Toy model: a startup has some kind of "health score" which varies according to exogenous events; "motivation" is some function of health and d(health)/dt, which is small when the startup is unhealthy and getting healthier; the startup dies if its health goes negative. In this model, motivation has no causal role, it's a pure epiphenomenon; but I bet you'll find that startups that die typically have very poor motivation shortly before their demise, and that poor motivation is a good short-term predictor of failure.

Of course, this is only a ridiculous toy model; in reality motivation clearly does have causal effects of its own, but it's also clearly affected by outside causes that themselves have an effect on the success of a startup. An observed correlation between demotivation and death is only weak evidence that demotivation is the underlying problem more than it's a symptom of having no customers, no money, technical problems you can't solve, etc.

(Doubtless you have more evidence than that correlation, having seen lots of startups succeed and fail.)

helveticaman• 17 years ago

What if the economy is a shitstorm? I'm not talking about right now, either; I'm talking about an Argentina-style, complete meltdown. What then? Can it become implausible to do well?

pg3• 17 years ago

Obviously there's a point at which it could, yes. If you lose civil order, for example, you lose most of the economy. That's what really destroyed the economy of the Roman Empire. There were of course deeper underlying causes, but the proximate cause was plundering and brigandage.

elai• 17 years ago

Well argentina isn't the only economy in the world ;)

AlphaEvolve• 17 years ago

This post really came at the right time for me. Such a moral boost. I created and is currently operating a startup in orange county, ca. I am also in the middle of a divorce. So to sum it up: Though times. Personal advise: If you believe your idea can provide value to clients, end-users, or to the community as a whole: Never ever give up. It will eventually pay off.

wastedbrains• 17 years ago

I read this whenever we are currently on a down swing on the roller coaster that is a startup.

How not to die

Related Stories