Hacker News new | past | comments | ask | show | jobs | submit login
Survivorship bias and startup hype (sonyaellenmann.com)
340 points by ifcologne on June 14, 2018 | hide | past | web | favorite | 118 comments

If you don't mind being relatively unheard of and only want to make enough money to be independent, then it actually isn't very hard to bootstrap a business. It does take time, effort, and a lot of mistake making, but it's like learning anything of value. I think it makes you a stronger and more independent person all around.

Even though this article attempts to deliver a dose of reality, it focuses on VC backed companies which is a completely different field. VC backed startups tend to do highly risky things in an effort to increase traction and valuation as quickly as possible. Most notably, spend a lot of investor money in pursuit of those things without having any real revenue. The odds of failure skyrocket.

I've seen a lot of commenters go on about survivorship bias. I believe that the real problem isn't being considered: the absolutely batshit advice and business building strategy that comes out of the world of venture capital and tech accelerators.

>, then it actually isn't very hard to bootstrap a business.

Your "not very hard" advice also inadvertently distorts the reality of possible outcomes. Most bootstrapped businesses fail as well.

Whether an entrepreneur gets some friends with a bootstrap mindset to start a consulting business, publish a new magazine, or open a restaurant, the odds are that all of those will most likely fail. Any business that can provide financial independence has long odds. (I'm not including solo businesses such as freelance web consultant, insurance agent, home inspector/appraiser, solo attorney, etc.)

Whether it's a bootstrap-business or VC-business, founders will still proceed forward even if it disagrees with "rational" calculations of successes/failures. It's human nature.

Yeah, but the reasons for failure are different. A startup usually fails because the founders are convinced they can't do whatever they want to do without investors and they never raise any money. OR, they do raise money and eventually run out of it because they never made enough money to support the costs that were enabled by ... the investment.

A bootstrapped business is alive as long as the founder is willing to continue trying to make it work. So a bootstrapped business doesn't fail per se — usually the founder fails to make money within the timeframe they'd hoped for and gives up.

I didn't start making real money with my business until I was 4 years in. There wasn't some magical day where I got lucky and started making money. The first 4 years was me basically being an idiot and making every mistake possible. It gave me the education I needed to price my product appropriately, position it, and sell it.

Some might say I'm a "survivor" but I don't think I am. I'm just determined as hell and didn't give up when I wasn't making any money. I worked freelance contracts and kept the thing going because I knew at the very least I was getting smarter even if I wasn't bringing in money, and eventually the education would probably prove useful or pay me back. Better and cheaper than grad school anyway, right? I sometimes think about the people who are "Masters of Business Administration" on paper and have never actually built or run a business.

The people I know who have successfully done it aren't geniuses with a bunch of money and lots of connections. They all just had the same determination to stick with it, adapt, learn, and find their way to what they want. There is definitely some pain and self-doubt involved, but like I originally said, you come out stronger for it.

>A bootstrapped business is alive as long as the founder is willing to continue trying to make it work. So a bootstrapped business doesn't fail per se

I'm going to rewrite your comment in more concrete terms and by doing so, you can let me know if I have parsed your meaning correctly.

For example, in California, the annual filing fee for C-Corp or LLC business entity is $800. So, for as long as the entrepreneur is diligently paying the $800 each year, the business has not been formally "shut down" and therefore, has not "failed". (Alive != Failed)

With that adminstrative setup as the baseline, we consider your next comment:

>didn't give up when I wasn't making any money. I worked freelance contracts and kept the thing going [...] even if I wasn't bringing in money,

I think this means that one can get any 9-5 day job and keep the "bootstrap business" as a side hustle for years that never brings in significant money. It can never be considered a "failure" because of 2 properties:

(1) the business wasn't formally wound down (just continue to pay $800/year admin fee).

(2) a money-losing business still provides "education" and founder comes out stronger. (Not sure if this also implies that failed VC business does not provide education and strength)

It seems like your advice is dependent on a hyper technicality of what "failure" means to you instead of engaging the spirit of the comparison between VC vs bootstrapped.

I go back to your first comment which already had as a condition "want to make enough money to be independent". Many current and former bootstrapped business founders will disagree that it's "not hard." It is extremely hard to create a self-sustaining bootstrap business that brings financial independence reasonably fast. (The "fast" being relative to working 30 years at a 9-5 job.)

Also, when you rely on phrases such as "determined as hell and didn't give up", you're inadvertently falling into the same "survivorship bias" for your bootstrapped business which the article complains about for VC businesses. Being "determined as hell" is an unfalsifiable attribute. Oh, you say your bootstrapped business failed?!? It happened because you weren't determined as hell.

I'm sympathetic to bootstrapped businesses because I sold custom software that way but touting the advantage to non-VC businesses as one that let's you redefine "failure" in an idiosyncratic way which keeps the business from meeting that criteria -- is not very compelling.

In any case, if the business is in a domain that can be bootstrapped, by all means, do that option. The founder will have more leverage. Leverage is good.

The point to my first comment is that advocates for "bootstrapped businesses" will have the same cognitive biases as the essays for VC businesses.

A few points:

* When I said, "I wasn't making any money," I meant not anything of significance. I wouldn't consider a thousand or two thousand per month significant, which is where I was at for a while. If I can't live on it and hire an assistant to delegate some work to, it's not significant. If I couldn't pay admin fees it would be a failure. But yes, I did redefine failure for bootstrapped startups because I think it's valid. VC-backed startups don't have this flexibility / time to learn the hard way.

* In regard to being determined, I know this sounds like survivorship bias but you really do get better at what you're attempting to do by sticking with it. Your product gets better. Your personal network grows. You're more yourself in more uncomfortable/growth situation. Your odds of failure drop the longer you work on achieving something — not surprising.

I'd guess 99% of the people who have taken at shot at learning a second language fail (or give up or get bored) before they become fluent. Does that mean the 1% who made it and say "practice every day" have some level of survivorship bias? It's kind of absurd. Bootstrapping a business to success is like anything that takes time and effort and is based far less on good fortune than people think.

You parsed what I said pretty deeply and I'm not that smart of guy, so if there are flaws in what I wrote here that's understood - but you might understand what Im trying to get at.

>In regard to being determined, I know this sounds like survivorship bias but you really do get better at what you're attempting to do by sticking with it

This is survivorship bias. You're a survivor and can look back on the set of decisions you've made in the past to rationalize your current success. I've seen countless businesses linger for years only to shutdown later and sell their assets.

I find learning a language a poor analog here as it's much simpler and generally within your control. More effort is more closely tied to more results. Language learners aren't typically affected by seasonality, fads, or market conditions.

It's all of these other factors that make analyzing business success factors tremendously difficult. Because it's so difficult, it's easier and more straightforward to assume "I'm successful, so my decisions were the right ones" than "I'm successful mostly due to luck".

>, I know this sounds like survivorship bias but you really do get better at what you're attempting to do by sticking with it.

Yes, I agree one can get better at something by continuing to practice it. But that's independent of what "survivorship bias" means. It seems like you're thinking of "survivorship bias" as a measure of honest self-reporting instead of an unemotional mathematical bias. Your "determined as hell" example is not the same as "language practice" in the next example:

>I'd guess 99% of the people who have taken at shot at learning a second language fail (or give up or get bored) before they become fluent. Does that mean the 1% who made it and say "practice every day" have some level of survivorship bias? It's kind of absurd.

Survivorship Bias means ignoring an uncounted population (counterexamples) when considering an attribute. Yes, your language example would be absurd since it includes both the successes who practiced every day _and_ the failures who didn't practice every day. This matches our intuition; my son doesn't practice Swahili language every day so he failed at knowing Swahili.

Your "determined as hell" anecdote isn't like that. To remove sampling bias of survivors from your advice, we have to count the _other_ entrepreneurs (not you) who were also "determined as hell" but still turned out to be business failures. The _other_ entrepreneurs can also create "better products" and "grow personal network" and still fail at the business.

>but you might understand what Im trying to get at.

I understand your positive advice for a successful bootstrap business. I took issue with _how_ you presented it because it copies the same logic flaws of survivor bias that the article is complaining about. It seemed like you were so passionate about your success anecdote that you overlooked what "survivor bias" actually means. Therefore, you didn't notice that it also applied to your comments of bootstrap being "not hard" if one is "determined as hell".

Any lessons learned that you want to share? As the owner of a very small consultancy shop I know it's very hard to extract generalized advice from your own failures, but still...

Sure, although my experience is in the product business. some of these might be applicable:

* Find a niche market where there is a need and the prospects aren't broke. Eg, they have money to spend for the potential of ROI with your product.

* Break your revenue goal into concrete building blocks. If you need $x/year to make this a profitable venture, how many customers do you need?

* If you need a lot of customers (IMO, > 25) you need to think about both marketing and making your product and making it possible for them to self-service. Eg, driving inbound so you don't have to rely so much on outbound.

* Right down your goal and make a plan to achieve it. Does that mean emailing / calling prospects? How many per day? How many trials does it take to convert one customer? These are simple metrics which will eventually help make your revenue "predictable." Great books on that topic.

* Sales is #1 priority. SV is hyperfocused on brilliant engineers. In bootstrapping, it's all about money. I'd spend more on a great salesperson than an engineer.

* Make yourself uncomfortable. Don't like cold calling? Cold call anyway. You'll get good (or at the very least not half bad) at everything.

* Take care of yourself. You are your own vehicle to success. Your mind and body are business partners. Your bad habits can hinder you severely. 9-5 employees can come to work hungover every day and probably get by fine - but an entrepreneur can't.

* If something isn't working, change it. If something's working, do more of it. That includes pivoting product, market, etc.

* Don't get too elated or depressed. There can be incredible highs and crushing lows. Failure is a teacher but success is a better teacher.

* Be humble. Recognize that you're an idiot and that you don't know shit, even when you think you know shit. It's how you keep your mind open to learning new things. Your own ego can be a roadblock to success.

* Learn to delegate. When you start out you'll do everything — but eventually you will find you have low-level tasks which, if offloaded, would make you far more productive. That's when it's time to hire. Resist the urge to just do it yourself. Your ultimate goal is that you become the brain of the company and you delegate the physical work to the staff you've built.

There's a lot more but those are the ones I could rattle off.

These are all the things I do to build my venture backed businesses :)

Well said. Sales is of critical importance. I try to outsource anything that costs less than my own hourly rate.

Damn this is by far the BEST advice i have read. HOLY SHIT! I would like to THANK YOU VERY MUCH whoever you are. I have been grinding it out with my tiny start up and its making $3k-$5k/month. I was thinking i need money to grow and should seek funding but then i realised what VC would fund me? i got no secret sauce no technology no VC would give a rats ass about my start up and anybody can copy me at any time. I been at it 2 1/2 years and i think now i will follow your advice and double down my efforts and use the profits to pay a contract SWIFT engineer to build an iOS. Then grow the thing AS much as i possibly can. I'm sorry for hitting you up for a sale but i always gotta be selling! If you have a tesla and would like to sell, please consider sending it to me: https://onlyusedtesla.com/

> Most bootstrapped businesses fail as well.

Depends how you define failure. The statistics I've seen suggest that around half are still in business 5 years after formation.

Most bootstrapped businesses fail as well.

While that's very true of startups I think there's more to it. A lot of businesses are founded in order to scale up to something that can give the founders a decent exit. If you start a business with the explicit intention of only growing to a point where it's going to give you a good lifestyle you're a lot less likely to fail. You don't need to take the risks necessary to grow. You can focus 100% on the business instead of distractions (for example, raising a round.)

You're still more likely to fail than success, but not with the sort of failure rate that startups typically see.

>> If you don't mind being relatively unheard of and only want to make enough money to be independent, then it actually isn't very hard to bootstrap a business.

100% agree agree agree.

TechCrunch, tech conferences and forums (including HN) fuel the startup porn aspect of fame. And the spotlighting of a handful of companies make people think that billions of dollars are out there for the taking. You aren't going to be famous and you don't need a billion dollars. Once you accept that your lens on the world comes into a lot more focus.

The last time I was in a position to raise money one of my friends said, "you're growing and there is a lot of profit...and you get to decide what to do with all the profit, include putting it in your own pocket? Raising money means you are now on a salary and have a boss. Why would you do that?" He was right and I've never thought about raising money for that project or any other ever again.

Most VC founders wait 7+ years for their cash-out moment, if it ever comes at all. Build a profitable business and stop worrying about the adoration of people you don't know. Make the money, keep the money.

> the absolutely batshit advice and business building strategy that comes out of the world of venture capital and tech accelerators

Part of this includes the whole concept of rushing a 'knocked together' MVP out there. It all depends on the product. Yeah, if you have an app to share cat pictures on a social network, then who cares if you have a flakey cobbled together app that doesn't upload half the time. The consequences are minimal.

If you have an app that is responsible for administering a life saving drug on a periodic basis, then you had better be sure the thing is polished, tested and reliable before even beginning to test it in the market.

The whole 'lean' system sausage factory is just that - creating a mass production line of ideas where only less than 1% float to the top. Very often only based on the fact that the creators shouted louder than others who may have had better (or more boring) ideas.

What's the alternative? Overfunded waterfall-style development on fewer things nobody wants?

The lean startup model seems like it works, even outside of tech. Even yes, in medicine that can kill as a failure mode. Drug development, for example, consists of trying a bunch of plausible things. (And worked until the effective, simple chemical space was exhausted.)

Drug development certainly does not consist of trying a bunch of plausible things on the public (or even on small carefully picked groups of human trial subjects, not until after extensive tests and trials have been exhausted). And thank fuck it doesn't.

Besides, it's rather disingenuous to act as though development falls on a binary between "overfunded waterfall-style development" and the current mess at many shops.

My apologies if my comment came off as disingenuous or overly binary. I'd agree you could take minimal MVPs too far. (Also, the phrase "minimal MVP" is RAS syndrome but how would I say it otherwise?)

What I haven't seen is "the current mess at many shops", and I'd be interested to hear about your experiences. I've only ever experienced failure by overplanning personally, and the stark divide between progress in things that can be cheaply iterated on and things that are expensive, preplanned disasters seems important to me.

This leads me to naturally think that if I wanted to make progress in e.g. medicine I should start by finding out how to make it cheap and iterable, rather than come up with comprehensive, well-funded plans.

...and moving a very small fraction of those drugs into long FDA trials.

I've been in a couple of those trials.

They make it very clear that the drug has already been tested extensively and also do fairly extensive testing on you to make sure your vitals are within the "normal" range.

I got a very clear impression that they're not throwing pills at the wall to see which one sticks.

I'm also working at a startup that's had three names and just as many products, funded by a VC who's just going to start another one with most of the same people whether this one works or not.

It doesn't matter one bit if this startup fails. The drug was pretty clearly indicated to work by the time a person swallowed a pill.

Different worlds.

Or if you're, say, building a system for self-driving cars.

This Silicon Valley arrogance and recklessness has killed people. It's not just academic.

Don't understand the hang up about this. Self Driving cars were always going to kill people. It's not possible to be 100% safe but even a minute improvement over human drivers would save thousands of lives a year.

Uber has had a fatal accident with only a few million miles driven when an average human will go more like a hundred million miles per fatality.

I don't understand why people insist on believing that they are already more safe than human drivers when the evidence suggests that they are currently much, much less safe.

I'm sorry where did I imply they were already safer? I think it's obvious they will continue to improve and surpassing human level is not far away.

Nonsense. Uber's self-driving cars currently have something like 30x the fatality rate of an average driver and that is in cherry-picked locations that they chose for testing and with a human driver constantly watching it to take over when it does something wrong. There is no evidence that they are anywhere close to surpassing human drivers.

Given a sample size of 1 I don't think there's much to be discerned form the data at all either way.

> There is no evidence that they are anywhere close to surpassing human...

Famous last words.

And even a slight worsening could be a worthwhile tradeoff for the economic gains it would give us.

> the absolutely batshit advice and business building strategy that comes out of the world of venture capital and tech accelerators

If the floor comes out (e.g. capital dries up), then tech founders will find themselves in "the emperor has no clothes" situations. The potential externalities of this are varied:

- A younger entrepreneurial business demographic who build initially unprofitable businesses based on venture capital and can no longer start businesses because there's no capital

- Entrepreneurs resorting to fraud (see -> Theranos, Hampton Creek, etc)

- Mental health issues, broken families, etc for people "crushing it" and realizing they're left with nothing 7 years later (see -> https://twitter.com/dhh/status/869361585613139968 )

- Fatigued investing community who no longer deploy capital because cap tables are screwed up

This is very true. Building on this, entrepreneurs also seem to make the mistake of taking a business idea that may work for something small, and pitching it to a VC. This usually involves taking an otherwise modest pitch and selling it as something that could change everything, forever.

I worked for a startup that did this. We took $2M to fund an idea that was going to change everything, but only in the .001% chance of hitting it big. Investors bought that pitch up. Over the course of 2 years we spent that money and tried to position the business to hit it big. In doing so, we forwent plenty of opportunities to grow the business incrementally; any time spent not aiming for the fences was a waste, because the business didn't have a future without a home run success. Taking $2M up front meant we had to meet absurdly high expectations for investors to return that money. Moderate success of the business meant a failure for investors. Even a sustainable business was a failure for investors. The only success story would be a 10x success.

In the end, we took a swing at the fences and missed. We had opportunity after opportunity to pursue revenue early, to grow slowly, to take on smaller customers. But we skipped all that because in the very beginning we lit a fuse by taking too much VC money thinking it would give us more freedom, but it chained us and ultimately ruined the company. Because VCs would rather have a ruined company than a 1x success (if VCs encouraged 1x successes, they'd see fewer 10x successes since fewer startups would even try).

In startup world, mediocre success meant failure. People don't understand that enough. You could make a useful product that people love, build a great company with great people and talent, attract customers, generate revenue and even profits. You could achieve all the hallmarks of a healthy, successful company and it would still be a failure in the eyes of some VCs.

> then it actually isn't very hard to bootstrap a business

As someone with a degree from Oxford (however much that may or may not mean) I find it tremendously hard even six years into it.

I mean this in support of your business-building quest, not as a slight: the worst businessman I know is a Professor at the Wharton School of business. Your market, competition, and execution are orders of magnitude more significant than your in-classroom education. Keep pushing, you'll get there!

No offense taken. I just mentioned my education to give one data point about my background.

(and thanks!)

Here I am, 13 years in. Still hard, and getting harder even though we are bigger.

"The financial challenges remain the same, you just add more '0's to them."

And you can no longer cover payroll on your own credit card...

Lol. Well played (and been there).

the absolutely batshit advice and business building strategy that comes out of the world of venture capital and tech accelerators.

Would you mind elaborating on this? What advice concretely are you talking about and what exactly is batshit about it?

Not OP but I would assume one of the issues would be building and growing a business while having no monetisation strategy, but chasing growth at all costs nevertheless.

Or monetisation can in no way sustain the current and future costs of the business.

There are many examples, like twitter, whatsapp, homejoy, various food delivery services, springpad... etc etc etc

Sometimes the strategy works though, for example Google and Facebook.

It seemed to have worked pretty well for WhatsApp, no?

Growing up I had a friend whose dad was a regular business starter. At first he had his own computer stores, which was put out of business by CompUSA, who sold parts cheaper than he could source them. Then he started his own internet cafes, which the ubiquity of broadband put out of business. Eventually he moved onto various profitable online websites.

The point is he made money every step of the way. You never heard of his companies, he never took VC, and despite starting companies at the edge of what was modern technology he never ran a "startup" by Paul Graham's definition, and he still made fuck you money.

I think the point is that while the economics might be sound in terms of anticipated risk, risk management, and expected growth; the 'folklore' that arises to relate that learned experience in a convenient, easily understood way is often only a shadow - a poor construct - of the experience that actually keeps these systems running.

Whether that's because 'trade secrets' or 'elevator speeches', whatever the reason, information gets lost, exaggerated, and permuted along the way. So sometimes it can easily become kind of batshit because of how far removed from reality - the actual guiding forces that sustain that economy, are.

I don't know if you can point out concrete examples, because for every concrete example there's probably a successful startup out there that followed that model. But the exception doesn't prove the rule in systems that are fundamentally opened, like the relationship between culture (consumer, startup, academia, enterprise) and the economy. That's because those rules (advice) influence the system.

You can't simultaneously account for the effect the advice will have on the system while following and dispensing the advice, unless the advice and it's effects on the system are immutable. But that's not how capitalism works.

This is a great article on two standpoints - it provides important high-level "abstract" advice to new founders without sounding like Sam Altman (more on that later). Secondly, it criticizes the general human tendency to look for examples of success, read biographies, watch interviews and think mimic those traits in themselves thinking that they've won the ticket to silicon valley lottery. I'd like to quote Richard Feynman, "... for nature cannot be fooled." and the world is an insanely complex place with a number of external factors that founders have no control over. Doesn't mean we should all give up on starting a company, but it should be humbling to learn this and internalize it.

Every single advice that Sam Altman gives has survivorship bias written all over it. I have never learned anything remotely useful from his vague generic articles and speeches. I'd rather watch Drew Hudson's startup pitch or read about Chad Rigetti and his quantum computer venture. Or maybe learn how Instacart managed to photograph every single grocery store item by literally going to the store and buying everything. Sam Altman's interviews are fantastic and I thank him for that but I think he should stop marketing himself as a startup guru of some kind. Each situation is unique and my personal opinion is that particular qualities such as leadership, convincing abilities, and charisma cannot be learned. It can only surface as a result of how one thinks and acts. This is why leadership courses are bullshit.

> Every single advice that Sam Altman gives has survivorship bias written all over it.

Every VC will talk like that, for a reason: they benefit from the view that your startup will be successful if you are smart and work hard. A VC doesn't care about the fate of the individual founder. They make money by spreading their risk and having many founders trying hard.

Sam usually puts a disclaimer into his articles. From his playbook: "A word of warning about choosing to start a startup: It sucks!...Joining an early-stage startup that’s on a rocketship trajectory is usually a much better financial deal."

I doubt these disclaimers work well when a crazed founder has dollar signs in his/her eyes. But he tries. Yes, VCs profit from a larger market of risk-bearing founders, but Sam was once a risk-taking founder too.

I always found this really interesting. If you join such rocketship often you find people swimming in their survivorship bias often coupled with general inexperience .

'The reason we're successful is that we got every engineer the top of the line MacBook! No one else does that!'

'The reason we're successful is that we got every engineer a reconditioned MacBook off of eBay! No one else does that!'

Though I have mixed emotions about the book, Gladwell's "outliers" does a good job reconstructing some famous success stories with a focus on the ecosystem and environment rather than (or on top on) the efforts of a single talented human.

->Every single advice that Sam Altman gives has survivorship bias written all over it. I have never learned anything remotely useful from his vague generic articles and speeches.

Every and never are really strong words, I don't think he intend to market himself as startup guru at least intentionally.

He is young and already had 10 years of experience of watching top notch startups rises and falls. Main message I am getting from him is practically advice about start-ups. like, do this and don't do this, then you may succeed.

> like, do this and don't do this, then you may succeed.

That's the whole point. It's 'may' and not 'will' or 'probably will'. Buy a lottery ticket and you may win.

Real advice should alter your chances, not leave them exactly the same as before.

If Sam Altman were a successful serial entrepreneur with a track record of one hit after another some of what he writes might be worth emulating but on the whole it is like observing the lottery playing crowds and noting that the ones carrying umbrellas ended up winning more lotteries so therefore you too should take your umbrella with you.

I think he's made a lot of money investing in startups so characterizing him as totally passive is a bit unfair. A more appropriate analogy would be something like poker, where there's some skill involved in addition to luck. But it's harder to teach investing in startups than playing poker

> I think he's made a lot of money investing in startups

That's not how I see it.

He got lucky being in one of the first YC batches, had his failing start-up Loopt acquired for a very slight premium over all investments up to that point.

Then for some reason Paul Graham made him CEO of YC. Now, Paul is anything but dumb so obviously Sam Altman has some qualities and quite possibly he will be instrumental in leveling up YC. But so far all that has happened is that he's tended the garden that others have built.

Making money when you start with a large pile of money is a reasonably sure thing.

I believe he made a good amount of money investing independent of YC before he joined. He was a scout for sequoia and invested from his proceeds from the sale of loopt. That doesn't necessarily qualify him to lead YC but I do think he has an angel investing track record independent of YC that is quite good compared to peers. But you're right, his relationship with PG probably had more to do with him getting his current job

Maybe, except poker has a fixed set of contexts. The hands you can hold, the hands your competition can hold, the winning and loss conditions, are fully enumerated. And you still can't assume that someone who did really well at a professional level can teach others to play at that level. How much worse in the land of startups when none of those things are fully enumerable or known?

If you want to draw that analogy, I think Pratchett's statement in Good Omens about God playing dice with the universe is more accurate - “God does not play dice with the universe; He plays an ineffable game of His own devising, which might be compared, from the perspective of any of the other players [i.e. everybody], to being involved in an obscure and complex variant of poker in a pitch-dark room, with blank cards, for infinite stakes, with a Dealer who won't tell you the rules, and who smiles all the time.”

God plays Calvinball with the universe.

I guess another way to think about it is with a sports analogy haha: he's like a coach of a professional team. Not all professional coaches were the best players, but many of them played professionally and were just good students of the game. That's maybe a better characterization of Sam: he played in the "big leagues", wasn't a huge success, but did well as a coach / angel afterwards

He is young and already had 10 years of experience of watching top notch startups rises and falls.

Is watching a few thousand baseball games enough experience to pitch at Fenway Park?

(I mean no disrespect to Sam; I've never met him and I have no knowledge of how good he is at advising startups. Starting and exiting Loopt shows he definitely knows more than most of us. I'm just skeptical about whether you can get useful experience of something passively.)

Data. YC has collected tons of data points since 2005. From more than 100k applications, roughly 2k batch startups (11 unicorns, several dozend +100MM startups), 3,5k founders, successful or not. Probably everything from founder CVs to week-per-week growth rates to the amount of office hours during a batch to VC deals/termsheet offerings after demo day to following series A, B, C rounds. Probably every single ratio you could imagine that might be somehow relevant. YC employs 50+ ppl, most of them engineers. They are basically a data-driven software business, with ML pre-selecting new applications.

So ... rather generic-sounding advice like 'do things that don't scale' or 'distribution is king' or 'make sth people love' is maybe as good and objective as it gets. These are probably the necessary heuristics to use while not being sufficient (luck, timing, talent...). From a merit-based point of view I'd say that SA is one of the top-10 global startup gurus. Because of the above.

> whether you can get useful experience of something passively.

depends on what you mean by passive.

Another way to figure this out in Sam's case isn't to look at his ability to pitch and physically play ball, but pick players and play money ball.

So the question that follows is then : Assuming one has the potential and skills to play money ball; is watching 10 years of startups enough to play?

PS. I also don't know Sam from Jack


Passively? He is leading YC, not a techcrunch journalist.

He hasn't been doing that for ten years.

Before that he was a startup founder, right? Then a YC partner.

He founded a single start-up when he was 19 that failed in its mission and was bought for a paltry sum (for YC, anyway). Other than that his experience in life and business consists primarily of having been born into affluence and all the privilege and advantage that brings.

It's so very Bay Area for someone like him to be president of a VC firm. His pre-YC resume wouldn't get most people past the resume filter in almost any other company, including a number that YC fund.

And most of his comments about business are not all that useful. They're exactly the sort of comments I'd expect from someone blessed with great luck and above average intellect but too little experience.

He was also a successful angel investor and I think a scout for sequoia before YC. I think he was more successful at the angel investing thing than at his startup

> do this and don't do this, then you may succeed

I like to think of startup advice as "necessary, but not sufficient". If you don't do the right things, you probably can't succeed. But even doing the right things doesn't mean you will succeed!

You don't believe leadership, persuasion, and charisma can be taught? You're denying the 1000's of years of study in rhetoric, persuasion, and logical reasoning going back to Ancient Greece? I agree with your post in general, but I think it's entirely reasonable to improve your interpersonal skills, emotional IQ, listening abilities, and leadership skills. Hell, even in scouts they know this, they improve upon all of those skills repeatedly as young men were given progressively more dutiful leadership positions.

Yes those things can be taught and they are being taught all the time. But, what portion of that really manifests into actual leadership and changing the world? You give an example of Scouts - did they ever become anything in their lives and change the world? Some do and some don't.

There is no definitive data that says you can go to a class, read a book or talk to someone and become a leader. In fact the opposite is probably true.

Leadership is an abstract quality that cannot, in my opinion, be actively managed. It just exudes from the person as they try to solve the problems, execute their vision, and build a rapport by delivering on promises. "Fake it until you make it" doesn't work actually work for most cases and neither do listening to lessons on leadership or reading a particular book.

But the world as whole, doesn't accept Luck. Not middle or even top management. Those who do, are actually founders, Bill Gate and Steve Jobs both talk about how lucky they are. I mentioned Luck in most of my interview and not a single one likes it. The older I am, the more I became aware how many of my success or wins , are not really because I am ten times better then others, yes it may have played a part, but there are lots of moving parts, and I am only a small part of it. In essence, I got lucky. There is no way, a human, or even machine could calculate what is going to happen next. The possibilities and probabilities are endless.

Right, humans seem hard-wired to discount/ignore luck, because it's hard to fit it into a conceptual framework. We prefer to create stories with clear causes even in situations which we abstractly know are random. (Ex: The myth of a "hot hand" or "I'm due for a change of luck.")

This extends to how they try to explain outcomes, so that the wall-street executive is either seen as a leadership-god or else as a corrupt shark, rather than a mostly-regular person with luck.

In closing, some humor I remember reading once:

> Most people won't admit how they got their current jobs unless you push them up against a built-in wall unit and punch them in the stomach until they spill their drink and start yelling, "I'LL NEVER INVITE YOU TO ONE OF MY PARTIES AGAIN, YOU DRUNKEN FOOL!"

> I think the reason these annoying people won't tell me how they got their jobs is because they are embarrassed to admit luck was involved. I can't blame them. Typically, the pre-luck part of their careers involved doing something enormously pathetic.

-- "The Dilbert Future", by Scott Adams

“Luck is a matter of preparation meeting opportunity” - Oprah

By and large you can't control what opportunities sprout up around you, but you can control how prepared you are to take advantage of them when they do. EG: you can spend more of your free time learning new technologies, so that you might at least be passable at them should the need arise.

I think the reason interviewers don't like talking about luck is because they'd rather hear what you're doing to help yourself at getting lucky. How do you best prepare yourself for the unknown opportunities coming in the future? Sure, its only half the equation, but its the half you get to control.

Generally I think you're right, but Seneca may have said that before Oprah did.

I agree luck is a factor. But you can increase your chances, by the choices you make, by perseverance ...

Such a nice article. But I have to note that the linked (2013) piece is even better: https://youarenotsosmart.com/2013/05/23/survivorship-bias/

It's also quite a bit longer, though I agree it's worth the read.

> You succumb to survivorship bias because you are innately terrible with statistics.

This one line is probably one of my biggest takeaways, and I wish we had a good term for our poor-intuition-about-stats.

Recently, I've seen it conflated with "innumeracy" here on HN, but that misses the point, as educating someone in statistics doesn't necessarily remove the natural intuition, which remains poor. We still need to be aware of it and the, not necessarily obvious, effects on our judgment it can have.

The massive amount of failed startups is caused by people who think this is a fun popular way to make money out of thin air.

If you're level-headed and sensible, you can do your analysis, make a business plan and figure out if your idea is marketable or not. Then, there's the business-doing portion, which is extremely complex, different from your technical work, and will make or break your company. But that only works if you have a financially viable, novel idea. Unless you intend to scam bad investors, you can't have a startup without real innovation, period.

The startups were meant to be a way for experts in some field to materialize an (actually) innovative idea with economic potential, and generally an idea that doesn't require absurd amounts of capital. This could be done in a typical corporate environment, but when the innovation originates from a single individual, the startup model allows him/her to better protect his interests and grow his business faster.

When running a startup became a trend (in many areas of Europe it did, I don't know about the Americas) there was a huge influx of people with no technical expertise, all of whom simple wanted to be a startup owner for the sake of it, with no plans or targets, just to be managers. These people fail over and over, sucking up EU and local government resources with them, and dragging the name of real startups in mud. Many of them are just the kids of wealthy individuals with connections. They leverage these connections to get investments and loans and waste 100s of thousands to millions on repeated failed attempts to materialize some retarded idea, like solar freaking highways or the 10,000th taxi hailing application on the market.

Survivorship bias is bad, but this situation would persist even without it.

> (in many areas of Europe it did, I don't know about the Americas) there was a huge influx of people with no technical expertise, all of whom simple wanted to be a startup owner for the sake of it, with no plans or targets, just to be managers.

I find this perspective quite fascinating, as it is something of the converse my impression of some of the complaints I've read about the startup scene here in the US (mostly SFBA/Silicon Valley).

Here, the complaint seems to be that (some) founders may have technical expertise but lack the business/managerial savvy (or, again, with no plans or targets) to turn it into a real product that makes money.

I'm unsure if either criticism, lack of business or technical acumen, is actually fair or accurate, but it may be a subtle indication of an over-focus on one over the other (by the investors) in each of our regions to the detriment of overall innovation.

To me survivorship bias is looking at someone who made it. Assigning some charateristics to that person. i.e. Hard working, motivated etc etc.

But, forgetting that lot's of people who didn't make it also had those charateristics.

Then you have to factor in luck. 2 Entrepreneurs with the same work ethic and idea but one gets lucky. The difference in money generated can be huge.

I think you should stay in school. I know it’s harder t justify in America where it’s really expensive, but the benefits can be immense if you use your time in the educational institutions correct. It’s really the most secure path to a meaningful life if you’re willing to work hard at bettering yourself.

That being said I think the article stumbles a little. Where there 1000 Steve Jobs and Wozes working on a personal computer at the time? Where there 1000 people working on an global online payment system when Musk was building PayPal?

I think timing is important, but I don’t think luck makes that much of a difference. You don’t obtain connections by being lucky, for instance, you obtain them because you want them and work hard at it.

"You don’t obtain connections by being lucky, for instance, you obtain them because you want them and work hard at it."- I have to disagree. You have little control who comes into your life outside your 1st and maybe 2nd degree of separation. You have almost no control over who will be assigned your freshman dormmate, your instructors, or even your boss. You can't work hard to be put in the same dorm as mark zuckerberg. That is just plain flat out luck. However, smart people are attracted to other smart people and that is why smart people tend to "Create" their own luck. They dont posses any special ability, per say, other than a desirable trait that others like them want to exalt.

The luck really starts with your parents and grandparents, as their financial status generally determines your early social network. http://philip.greenspun.com/bg/ discussed some of Gates' early advantages, like having a trust fund to fall back on if risks don't pan out.

But that excludes countless counterexamples with people being born with nothing and becoming self-made millionaires. Again that's not luck its a product of work.

And that excludes even more countless examples of those who work hard and fail.

There are vastly more of the latter than the former.

Work - if it’s the right kind of work - can only ever increase your odds. It can never guarantee success.

What’s needed is hard data about the relative influence of market fit, hours worked, personality, network and wealth of origin, and so on.

Analysis paralysis is arguably as dangerous as survivorship bias.

Of course winning the genetic lottery bestows tremendous advantage. And of course luck is a huge variable.

But the intersection of luck, preparation, and willingness is serendipity. And every successful entrepreneur has that in spades in the end, regardless of where they started.

Sure that sounds like another empty platitude, but the fact of the matter is there is no formula or one simple trick.

I fear that winning the genetic lottery is more than just a simple advantage. In the case of being born into a wealthy family, it turns the situation from "it's all up to you to succeed" into "all you have to do is not fuck up." Since tremendous luck, good or bad, is relatively rare the condition from generation to generation just doesn't change much - regardless of hard work - except in so much as society itself grows wealthier.

>You have little control who comes into your life outside your 1st and maybe 2nd degree of separation.

I have not found this to be true, personally. I've established my own network over 30 years by having an attitude entirely opposite to that which you profess is the natural order of things - if you work hard to network, you can make contact with a great deal many people, and luck has nothing to do with it.

We're not all pushed around by the fates. Some of us, push the rock hard. There really are a different set of results in both spheres.

For someone like myself who is still young, what would you say are the keys to becoming a better networker?

1. The less judgemental you are the more people you'll allow yourself to talk to.

2. Lots of people want to be talked to but they don't want to take the risk of approaching and getting rejected. Master the art of gracefully accepting rejection and you'll feel confident in approaching people.

I totally agree with you - prejudice is the #1 enemy of the positive networker.

I would add this: Learn to experience things from the perspective of the other person with whom you are attempting to network without giving up too much of your own personal identity in the process, and you will go very far.

The counterargument to this is that enterprise sales is a job. If you've worked alongside these people you watch them do it.

They go to events, identify useful people, create a relationship with or without introductions, use them to access decision makers...

The fact is, people make a career out of creating connections from nothing and using them to land buisness. A lot of people.

Yea that's a valid point. The thing about sales that always bothered me is that the psychology often borders on sociopathy. How can you love your fellow man while simultaneously seeking to get the better of him?

You're not seeking to get the better of them, you're seeking to get a fair trade and for startups one that will save them a significant amount of money.

> You have almost no control over who will be assigned your freshman dormmate, your instructors, or even your boss

You can, however, randomly go up to people and say “hi”. Most people don’t do this. (Which is why it works.)

Randomly... First of all, it's rarely really random -- you will approach people based on your internal biases, whether you're aware of them or not. And second, even if you did manage to do it truly randomly, that just proves the GP's point.

Young people usually don't look for "secure" though. But I would say if you don't have rich parents it is even the quickest way with the most success. Even the chance to a successful startup increases if one follows a stable career path in the beginning. Everyone who succeeded without it, like Bill Gates, actually succeeded because they had a secret joker through the success of their parents.

I myself had to become 30+ before I understood this, though.

Agree 100%. Reminds me of what they say in Wall Street, to be successful dont try to time the market, put time in the market.

When I dropped out of school I thought the odds on the outcome of staying and graduating were worse than leaving. The bad outcome having been getting an average salary at a job I was unhappy with. I dropped out with a tiny income from my business, and the extra free time made a massive difference within 12 months.

The biggest lesson I’ve learned was to think through problems myself. There are a lot of degrees and dimensions to success. If you live in a first world country, are physically healthy, and everything is working with your brain, you are 90% of the way there. Exceptionalism is not a requirement to success.

That in itself sounds exceptional.

I've worked in a lot of coworking spaces where inevitably someone starts a group to talk about the harsh realities of starting a business. It sounds like Gambler's Anonymous.

Surprised to hear only 70% fail. Those odds actually sound pretty good. I've made enough in the startup world to pay the bills for a decade, but probably below my market rate at a big tech co.

I don't regret a thing. Each time I get a bit further into the wealth creation game. But I also wouldn't recommend it to anyone, and actively work on dissuading new founders from trying.

It is a relief to see this view reflected.

In an efficient market for business talent, VC should not exist. They could be seen as in effect talent arbitrageurs. Something about business culture over a certain size creates a massive inefficiency where someone with the level of skill required to build a successful product is not snapped up by existing firms.

Personally, I like anyone who takes an arbitrageurs view because to me betting against bullshit is a costly signal of outlier intelligence.

However, viewed this way, it's not about kids in hoodies. College dropouts were useful because the key signal (ivy certificate) that gets them picked up by a firm that would smother their productivity, had been suppressed. That they were selected for the school in the first place was a great negative filter as well, so betting on the remainder is practically a no-brainer.

Given the exponential distribution of returns, from a statistical perspective, there isn't much value in explanations for failure because it's by far the most probable outcome anyway. e.g. who cares why startups fail, %90+ of them necessarily must, and merely avoiding a subset of failure scenarios is not equivalent to success. In that environment, there is only one necessary condition for 80th+ percentile talent to maintain: survival.

Being the person who climbs up that return curve means surviving the twin perils running out of runway, and of being snapped up by a firm for a job you were originally more suited to anyway. This second peril is the other major attrition factor that prevents people from benefiting from survivor bias.

The business of VCs who throw money at geeks in hoodies is to mitigate the former and get exposure to that overlooked subset of the latter. Given the remaining members of that sample, the people who also selected out of being snapped up for jobs will be necessarily over represented - and on that very steep curve. I think this explains why investors like colorful founder stories, because they are (consciously or not) looking for how the founders 1-stddev+ talent signal might have been suppressed.

Good news is: the confluence of a) an exponential distribution of success and returns, b) the guaranteed competitor attrition from failures, and c) the further guaranteed competitor attrition from the job market - survivor bias may be less of a bug, and rather, a feature.

Are start uppers copying how Carmack started to do software? I don't understand why there is so much need for raising money when all you need is a roof and a computer to do this job.

I know that not everyone is talented like carmack, but sometimes I wonder if money is not spoiling and troubling the IT sector.

So in fact a startup might think it failed because it doesn't have enough money or because investors don't want to raise funds, but the reality is that software is so cheap to make, I really don't understand the world of investing in IT.

Survivorship bias is a problem in science, too, IIRC. Eg, "I looked for evidence that eating onions causes cancer and found nothing" doesn't get published, but the "successful" study does. I remember hearing of efforts to make researchers state their intended research up front and publish the result either way.

Similarly, shouldn't business schools should be eager to do studies following new businesses, bootstrapped and funded, and analyzing their success or failure? YC and bootstrapping communities could provide a lot of interviews if they wanted to.

(This seems obvious and has probably been done.)

I would like to see an analysis of startups that passed two years of age. That sounds to me like enough time to learn how it feels and learn and experience most of the issues in running a small business, and not enough time that it's "wasted".

If it turns out that 70% of 3 or 4 year old startups fail, then I'd say it's a seriously bad tradeoff.

There's always this perverse focus on who is making this thing, when there should be a focus on who is buying this thing.

Seems like the number one indicator of success in many of these businesses is when people actually purchase or use the product.

"In the same vein, a stunning proportion of partners at VC firms graduated from a handful of tony universities, as if the seal on a person’s diploma were what indicated investing abilities. (Granted, the incidence of leveraged social connections and postgraduate degrees may amplify that trend.)"

Or it could be that these "tony schools" are primarily designed to select for hyper-ambitious people with high IQs. Crazy, I know.

Considering around 1/3 of students at schools like Harvard are Legacy Admits who would never otherwise get admitted on their merits, I don't think your argument holds much weight. You seem to be admitting that they are selecting for elite education, if they were selecting ambition and intelligence there are countless people like that who went to a big State school instead of Stanford yet they are not represented in "Elite" firms whether that's VC, IB, Tech etc..

So was Elon Musk lucky in all four of his main ventures? (zip2, paypal, tesla, spacex) Or can the luck factor be overcome if you are good enough.

I think previous failures and successes up your chances for success in subsequent adventures. The failures teach. The successes teach and come with a payout. The subsequent ventures make it easier to retain control with your private money, as does the trust gained from previous successes, and your own experience.

I think Musk has improved his chances of success by the principles above and is improving his future chances by consolidating technologies in his companies. Solar City, Tesla battery technologies, The Boring Company's use of Model X chassis, The Boring company using the Hyperloop concept?

Of course, none of this would be possible without some runway, either investors or private money.

There is more information in Success. If the likelihood of success is very small you've potentially found at least some causal factor to help succeed - the needle in a haystack. If you fail you've found one of the million reasons you can fail. It's simply less informative.

It's not being lucky or being good. It's being lucky and being good. Nobody can succeed without both.

Of course there are occasional exceptions where somebody is successful purely through luck. However, very rarely is somebody successful with no amount of luck at all. It's hard to pin down just how much of an influence random chance has had in this chaotic world.

However there are plenty of very good and hardworking people who have yet to overcome a lack of good circumstances. The world is filled with failed geniuses.

I guess there's some survivorship bias for startups if we define startups as companies that take VC funding and intend to grow really fast, but those are less than 0.5% of all companies. The fact is that people with domain expertise, financial discipline, and a real product or service manage to do quite well in business and the success rate is much higher (I believe around 50%).

Do you have a reference for that number? That sounds really high.

Yes.[0] See item number 6. Anecdotally I know a number of people who have started service businesses and they were all quite successful. They don't have FU money and they probably have to keep working in some capacity, but they seem pretty satisfied.

[0]: https://www.sba.gov/sites/default/files/advocacy/SB-FAQ-2017...


I agree about some services businesses. Hell, I'm running my second successful consultancy, and have been known to recommend to people not to pursue the standard startup route.

Still, considering the sheer number of e.g. restaurant openings, small store openings, etc, which I assume have a relatively high 5-year failure rate, made me suspect the number was high.

I do wonder how your reference counts an establishment surviving after 5 years. E.g. if I start a company of 1 person (myself), and I keep it around after 4 years, but I'm working a full-time job besides, I understand from the reference that it would be counted as a firm that is still going after 5 years, although I wouldn't classify it as such. But I might be reading this wrong.

Now my question is, in all of those success cases, what was more important, luck or work and dedication? Because if someone wins the lottery, it doesn't matter if they bought 100 tickets every day. Luck would still be the main factor of the success. Would also be the same for these companies like Facebook and Google?

You should be able to 'afford' to fail? If you can't afford to fail then the risks are possibly too high.

It will likely set you back decades, destroy your finances and credit, possibly impact your family life if you have one or further the possibility of starting one and take a serious toll on your health.

I wonder if there are any stats around probability of success (measured by not going bankrupt) at different levels of revenue?

For instance, what percentage of companies that are able to reach $5k monthly recurring revenue go under then the same metric at $10k, $20k, etc etc

Anybody seen anything like that or have an idea?

"How I won The Lottery" (joke tech talk) https://www.youtube.com/watch?v=l_F9jxsfGCw

The real problem is that 100K and some advice sets up the vast majority for failure right out of the gate. Also weeds out the flakes, so there's that, too.

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact