Hacker News Reader

I agree. Perhaps some of the root cause here is a faulty premise. The idea that "25% of the companies are on a trajectory that could lead to a multi-billion dollar company" feels wildly optimistic to me. (And I say that with full respect of YC and the top-teir companies in the program.)

Strong early growth is a very positive sign. But growth is a funny thing and can be hard to sustain. The true market test may come later... from when you are trying to grow a $50 million dollar business into a $500 million dollar business.

Relevant xkcd: https://xkcd.com/605/

> The idea that "25% of the companies are on a trajectory that could lead to a multi-billion dollar company" feels wildly optimistic to me

The key word in that sentence is "could", not "will". I'm pretty sure after all these years YC have enough experience and data to identify the "could"'s. But they're still just "could"'s. Doesn't seem "wildly" optimistic to me, just optimistic, based on previous track records.

Sure--the vast majority of ideas just can't support a multi-billion dollar business.

But a lot can, and poor execution is why they don't.

Don't you feel the least bit of reservation selling young kids on the idea that they're likely to create a "multi-billion dollar business" by running through a grist-mill of a program that -- by your own admission -- isn't sustainable for most of them?

The vast majority will not achieve that goal (you don't need them to, either -- for sustainability, all you need is a few winners), and in the process, I can't help but feel that they're stunting their career, technical, and personal growth by approaching critical period of their lives with the belief that your economic model of the universe applies equally to them.

No, I don't.

We try to be up front that they have a high chance of failure. But I truly believe that there is nothing better for their career, technical, and personal growth than doing a startup they're passionate about (we try to filter out the people that want to do a startup for a resume item).

Also, one of the advantage of YC is that if the startup doesn't work out, founders have made a lot of connections in the network and can usually find something interesting to do next.

As I've said before, I think the actual risky path is not pursuing what you really want to do and then spending the rest of your life regretting it.

Do you think, its because YC didn't change the actual (slacking) nature of founders in 3 months?

Peer pressure once there is no longer there and you can't replicate that easily.

I don't think it's fair to use the term slacking here. At worst they're people who focus on the wrong things and don't prioritize their resources optimally. Considering the number who burnout, they're likely working very hard, just not on the things that will matter in the end.

I also think it's real and sometimes hard work, like migrating from Angular to React which could make a company grow immensely, especially if even more time is invested in blogging about the process.

I don't think migrating from Angular to React is real work. If it takes a long time, it's probably best to stay with Angular. A failed company using Angular is the same as a failed company using React. Better to spend your time iterating on the product and acquiring customers. If you succeed you can always migrate later. It will cost more, but you will have more resources too. Doing something like that in an early stage startup just has too high of an opportunity cost. It's exactly the kind of fake work Altman is talking about.

I was trying to be sarcastic but in hindsight it was a bit too plausible around here heh.

I was hoping you were being sarcastic. The trouble with sarcasm is it's indistinguishable from foolish sincerity, which abounds on the interwebs. Without facial or voice cues, there's no way to know. That's why you have to use the /sarcasm tag.

I had an idea 5 years ago for a company. When I had built the right experience and the legal environment changed to support the business I left my job and pursued it relentlessly.

We added a few people (part-time) raised a little money, but ultimately we couldn't get the product market fit we needed to feel comfortable that we could grow and keep our integrity (we were a financial product). A few weeks ago I shut down the business and I'm about to find something new.

The fact that I lost a year of income is completely irrelevant vs the regret I would have faced had I never made the leap. Even without YC my network has strengthened considerably and I learned more in a year than I thought possible.

I have read a ton, watched all the startup school videos, talked to a lot of founders and tried to understand how to do this well. The fake work problem is real and even while paying extreme attention to it, often times I would waste a few hours doing something that would not increase the probability of success.

Code and talk to customers. It's so, so crucial.

Starting a fast growth company is not for everyone, but for those that want to do it, YC's information is gold. Thanks for everything you guys have put out.

My impression of YC's marketing is an appeal to hubris (and often, the hubris of youth). Most will fail, but none believe it will be them, and in the process, they position themselves outside of an environment in which they'd otherwise have access to people with technical and sustainable business experience from which they can learn what (and what not) to do.

I'm a lot happier having spent 10 years building a sustainable company doing what I really want to do, than I ever was at exciting startups during the v1 .com bubble.

Getting to that point, however, required putting in a lot more than 3 months -- or 3 years -- of backbreaking effort.

sama's response is correct. There's no real downside to doing YC when you're young. If you succeed, you get rich. If you don't, you can probably find work at another YC company that is succeeding. If you can't, then having YC on your resume will probably help you land a job at another startup. No YC founder will ever go hungry.

There's a long, long road between not "going hungry" and maximizing your individual potential, and anyone that survived the first .com bubble unscathed knows that the latter has far more lasting power.

I'd posit that very few who go through YC stunt their career growth even when failing.

The median age of YC founders is 30. These aren't the pliable, simple-minded children of your caricature. Many of them are accomplished entrepreneurs who understand power law dynamics perfectly.

Further, I think nearly every founder who has gone through YC would report that, succeed or fail, they are net-benefited by the experience. Rather than stunting an individual, the exercise of founding a startup is probably one of the best ways to rapidly increase your skill-set over a short time frame.

I would agree with you that attempting a startup is a net positive career, and even life, experience.

I would disagree that you should grind away for it on years. If YC is confident that hustle/effort is the problem, they should be hiring engineers directly, pay them market rate, and keep all of the equity in their projects to themselves instead of having founders shoulder the risk.

I think a lot of your replies and central argument is very thoughtful, but the one thing to consider is that successful startups are the result of unreasonable exertions of will, unreasonable amounts of luck, and an unreasonable amount of risk.

If it were reasonable, I don't think the value created would be so great.

While I don't disagree with your points on a macro, startups are very specific entities, and their success is contingent upon founders (and team members) who are willing to do unreasonable things. And also be extremely lucky.

Thank you. I agree with your points as well.

Isn't it easier to take risks when you are younger? So even if it's not sustainable, wouldn't it be better advised to do something with fewer chances of success early one in life ?

It's easier to take risks, and it's also cheaper for YC.

Your chances of success, however, are much higher with more experience and resources at your disposal.

The people he's telling to grind are those that have millions in funding. So you work hard, pay yourself, and if it fails you had a great job and a great experience, and you can go get a better job than you could have when you started.

Sure, you worked harder than someone who has a boring 9-5, but you're not turning yourself an indentured servant until the end of time. Your only risk is the additional time/effort.

You say a vast majority can't, but then go on to say that a lot can.

What do you define as "a lot"? What do you define as a "vast majority"?

It would be interesting to put together a list of these "good idea, but poor execution" ideas. It would probably be a good list of things to try again.

Even more sadistic, have an entire batch dedicated to ONE of those ideas and see which company has the least poor execution.

That's what you get from an X prize, usually. It would be interesting to see an X prize that also had some base level of pre-funding for all entrants.

In the world of business, good ideas are literally priceless - in that they have literally no value. Good ideas aren't even that important to a business. Good execution is what matters. Even sub-optimal ideas can be billion dollar businesses. I still think that the basic idea behind twitter is only mediocre, but it is so excellently executed that it is a joy to use. The same applies for pinterest. They are just not great "ideas." But the execution is phenomenal.

Let's put it another way. When you say:

>It would be interesting to put together a list of these "good idea, but poor execution" ideas.

I read:

>It would be interesting to put together a list of these "execution plans but poor execution" ideas.

There is absolutely no way to separate out the "good idea" from the "good idea but poor execution" because business is the art of execution.

But maybe I'm crazy.

yes and no. If as a game you were given eight years to read and learn (but not take with you) whatever documentation you wanted about technical knowledge (only), and then sent back to New York in 1860 with $1,000,000 in 1860 bills and the goal of the game is to make as large a fortune as you can in 20 years... then it should be obvious to anyone that the ideas you would have are incredibly valuable...literally priceless. They represent without exaggeration trillions of dollars of research over 155 years.

You can patent entire industries and fundamental advances. The hard part is just choosing what to focus on, that you can actually do in 1860 starting with $1,000,000 and maybe the patent system (though it's probably unfair that you're using a time machine - this is likely quite immoral as you're not the true inventor of anything.)

And then doing it.

In other words: the outcome of the game is about execution. But what you're executing is idea. The two go hand in hand. You have a real chance to become worth 2015$ 1 trillion within the 20 years - or generate a one-million fold return on the million. You can be worth as much as entire continents. But you have to do it right.

For starters, you have to be someone whose brain can ship those technical ideas back to 1860 in one piece. This is quite similar to someone with an idea today. Not everyone qualifies.

While you are mechanically correct, your ideas would absolutely be hugely valuable, you have the benefit of already knowing what macroeconomic trends and historical events will happen.

Put another way, you could probably make exactly the same amount of money, if not more, with a "back to the future 2" scenario where you had zero ideas, but had a chart of various events that you could wager on or invest based on.

When we look at it from that perspective, it's only execution - can you correctly execute about the information you already have.

Soooo - yeah, it's not a great analogy. In general, having a perfect knowledge of what happens makes for bad analogies.

Well, I said you can only use technical knowledge. If you really want to be anal about the analogy then the game is you can only dictate technical knowledge to someone in 1860 (or I guess show them diagrams, but only of technical things - you are not allowed to talk about anything else) and give them starting capital. Then how well you do at the game depends on the ability of the person you picked to...execute. (As well as understand the technical knowledge they've been dictated/shown) It's not hard to understand what I'm getting at. But if you pick someone who can't understand the ideas, you won't do well at the revised game.

Why this is a workable analogy is because this really is similar to the position that founders with technical ideas are in. They have this vision that something can more or less work - but even if it does they still have to execute on it. That's my point.

Everything you dictate in this revised game is 100% guaranteed to work, because you're reading it straight out of technical manuals for the scientists who build these things today. But that still doesn't make a company that manufacture these things pop out of thin air. You can transfer a very large part of a $1 trillion in R&D back in time -- the ideas; but you won't suddenly have a company as a result. Someone still has to make it all and sell it all.

As an example, even if Babbage had invented modern transistors, he didn't "almost" create Intel and IBM. These companies as such are quite separate from the idea.

"and poor execution is why they don't"

Luck and timing no doubt play a very important role in addition to simply poor execution. Also even excellent execution doesn't take into account how the market will respond to what you are making or trying to sell which can't always be determined prior to actually going through the motions.

Comparing those numbers to the 5 and 10 year survival rate of more typical small businesses (50% and 33% respectively[1]) puts this into perspective.

This suggests that creating a business with the intent of not exiting, but retaining and growing it gives better odds. While a more typical small business has its own stressors and certainly can have long hours, in my experience the level of intensity is much more sustainable.

The down side - you're much less likely to walk away with large sums of money. On the up side, you're much more likely to create a sustainable venture that has a lot more success paths than the binary 'get acquired or go public' routes available to vc-funded startups.

[1] pdf - https://www.sba.gov/sites/default/files/FAQ_Sept_2012.pdf

Many of those small businesses that survive are simply the founder buying themself a job of some sort.

It's no surprise that someone who starts a landscaping or property management company, where the prime thing being sold is hours of labor, would have a higher 5/10 year survival rate than a company selling bits, ads, or other types of digital engagement.

Small businesses are defined as 1-500 employees - I suspect that includes a fair range of company types. Lumping them together as a single thing strikes me as a bit disingenuous.

A quick search turned this up:

http://smallbiztrends.com/2012/09/failure-rates-by-sector-th...

This shows all business types following roughly the same trend. 'services' [which would include software and property management] have the lowest 5 yr survival rate at 36% and mining the highest at > 50%.

This is data is from the 2010 census, examining businesses started in 2005 - likely a different methodology than the SBA used. This which probably accounts for the discrepancy from my SBA stats in the original post.

It's also arguable that Internet businesses tend toward a few winners take-all scenario and that drastically increases the tech start-up failure rate. Whereas there are 150,000 convenience stores and 72,000 pizza restaurants in the US.

The number of businesses that can be supported in the physical world is dramatically higher than for Internet services. Why? Geographic restriction and franchising.

There are going to be two dozen major cloud hosting companies, and maybe 50-100 smaller players. There are 7,000 frozen yogurt shops.

>you're much less likely to walk away with large sums of money

It would be great to have data on this. How many SMB owners make a few hundreds $K a year, for many years, ending up making a lot more than a very successful/lucky early employee making a couple of $M in one shot after many years. Especially in the software field.

1st article relies entirely on startup accelerators for stats, which means 92% of companies that enter startup accelerators fail, which is a totally different statistic.

2nd article just links to other articles on fortune which don't have anything to back them up.

So I'd be dubious of those figures.

I'm open to a debate on this. What's the failure rate only of startups that go through YC?

"Y-Combinator’s success is undeniable: the program has produced 37 start-ups worth over $40 million each in the past five years. In its first six years, 72 percent of Y.C.’s 249 start-ups raised money after Demo Day, and today, the average value of a Y.C.-financed start-up is $22.4 million. However, these statistics also reveal a grimmer reality: 93 percent of the 511 companies accepted by Y-Combinator have failed. Even more alarming, only 3 to 5 percent of the companies that apply to Y-Combinator are even accepted, meaning that only one in every 200 companies that applies to Y.C. eventually succeeds."

Emphasis mine.

[+] http://koltai.co/notebook/the-risky-business-of-entrepreneur...

EDIT: Please disregard this post. Left for historical purposes. sama notes that the stats are inaccurate.

EDIT 2: The citations in that link I mention above are from nytimes and business insider. They might be accurate, if a bit old, but require verification. Treat them appropriately.

Do all non YC accepted startups fail?

It appears, from data, that a majority of them do.

I don't know where that statistic came from but it's nowhere close to accurate. Less than half of the startups we funded in the first 5 years of YC have died.

Also, he average valuation of all YC startups through w2014 (i.e. with enough time to have raised a first round) is now over $100MM.

What were the causes of failures for those YC alums that did fail? (Again, not trying to be combative whatsoever. I'm genuinely curious, if you're able to provide the data without compromising YC from a business perspective).

Think PG has documented this well, but in summary

- cofounder disputes - market timing - not continuing to find growth

are top 3

Personally, I'm less interested than the average valuation and more interested in their P&Ls. Of course, that's not the game YC is playing.

Are there any stats for YC-funded startups that have grown to sustained profitability, vs continuing to exist through additional rounds of funding?

According to Wikipedia, their best-known funded companies are: Dropbox, Airbnb, Coinbase, Stripe, Reddit, Zenefits, Instacart and Weebly.

Dropbox - Not public, probably not profitable

Airbnb - Not public, might be profitable, lots of legal questions.

Coinbase - Not public, raised $100m+, no idea if it's profitable

Stripe - Not public, raised almost $200m, almost certainly not profitable

Reddit - Acquired by Conde Nast (private), might be profitable.

Zenefits - Not public, probably not profitable, raised ~$500m

Instacart - Not public, not profitable, raised ~$300m

Weebly - Not public, might be profitable

So I'd say their track record isn't great on that front. I don't know of any meaningful ($1bn+) IPOs of YC companies (not to say that they don't exist). The fact that I had no idea what Weebly was until a moment ago probably means it's not one of their success stories. I suspect DropBox will IPO at some point, but after Box's IPO, they have an uphill battle to prove that they won't be the same dud. Coinbase I had forgotten existed, so I'm really curious what their deal is (especially considering their valuation is so high, I wonder if they can support it anymore).

Thanks for providing this. I'd like to turn this into a wiki or GH repo, aggregating all of the data.

Calling Dropbox having an uphill battle to prove their valuation is quite an understatement.

Box is trading at about 6x their forward revenue in valuation, versus cloud industry's 10-12x average, and during the last Dropbox round it was rumored it was valued at 40X its forward revenue.

Meanwhile Box is winning major partnerships and customers left and right and the market still hates it, I really wonder how Dropbox is gonna tell their story considering their vast consumer user base is a liability instead of asset when it comes to profitability.

"Less than half of the startups we funded in the first 5 years of YC have died."

But how many were successful, in the sense of paying back their investors? The most likely outcome of most VC investments is a "zombie", a company which can pay its own operating systems but is a net loss to investors. The usual VC stats are that about 10% are wins, 20% go bankrupt, and the rest drag on for years in zombie mode.

I can't reply to sama's comment below, for some reason, but I would be interested to hear the median valuation of all YC backed startups. My hypothesis is that number tells a very different story.

If we are taking the amount of risk I'd like us to be taking, it should be near 0.

>Its been proved scientifically that people encounter diminising returns past 40-50 hour work weeks.

Sure, but I don't think there have been any studies that find negative returns after 40-50 hours.

If we are optimizing for maximum returns, rather than maximal efficiency, you would advocate for increasing hours worked until the result of one hour of work was zero.

Many activities have diminishing returns after a certain amount of work (oil drilling, distance running, mining bitcoin, etc.), but it doesn't mean that the rational decision is to stop when the point of diminishing returns is reached.

The diminishing returns affect all future hours, not the marginal hour. If you're tired all of your work will suck until you recover. That marginal hour might damage your ability to work more than the gain from working that one hour.

Programming in particular has negative effects after the point of being tired. You make mistakes, you commit bad code, you have to maintain bad code for the lifetime of hte product.

That one grind where you kept pushing could keep affecting the productivity of the whole engineering team for the next 2 or 3 years. I've seen it happen. A lot.

I would assume doing strategizing or negotiating when tired could equally have devastating long-term effects on the business side.

Work as long as you want, just don't work tired. It isn't worth it.

That's exactly what has been found, as cited on another comment in this discussion:

http://www.lostgarden.com/2008/09/rules-of-productivity-pres...

The cited research claims total overall productivity of working 60 hours a week starts showing a deficit to working 40 hour weeks, after about 4 weeks.

> Its been proved scientifically that people encounter diminising returns past 40-50 hour work weeks

In the average case.

People who build billion-dollar businesses are not the average case.

That isn't to say everyone should grind super hard. People should, however, have the self-awareness to recognize whether or not they are capable of high levels of achievement. Few people are.

Travis Kalanick is grinding super hard. Are people going to call him average when Uber fails? Elon Musk is grinding super hard, but he has different market and regulatory forces to contend with (no one stops you from going to space or making batteries and electric cars, everyone is going to stop you from breaking transportation and labor laws).

Being "superhuman" doesn't make you terribly more likely to succeed.

"Travis Kalanick is grinding super hard. Are people going to call him average when Uber fails?"

He'll be long gone before Uber fails. They're probably going to dump all of the liability onto the public markets, he'll probably resign to focus on other projects or some such, and leave some MBA schmuck to deal with the mess.

How many hours can someone, average or not, work before they start making more bad decisions than good ones? The problem with 50 hour+ work weeks is at some point, depending on the individuals outside life, they enter a period of sleep deprivation.

Is it a characteristic of most successful founders that they can make good decisions in high stress situations while sleep deprived?

Surgical residents work 100+ hours a week. They are working on life/death situations. Somehow they manage to not kill most people.

I am always surprised that letting residents work such long hours is acceptable given the risks involved. Does anyone want to operated on by someone who is late into their third back to back shift?

If the choice comes down to having a sleep deprived surgeon do your surgery, versus not having the surgery, then some people will choose the sleep deprived surgeon.

Some people might, but it does not seem too common for a resident surgeon to tell a patient before cutting them open something along these lines: “Hay good fellow I haven’t slept for 48 hours and I can hardly tell which end of the scalpel to hold - do you want me to operate or not?”

Perhaps the AMA shouldn't restrict the number of residents then, in order to keep surgeon salaries high. Lives > Profit.

Well there is this slight issue. Actually the bigger problem is that AMA allows surgeons to keep their license without spending a reasonable amount of time in hospitals. There are enough surgeons trained to ensure that hospitals are always staffed with well rested people, they just aren’t willing to work when they are actually needed.

So import doctors from other countries. Its okay to depress tech worker salaries using H1Bs but not doctors?

There is no need to import doctors (the USA does do a lot of this anyway). All that is needed is to make part of the registration process that you have to x number of hours in a hospital outside of 9 to 5 (surgeons though tend to start earlier). You just set x to whatever is required to ensure that you don’t have residents working 100 hour weeks.

> Somehow they manage to not kill most people.

Sure, but they're killing a lot more people than the docs getting sleep.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1705824/

"Interns working five or more extended-duration shifts per month reported more attentional failures during lectures, rounds, and clinical activities, including surgery and reported 300% more fatigue-related preventable adverse events resulting in a fatality."

I don't want to put words in Sam's mouth, but I read his point as referring to the patterns of thinking and behavior applied by the founders more than hours worked. I equate "intensity" with "focus," as I've seen (and been a part of) many startups that have lost focus after achieving a modicum of success and taking their foot off the gas.

He's not encouraging grinding harder, but smarter.

Did you read beyond the first two paragraphs?